-*- mode: org -*-
Environmental management, ecological, and remote sensing applications
*   9(<-539): Semisupervised PSO-SVM regression for biophysical parameter estimation

In this paper, a novel semisupervised regression approach is proposed
to tackle the problem of biophysical parameter estimation that is
constrained by a limited availability of training (labeled)
samples. The main objective of this approach is to increase the
accuracy of the estimation process based on the support vector machine
(SVM) technique by exploiting unlabeled samples that are available
from the image under analysis at zero cost. The integration of such
samples in the regression process is controlled through a particle
swarm optimization.(PSO) framework that is defined by considering
separately or jointly two different optimization criteria, thus
leading to the implementation of three different inflation
strategies. These two criteria are empirical and structural
expressions of the generalization capability of the resulting
semisupervised PSO-SVM regression system. The conducted experiments
were focused on the problem of estimating chlorophyll concentrations
in coastal waters from multispectral remote sensing images. In
particular, we report and discuss resu s of experiments that are
designed in such a way as to test the proposed approach in terms
of: 1) capability to capture useful information from a set of
unlabeled samples for improving the estimation accuracy; 2)
sensitivity to the number of exploited unlabeled samples; and 3)
sensitivity to the number of labeled samples used for supervising the
inflation process.

2007


*  12(<-211): Predicting groundwater level fluctuations with meteorological effect implications-A comparative study among soft computing techniques

The knowledge of groundwater table fluctuations is important in
agricultural lands as well as in the studies related to groundwater
utilization and management levels. This paper investigates the
abilities of Gene Expression Programming (GEP), Adaptive Neuro-Fuzzy
Inference System (ANFIS), Artificial Neural Networks (ANN) and Support
Vector Machine (SVM) techniques for groundwater level forecasting in
following day up to 7-day prediction intervals. Several input
combinations comprising water table level, rainfall and
evapotranspiration values from Hongcheon Well station (South Korea),
covering a period of eight years (2001-2008) were used to develop and
test the applied models. The data from the first six years were used
for developing (training) the applied models and the last two years
data were reserved for testing. A comparison was also made between the
forecasts provided by these models and the Auto-Regressive Moving
Average (ARMA) technique. Based on the comparisons, it was found that
the GEP models could be employed successfully in forecasting water
table level fluctuations up to 7 days beyond data records. (C) 2013
Elsevier Ltd. All rights reserved.

2013





*  57(<-264): Exploiting the Classification Performance of Support Vector Machines with Multi-Temporal Moderate-Resolution Imaging Spectroradiometer (MODIS) Data in Areas of Agreement and Disagreement of Existing Land Cover Products

Several studies have focused in the past on global land cover (LC)
datasets harmonization and inter-comparison and have found significant
inconsistencies. Despite the known discrepancies between existing
products derived from medium resolution satellite sensor data, little
emphasis has been placed on examining these disagreements to improve
the overall classification accuracy of future land cover maps. This
work evaluates the classification performance of a least square
support vector machine (LS-SVM) algorithm with respect to areas of
agreement and disagreement between two existing land cover maps. The
approach involves the use of time series of Moderate-resolution
Imaging Spectroradiometer (MODIS) 250-m Normalized Difference
Vegetation Index (NDVI) (16-day composites) and gridded climatic
indicators. LS-SVM is trained on reference samples obtained through
visual interpretation of Google Earth (GE) high resolution
imagery. The core of the training process is based on repeated random
splits of the training dataset to select a small set of suitable
support vectors optimizing class separability. A large number of
independent validation samples spread over three contrasting regions
in Europe (Eastern Austria, Macedonia and Southern France) are used to
calculate classification accuracies for the LS-SVM NDVI-derived LC map
and for two (globally available) LC products: GLC2000 and
GlobCover. The LS-SVM LC map reported an overall accuracy of
70%. Classification accuracies ranged from 71% where GlobCover and
GLC2000 agreed to 68% for areas of disagreement. Results indicate that
existing LC products are as accurate as the LS-SVM LC map in areas of
agreement (with little margin for improvements), while classification
accuracy is substantially better for the LS-SVM LC map in areas of
disagreement. On average, the LS-SVM LC map was 14% and 18% more
accurate compared to GlobCover and GLC2000, respectively.

2012






*  68(<- 67): A real-time forecasting model for the spatial distribution of typhoon rainfall

Accurate forecasts of hourly rainfall are necessary for early warning
systems during typhoons. In this paper, a typhoon rainfall forecasting
model is proposed to yield 1- to 6-h ahead forecasts of hourly
rainfall. First, an input optimization step integrating
multi-objective genetic algorithm (MOGA) with support vector machine
(SVM) is developed to identify the optimal input combinations. Second,
based on the results of the first step, the forecasted rainfall from
each station is used to obtain the spatial characteristics of the
rainfall process is presented. An actual application to Tsengwen river
basin is conducted to demonstrate the advantage of the proposed
model. The results clearly indicate that the proposed model
effectively improves the forecasting performance and decreases the
negative impact of increasing forecast lead time. The optimal input
combinations can be obtained from the proposed model for different
stations with different geographical conditions. In addition, the
proposed model is capable of producing more acceptable the results of
rainfall maps than other model. In conclusion, the proposed modeling
technique is useful to improve the hourly typhoon rainfall forecasting
and is expected to be helpful to support disaster warning systems. (C)
2014 Elsevier B.V. All rights reserved.

2015




* 106(<- 22): A review on sustainable construction management strategies for monitoring, diagnosing, and retrofitting the building's dynamic energy performance: Focused on the operation and maintenance phase

According to a press release, the building sector accounts for about
40% of the global primary energy consumption. Energy savings can be
achieved in the building sector by improving the building's dynamic
energy performance in terms of sustainable construction management in
the urban-based built environments (referred to as an "Urban
Organism"). This study implements the concept of "dynamic approach" to
reflect the unexpected changes in the climate and energy environments
as well as in the energy policies and technologies. Research in this
area is very significant for the future of the building, energy, and
environmental industries around the world. However, there is a lack of
studies from the perspective of the dynamic approach and the system
integration, and thus, this study is designed to fill the research
gap. This study highlights the state-of-the-art in the major phases
for a building's dynamic energy performance (i.e., monitoring,
diagnosing, and retrofitting phases), focusing on the operation and
maintenance phase. This study covers a wide range of research works
and provides various illustrative examples of the monitoring,
diagnosing, and retrofitting of a building's dynamic energy
performance. Finally, this study proposes the specific future
developments and challenges by phase and suggests the future direction
of system integration for the development of a carbon-integrated
management system as a large complex system. It is expected that
researchers and practitioners can understand and adopt the holistic
approach in the monitoring, diagnosing, and retrofitting of a
building's dynamic energy performance under the new paradigm of an
"Urban Organism". (C) 2015 Elsevier Ltd. All rights reserved.

2015




* 107(<-360): Optimization methods applied to renewable and sustainable energy: A review

Energy is a vital input for social and economic development. As a result of the generalization of agricultural, industrial and domestic activities the demand for energy has increased remarkably, especially in emergent countries. This has meant rapid grower in the level of greenhouse gas emissions and the increase in fuel prices, which are the main driving forces behind efforts to utilize renewable energy sources more effectively, i.e. energy which comes from natural resources and is also naturally replenished. Despite the obvious advantages of renewable energy, it presents important drawbacks, such as the discontinuity of generation, as most renewable energy resources depend on the climate, which is why their use requires complex design, planning and control optimization methods. Fortunately, the continuous advances in computer hardware and software are allowing researchers to deal with these optimization problems using computational resources, as can be seen in the large number of optimization methods that have been applied to the renewable and sustainable energy field. This paper presents a review of the current state of the art in computational optimization methods applied to renewable and sustainable energy, offering a clear vision of the latest research advances in this field. (C) 2010 Elsevier Ltd. All rights reserved.

2011




* 125(<-205): Development of an effective data-driven model for hourly typhoon rainfall forecasting

In this paper, we proposed a new typhoon rainfall forecasting model to
improve hourly typhoon rainfall forecasting. The proposed model
integrates multi-objective genetic algorithm with support vector
machines. In addition to the rainfall data, the meteorological
parameters are also considered. For each lead time forecasting, the
proposed model can subjectively determine the optimal combination of
input variables including rainfall and meteorological parameters. For
1- to 6-h ahead forecasts, an application to high- and low-altitude
metrological stations has shown that the proposed model yields the
best performance as compared to other models. It is found that
meteorological parameters are useful. However, the use of the optimal
combination of input variables determined by the proposed model yields
more accurate forecasts than the use of all input variables. The
proposed model can significantly improve hourly typhoon rainfall
forecasting, especially for the long lead time forecasting. (C) 2013
Elsevier B.V. All rights reserved.

2013




* 134(<-180): Interest in intermediate soft-classified maps in land change model validation: suitability versus transition potential

This study compares two types of intermediate soft-classified
maps. The first type uses land use/cover suitability maps based on a
multi-criteria evaluation (MCE). The second type focuses on the
transition potential between land use/cover categories based on a
multi-layer perceptron (MLP). The concepts and methodological
approaches are illustrated in a comparable manner using a Corine data
set from the Murcia region (2300km(2), Spain) in combination with maps
of drivers that were created with two stochastic, discretely
operating, commonly used tools (MCE in CA_MARKOV and MLP in Land
Change Modeler). The importance of the different approaches and
techniques for the obtained results is illustrated by comparing the
specific characteristics of both approaches by validating the
suitability versus transition potential maps to each other using a
Spearman correlation matrix and, between the Corine maps, using
classical ROC (receiver operating characteristic) statistics. Then, we
propose a new use of ROC statistics to compare these intermediate
soft-classified maps with their respective hard-classified maps of the
models for each category. The validation of these results can be
beneficial in choosing a suitable model and provide a better
understanding of the implications of the different modeling steps and
the advantages and limitations of the modeling tools.

2013




* 150(<-660): Land-use suitability analysis in the United States: Historical development and promising technological achievements

Various methods of spatial analysis are commonly used in land-use
plans and site selection studies. A historical overview and discussion
of contemporary developments of land-use suitability analysis are
presented. The paper begins with an exploration into the early 20th
century with the infancy of documented applications of the technique,
The article then travels through the 20th century, documenting
significant milestones. Concluding with present explorations of
advanced technologies such as neural computing and evolutionary
programming, this work is meant to serve as a foundation for
literature review and a premise for the exploration of new
advancements as we enter into the 21st century.

2001




* 162(<-646): Decision support to assist environmental sedimentology modelling

Attention is drawn to the importance of spatial aspects when adopting
a modelling approach to predict the likely character of sediment. This
requires an understanding of the processes controlling transport,
deposition and remobilization, singly and I in combination. The
advantages of incorporating expert systems are examined alongside
recently developed GIS techniques utilising multiple criteria and
fuzzy sets.

2003






* 168(<-129): A comparison study of three statistical downscaling methods and their model-averaging ensemble for precipitation downscaling in China

This study evaluated the performance of three frequently applied
statistical downscaling tools including SDSM, SVM, and LARS-WG, and
their model-averaging ensembles under diverse moisture conditions with
respect to the capability of reproducing the extremes as well as mean
behaviors of precipitation. Daily observed precipitation and NCEP
reanalysis data of 30 stations across China were collected for the
period 1961-2000, and model parameters were calibrated for each season
at individual site with 1961-1990 as the calibration period and
1991-2000 as the validation period. A flexible framework of
multi-criteria model averaging was established in which model weights
were optimized by the shuffled complex evolution algorithm. Model
performance was compared for the optimal objective and nine more
specific metrics. Results indicate that different downscaling methods
can gain diverse usefulness and weakness in simulating various
precipitation characteristics under different circumstances. SDSM
showed more adaptability by acquiring better overall performance at a
majority of the stations while LARS-WG revealed better accuracy in
modeling most of the single metrics, especially extreme indices. SVM
provided more usefulness under drier conditions, but it had less skill
in capturing temporal patterns. Optimized model averaging, aiming at
certain objective functions, can achieve a promising ensemble with
increasing model complexity and computational cost. However, the
variation of different methods' performances highlighted the tradeoff
among different criteria, which compromised the ensemble forecast in
terms of single metrics. As the superiority over single models cannot
be guaranteed, model averaging technique should be used cautiously in
precipitation downscaling.

2014





* 172(<-556): Soft combination of local models in a multi-objective framework

Conceptual hydrologic models are useful tools as they provide an
interpretable representation of the hydrologic behaviour of a
catchment. Their representation of catchment's hydrological processes
and physical characteristics, however, implies a simplification of the
complexity and heterogeneity of reality. As a result, these models may
show a lack of flexibility in reproducing the vast spectrum of
catchment responses. Hence, the accuracy in reproducing certain
aspects of the system behaviour may be paid in terms of a lack of
accuracy in the representation of other aspects. By acknowledging the
structural limitations of these models, we propose a modular approach
to hydrological simulation. Instead of using a single model to
reproduce the full range of catchment responses, multiple models are
used, each of them assigned to a specific task. While a modular
approach has been previously used in the development of data driven
models, in this study we show an application to conceptual models. The
approach is here demonstrated in the case where the different models
are associated with different parameter realizations within a fixed
model structure. We show that using a 'composite' model, obtained by a
combination of individual 'local' models, the accuracy of the
simulation is improved. We argue that this approach can be useful
because it partially overcomes the structural limitations that a
conceptual model may exhibit. The approach is shown in application to
the discharge simulation of the experimental Alzette River basin in
Luxembourg, with a conceptual model that follows the structure of the
HBV model.

2007





* 173(<- 53): OR models with stochastic components in disaster operations management: A literature survey

The increasing number of affected people due to disasters, the
complexity and unpredictability of these phenomena and the different
problems encountered in the planning and response in different
scenarios, establish a need to find better measures and practices in
order to reduce the human and economic loss in this kind of
events. However this is not an easy task considering the great
uncertainty these phenomena present including the multiple number of
possible scenarios in terms of location, probability of occurrence and
impact, the difficulty in estimating the demand and supply, the
complexity of determining the number and type of resources both
available and needed and the intricacy to establish the exact location
of the demand, the supply and the possible damaged infrastructure,
among many others. Disaster Operations Management has become very
popular and, considering the properties of disasters, the use of tools
and methodologies such as OR have been given a lot of attention in
recent years. The present work provides a literature review on the OR
models with some stochastic component applied to Disaster Operations
Management (DOM), along with an analysis of these stochastic
components and the techniques used by different authors to cope with
them as well as a detailed database on the consulted papers, which
differentiates this research from other reviews developed during the
same period, in order to give an insight on the state of the art in
the topic and determine possible future research directions. (C) 2014
Elsevier Ltd. All rights reserved.

2015



* 174(<- 91): Multi-objective ecological reservoir operation based on water quality response models and improved genetic algorithm: A case study in Three Gorges Reservoir, China

This study proposes a self-adaptive GA-aided multi-objective ecological reservoir operation model (SMEROM) and applies it to water quality management in the Xiangxi River near to the Three Gorges Reservoir, China. The SMEROM integrates statistical water quality models, multi-objective reservoir operations, and a self-adaptive GA within a general framework. Among them, the statistical water quality models of the Xiangxi River are formulated to deal with the relationships between reservoir operation and water quality, which are embedded in constraints of the SMEROM. The multiple objective functions, including maximizing hydropower generation, minimizing loss of flood control, minimizing rate of flood risk, maximizing the average remaining capacity of flood control and maximizing the benefit of shipping, are considered simultaneously to obtain comprehensive benefit among the environment, society and economy. The weighting method is employed to convert the multiple objectives to a single objective. To solve the complex SMEROM, an improved self-adaptive GA is employed through incorporating simulated binary crossover and self-adaptive mutation. To demonstrate the advantage of the developed SMEROM model, the solutions through ecological reservoir operation are compared with those through the traditional reservoir operation and the practical operation in 2011, in terms of water quality, reservoir operation and objective function values. The results show that most of benefit in the ecological operation is better than that in the traditional or practical operations except for the hydropower benefit and loss benefit of flood control. This is because flood control and environmental protection are reasonably considered in the ecological operation. (C) 2014 Elsevier Ltd. All rights reserved.

2014



* 175(<-112): Simulation-optimization modeling for conjunctive water use management

Good quality surface water and groundwater resources are limited furthermore they are shrinking because of the urbanization, contamination, and climate change impacts. In this backdrop, the proper allocation and management of these resources is a critical challenge for satisfying the rising water demands of agricultural sector. Because irrigated agriculture is the largest user of all the developed water resources and consumes over 70% of the abstracted freshwater globally. The computer-based models are useful tools for achieving the optimal allocation of limited water resources for the conjunctive use planning and management in irrigated agriculture. Various simulation and optimization modeling approaches have been used to solve the water allocation problems. Optimization models have been shown to be of great importance when used with simulation models and the combined use of these two approaches gives the best results. The reviews on the combined applications of simulation and optimization modeling for the conjunctive use planning and management of surface water and groundwater resources for sustainable irrigated agriculture are done and presented in this paper. Conclusions are provided based on this review which could be useful for all the stakeholders. (C) 2014 Elsevier B.V. All rights reserved.

2014



* 176(<-227): A Multiobjective Optimisation Model for Groundwater Remediation Design at Petroleum Contaminated Sites

This study proposes a fuzzy multi-objective model for groundwater remediation in petroleum-contaminated aquifers. The optimisation system is designed based on the PAT technology, and includes two objectives (i.e. total pumping rate and average post-remedial contaminant concentration). The relationship between pumping rates and contamination concentrations at all monitoring wells after remediation are determined by a proxy model, which integrates simulation, inference, and optimisation technologies and is composed of intercept, linear, interactive, and quadratic options. Fuzzy algorithms are used to solve the formulated multi-objective optimisation problem to find optimal solutions. The model is then applied to a petroleum-contaminated aquifer in western Canada. The trade-off and lambda analyses of the results indicate that the fuzzy multi-objective model has great potential in groundwater remediation applications as it can: (1) provide reliable groundwater remediation strategies, (2) reduce computational costs in the optimization processes, and (3) balance the trade-off between remediation costs and remediation outcomes.

2013



* 177(<-257): System optimization for eco-design by using monetization of environmental impacts: a strategy to convert bi-objective to single-objective problems

Eco-design is an essential way to reduce the environmental impacts and economic cost of processes and systems, as well as products. Until now, the majority of eco-design approaches have employed multi-objective optimization methods to balance between environmental and economic performances. However, the methods have limitations because multi-objective optimization requires decision makers to subjectively assign weighting factors for objectives, i.e., environmental impacts and economic cost. This implies that, depending on decision makers' preference and knowledge, different design solutions can be engendered for the same design problem. Thus, this study proposes an eco-design method which can generate a single design solution by developing mathematical optimization models with a single-objective function for environmental impacts and economic cost. For the formulation of the single-objective function, environmental impacts are monetized to external cost by using the Environmental Priority Strategies. This enables the tradeoffs between environmental impacts and economic cost in the same unit, i.e., monetary unit. As a case study, the proposed method is applied to the eco-design of a water reuse system in an industrial plant. This study can contribute to improving the eco-efficiency of various products, processes, and systems. (C) 2012 Elsevier Ltd. All rights reserved.

2013





* 178(<-364): Paradigm shift in urban energy systems through distributed generation: Methods and models

The path towards energy sustainability is commonly referred to the
incremental adoption of available technologies, practices and policies
that may help to decrease the environmental impact of energy sector,
while providing an adequate standard of energy services. The
evaluation of trade-offs among technologies, practices and policies
for the mitigation of environmental problems related to energy
resources depletion requires a deep knowledge of the local and global
effects of the proposed solutions. While attempting to calculate such
effects for a large complex system like a city, an advanced
multidisciplinary approach is needed to overcome difficulties in
modeling correctly real phenomena while maintaining computational
transparency, reliability, interoperability and efficiency across
different levels of analysis. Further, a methodology that rationally
integrates different computational models and techniques is necessary
to enable collaborative research in the field of optimization of
energy efficiency strategies and integration of renewable energy
systems in urban areas. For these reasons, a selection of currently
available models for distributed generation planning and design is
presented and analyzed in the perspective of gathering their
capabilities in an optimization framework to support a paradigm shift
in urban energy systems. This framework embodies the main concepts of
a local energy management system and adopts a multicriteria
perspective to determine optimal solutions for providing energy
services through distributed generation. (C) 2010 Elsevier Ltd. All
rights reserved.

2011





* 180(<-605): Multi-criteria decision analysis for the optimal management of nitrate contamination of aquifers

We present an integrated methodology for the optimal management of
nitrate contamination of ground water combining environmental
assessment and economic cost evaluation through multi-criteria
decision analysis. The proposed methodology incorporates an integrated
physical modeling framework accounting for on-ground nitrogen loading
and losses, soil nitrogen dynamics, and fate and transport of nitrate
in ground water to compute the sustainable on-ground nitrogen loading
such that the maximum contaminant level is not violated. A number of
protection alternatives to stipulate the predicted sustainable
on-ground nitrogen loading are evaluated using the decision analysis
that employs the importance order of criteria approach for ranking and
selection of the protection alternatives. The methodology was
successfully demonstrated for the Sumas-Blaine aquifer in Washington
State. The results showed the importance of using this integrated
approach which predicts the sustainable on-ground nitrogen loadings
and provides an insight into the economic consequences generated in
satisfying the environmental constraints. The results also show that
the proposed decision analysis framework, within certain limitations,
is effective when selecting alternatives with competing demands. (c)
2004 Elsevier Ltd. All rights reserved.

2005





* 182(<-284): Planning of Groundwater Supply Systems Subject to Uncertainty Using Stochastic Flow Reduced Models and Multi-Objective Evolutionary Optimization

The typical modeling approach to groundwater management relies on the combination of optimization algorithms and subsurface simulation models. In the case of groundwater supply systems, the management problem may be structured into an optimization problem to identify the pumping scheme that minimizes the total cost of the system while complying with a series of technical, economical, and hydrological constraints. Since lack of data on the subsurface system most often reflects upon the development of groundwater flow models that are inherently uncertain, the solution to the groundwater management problem should explicitly consider the tradeoff between cost optimality and the risk of not meeting the management constraints. This work addresses parameter uncertainty following a stochastic simulation (or Monte Carlo) approach, in which a sufficiently large ensemble of parameter scenarios is used to determine representative values selected from the statistical distribution of the management objectives, that is, minimizing cost while minimizing risk. In particular, the cost of the system is estimated as the expected value of the cost distribution sampled through stochastic simulation, while the risk of not meeting the management constraints is quantified as the expected value of the intensity of constraint violation. The solution to the multi-objective optimization problem is addressed by combining a multi-objective evolutionary algorithm with a stochastic model simulating groundwater flow in confined aquifers. Evolutionary algorithms are particularly appropriate in optimization problems characterized by non-linear and discontinuous objective functions and constraints, although they are also computationally demanding and require intensive analyses to tune input parameters that guarantee optimality to the solutions. In order to drastically reduce the otherwise overwhelming computational cost, a novel stochastic flow reduced model is thus developed, which practically allows for averting the direct inclusion of the full simulation model in the optimization loop. The computational efficiency of the proposed framework is such that it can be applied to problems characterized by large numbers of decision variables.

2012




* 199(<- 50): Landslide susceptibility mapping using GIS-based statistical models and Remote sensing data in tropical environment

This research presents the results of the GIS-based statistical models
for generation of landslide susceptibility mapping using geographic
information system (GIS) and remote-sensing data for Cameron Highlands
area in Malaysia. Ten factors including slope, aspect, soil,
lithology, NDVI, land cover, distance to drainage, precipitation,
distance to fault, and distance to road were extracted from SAR data,
SPOT 5 and WorldView-1 images. The relationships between the detected
landslide locations and these ten related factors were identified by
using GIS-based statistical models including analytical hierarchy
process (AHP), weighted linear combination (WLC) and spatial
multi-criteria evaluation (SMCE) models. The landslide inventory map
which has a total of 92 landslide locations was created based on
numerous resources such as digital aerial photographs, AIRSAR data,
WorldView-1 images, and field surveys. Then, 80% of the landslide
inventory was used for training the statistical models and the
remaining 20% was used for validation purpose. The validation results
using the Relative landslide density index (R-index) and Receiver
operating characteristic (ROC) demonstrated that the SMCE model
(accuracy is 96%) is better in prediction than AHP (accuracy is 91%)
and WLC (accuracy is 89%) models. These landslide susceptibility maps
would be useful for hazard mitigation purpose and regional planning.

2015



* 200(<- 82): GIS-based multicriteria evaluation with multiscale analysis to characterize urban landslide susceptibility in data-scarce environments

Landslides can have a severe negative impact on the socio-economic and
environmental state of individuals and their communities. Minimizing
these impacts is dependent on the effective identification of risk
areas using a susceptibility analysis process. In such a process,
output maps are generated to determine various levels of threat to
human populations. However, the reliability of the process is
controlled by critical factors such as data availability and data
quality. In data-scarce environments, susceptibility analysis done at
multiple interlocking geographic scales can provide a convergence of
evidence to reliably identify risk areas. In this study, multiscale
analysis and fuzzy sets are combined with GIS-based multicriteria
evaluation (MCE) to determine landslide susceptibility levels for
areas of the Metro Vancouver region, British Columbia,
Canada. Landslide-conditioning parameters are chosen based on their
relevance and effect on a particular scale of analysis. These
parameters are derived for three geographic scales using digital
elevation models, drainage networks and road networks. An analytical
hierarchy process (AHP) analysis provides relative weights of
importance to combine variables. The landslide susceptibility analysis
is done for regional, municipal and local scales at resolutions of 50
m, 10 m, and 1 m respectively. At each scale, susceptibility models
are validated against real inventory data using the seed cell area
index (SCAI) method. The strong inverse correlation between the map
classes and the SCAI adds to confidence in the results. The developed
approach can enable analysts in data-scarce environments to reliably
identify susceptible areas thereby improving hazard mitigation,
emergency services targeting, and overall community planning. (C) 2014
Elsevier Ltd. All rights reserved.

2015



* 201(<-123): Landslide susceptibility mapping using GIS-based multi-criteria decision analysis, support vector machines, and logistic regression

Identification of landslides and production of landslide susceptibility maps are crucial steps that can help planners, local administrations, and decision makers in disaster planning. Accuracy of the landslide susceptibility maps is important for reducing the losses of life and property. Models used for landslide susceptibility mapping require a combination of various factors describing features of the terrain and meteorological conditions. Many algorithms have been developed and applied in the literature to increase the accuracy of landslide susceptibility maps. In recent years, geographic information system-based multi-criteria decision analyses (MCDA) and support vector regression (SVR) have been successfully applied in the production of landslide susceptibility maps. In this study, the MCDA and SVR methods were employed to assess the shallow landslide susceptibility of Trabzon province (NE Turkey) using lithology, slope, land cover, aspect, topographic wetness index, drainage density, slope length, elevation, and distance to road as input data. Performances of the methods were compared with that of widely used logistic regression model using ROC and success rate curves. Results showed that the MCDA and SVR outperformed the conventional logistic regression method in the mapping of shallow landslides. Therefore, multi-criteria decision method and support vector regression were employed to determine potential landslide zones in the study area.

2014



* 202(<-128): GIS-based landslide susceptibility mapping with probabilistic likelihood ratio and spatial multi-criteria evaluation models (North of Tehran, Iran)

The aim of this study is to produce landslide susceptibility mapping by probabilistic likelihood ratio (PLR) and spatial multi-criteria evaluation (SMCE) models based on geographic information system (GIS) in the north of Tehran metropolitan, Iran. The landslide locations in the study area were identified by interpretation of aerial photographs, satellite images, and field surveys. In order to generate the necessary factors for the SMCE approach, remote sensing and GIS integrated techniques were applied in the study area. Conditioning factors such as slope degree, slope aspect, altitude, plan curvature, profile curvature, surface area ratio, topographic position index, topographic wetness index, stream power index, slope length, lithology, land use, normalized difference vegetation index, distance from faults, distance from rivers, distance from roads, and drainage density are used for landslide susceptibility mapping. Of 528 landslide locations, 70 % were used in landslide susceptibility mapping, and the remaining 30 % were used for validation of the maps. Using the above conditioning factors, landslide susceptibility was calculated using SMCE and PLR models, and the results were plotted in ILWIS-GIS. Finally, the two landslide susceptibility maps were validated using receiver operating characteristic curves and seed cell area index methods. The validation results showed that area under the curve for SMCE and PLR models is 76.16 and 80.98 %, respectively. The results obtained in this study also showed that the probabilistic likelihood ratio model performed slightly better than the spatial multi-criteria evaluation. These landslide susceptibility maps can be used for preliminary land use planning and hazard mitigation purpose.

2014


* 203(<-130): GIS-based groundwater spring potential assessment and mapping in the Birjand Township, southern Khorasan Province, Iran

Three statistical models-frequency ratio (FR), weights-of-evidence (WofE) and logistic regression (LR)-produced groundwater-spring potential maps for the Birjand Township, southern Khorasan Province, Iran. In total, 304 springs were identified in a field survey and mapped in a geographic information system (GIS), out of which 212 spring locations were randomly selected to be modeled and the remaining 92 were used for the model evaluation. The effective factors-slope angle, slope aspect, elevation, topographic wetness index (TWI), stream power index (SPI), slope length (LS), plan curvature, lithology, land use, and distance to river, road, fault-were derived from the spatial database. Using these effective factors, groundwater spring potential was calculated using the three models, and the results were plotted in ArcGIS. The receiver operating characteristic (ROC) curves were drawn for spring potential maps and the area under the curve (AUC) was computed. The final results indicated that the FR model (AUC = 79.38 %) performed better than the WofE (AUC = 75.69 %) and LR (AUC = 63.71 %) models. Sensitivity and factor analyses concluded that the bivariate statistical index model (i.e. FR) can be used as a simple tool in the assessment of groundwater spring potential when a sufficient number of data are obtained.

2014


* 204(<-299): A comparison of landslide susceptibility maps produced by logistic regression, multi-criteria decision, and likelihood ratio methods: a case study at Izmir, Turkey

The main purpose of this study is to compare the use of logistic regression, multi-criteria decision analysis, and a likelihood ratio model to map landslide susceptibility in and around the city of Izmir in western Turkey. Parameters, such as lithology, slope gradient, slope aspect, faults, drainage lines, and roads, were considered. Landslide susceptibility maps were produced using each of the three methods and then compared and validated. Before the modeling and validation, the observed landslides were separated into two groups. The first group was for training, and the other group was for validation steps. The accuracy of models was measured by fitting them to a validation set of observed landslides. For validation process, the area under curvature (AUC) approach was applied. According to the AUC values of 0.810, 0.764, and 0.710 for logistic regression, likelihood ratio, and multi-criteria decision analysis, respectively, logistic regression was determined to be the most accurate method among the other used landslide susceptibility mapping methods. Based on these results, logistic regression and likelihood ratio models can be used to mitigate hazards related to landslides and to aid in land-use planning.

2012


* 205(<-407): Landslide susceptibility mapping for Ayvalik (Western Turkey) and its vicinity by multicriteria decision analysis

This paper presents the results of geographical information system (GIS)-based landslide susceptibility mapping in AyvalA +/- k, western Turkey using multi-criteria decision analysis. The methodology followed in the study includes data production, standardization, and analysis stages. A landslide inventory of the study area was compiled from aerial photographs, satellite image interpretations, and detailed field surveys. In total, 45 landslides were recorded and mapped. The areal extent of the landslides is 1.75 km(2). The identified landslides are mostly shallow-seated, and generally exhibit progressive character. They are mainly classified as rotational, planar, and toppling failures. In all, 51, 45, and 4% of the landslides mapped are rotational, planar, and toppling types, respectively. Morphological, geological, and land-use data were produced using existing topographical and relevant thematic maps in a GIS framework. The considered landslide-conditioning parameters were slope gradient, slope aspect, lithology, weathering state of the rocks, stream power index, topographical wetness index, distance from drainage, lineament density, and land-cover and vegetation density. These landslide parameters were standardized in a common data scale by fuzzy membership functions. Then, the degree to which each parameter contributed to landslides was determined using the analytical hierarchy process method, and the weight values of these parameters were calculated. The weight values obtained were assigned to the corresponding parameters, and then the weighted parameters were combined to produce a landslide susceptibility map. The results obtained from the susceptibility map were evaluated with the landslide location data to assess the reliability of the map. Based on the findings obtained in this study, it was found that 5.19% of the total area was prone to landsliding due to the existence of highly and completely weathered lithologic units and due to the adverse effects of topography and improper land use.

2010


* 206(<-510): Landslide susceptibility mapping for a landslide-prone area (Findikli, NE of Turkey) by likelihood-frequency ratio and weighted linear combination models

Landslides are very common natural problems in the Black Sea Region of Turkey due to the steep topography, improper use of land cover and adverse climatic conditions for landslides. In the western part of region, many studies have been carried out especially in the last decade for landslide susceptibility mapping using different evaluation methods such as deterministic approach, landslide distribution, qualitative, statistical and distribution-free analyses. The purpose of this study is to produce landslide susceptibility maps of a landslide-prone area (Findikli district, Rize) located at the eastern part of the Black Sea Region of Turkey by likelihood frequency ratio (LRM) model and weighted linear combination (WLC) model and to compare the results obtained. For this purpose, landslide inventory map of the area were prepared for the years of 1983 and 1995 by detailed field surveys and aerial-photography studies. Slope angle, slope aspect, lithology, distance from drainage lines, distance from roads and the land-cover of the study area are considered as the landslide-conditioning parameters. The differences between the susceptibility maps derived by the LRM and the WLC models are relatively minor when broad-based classifications are taken into account. However, the WLC map showed more details but the other map produced by LRM model produced weak results. The reason for this result is considered to be the fact that the majority of pixels in the LRM map have high values than the WLC-derived susceptibility map. In order to validate the two susceptibility maps, both of them were compared with the landslide inventory map. Although the landslides do not exist in the very high susceptibility class of the both maps, 79% of the landslides fall into the high and very high susceptibility zones of the WLC map while this is 49% for the LRM map. This shows that the WLC model exhibited higher performance than the LRM model.

2008



* 208(<- 69): Improvement in estimation of soil water retention using fractal parameters and multiobjective group method of data handling

Soil water retention characteristic is required for modeling of water
and substance movement in unsaturated soils and need to be estimated
using indirect methods. Point pedotransfer functions (PTFs) for
prediction of soil water content at matric suctions of 1, 5, 25, 50,
and 1500kPa were developed and validated using a data-set of 148 soil
samples from Hamedan and Guilan provinces, Iran, by multiobjective
group method of data handling (mGMDH). In addition to textural and
structural properties, fractal parameters of the power-law fractal
models for both particles and aggregates distributions were also
included as predictors. Their inclusion significantly improved the
PTFs' accuracy and reliability. The aggregate size distribution
fractal parameters ranked next to the particle size distribution (PSD)
in terms of prediction accuracy. The mGMDH-derived PTFs were
significantly more reliable than those by artificial neural networks
but their accuracies were practically the same. Similarity between the
fractal behavior of particle and void size distributions may
contribute to the improvement of the derived PTFs using PSD fractal
parameters. It means that both distributions of the pore and particle
size represent the fractal behavior and can be described by fractal
models.

2015




* 209(<-251): Artificial Bee Colony approach to information granulation-based fuzzy radial basis function neural networks for image fusion

This paper mainly proposed a novel method of Artificial Bee Colony
(ABC) optimized fuzzy radial basis function neural networks with
information granulation (IG-FRBFNNs) for solving the image fusion
problem. Image fusion is the process of combining relevant information
from two or more images into a single image. The fuzzy RBF neural
networks exploit the Fuzzy C-Means (FCM) clustering to form the
premise part of the rules. As the consequent part of the model (being
the local model representing input output relation in the
corresponding sub-space), four types of polynomials are considered,
with the ordinary least square (OLS) learning being exploited to
estimate the values of the coefficients of the polynomial. Since the
performance of the IG-FRBFNN model is directly affected by the
parameters such as the fuzzification coefficient used in the FCM, the
position of their centers and the values of the widths, ABC algorithm
is exploited to carry out the structural and parametric optimization
of the model respectively while the optimization is of multi-objective
character as it is aimed at the simultaneous minimization of
complexity and maximization of accuracy. Subsequently, the proposed
approach can dynamically obtain optimal image fusion weights based on
regional features, so as to optimize performance of image
fusion. Series of experimental results are presented to verify the
feasibility and effectiveness of the proposed approach. (C) 2012
Elsevier GmbH. All rights reserved.

2013



* 210(<-260): A Neural Network Based Intelligent Predictive Sensor for Cloudiness, Solar Radiation and Air Temperature

Accurate measurements of global solar radiation and atmospheric
temperature, as well as the availability of the predictions of their
evolution over time, are important for different areas of
applications, such as agriculture, renewable energy and energy
management, or thermal comfort in buildings. For this reason, an
intelligent, light-weight and portable sensor was developed, using
artificial neural network models as the time-series predictor
mechanisms. These have been identified with the aid of a procedure
based on the multi-objective genetic algorithm. As cloudiness is the
most significant factor affecting the solar radiation reaching a
particular location on the Earth surface, it has great impact on the
performance of predictive solar radiation models for that
location. This work also represents one step towards the improvement
of such models by using ground-to-sky hemispherical colour digital
images as a means to estimate cloudiness by the fraction of visible
sky corresponding to clouds and to clear sky. The implementation of
predictive models in the prototype has been validated and the system
is able to function reliably, providing measurements and four-hour
forecasts of cloudiness, solar radiation and air temperature.

2012





* 214(<-230): Evaluation of several rainfall products used for hydrological applications over West Africa using two high-resolution gauge networks

The evaluation of rainfall products over the West African region will
be an important component of the Megha-Tropiques (MT) Ground
Validation (GV) plan. In this paper, two dense research gauge networks
from Benin and Niger, integrated in the MT GV plan, are presented and
are used to evaluate several currently available global or regional
satellite-based rainfall products. Eight productsthe Precipitation
Estimation from Remotely Sensed Information using Artificial Neural
Networks (PERSIANN), Climate Prediction Center Morphing method
(CMORPH), Tropical Rainfall Measuring Mission (TRMM) Multi-satellite
Precipitation Analysis (TMPA) 3B42 real-time and gauge-adjusted
version, Global Satellite Mapping of Precipitation (GSMaP), Climate
Prediction Center (CPC) African Rainfall Estimate (RFE), Estimation
des Precipitation par SATellite (EPSAT), and Global Precipitation
Climatology Project One Degree Daily estimate (GPCP-1DD)are compared
to the ground reference. The comparisons are carried out at daily, 1
degrees resolution, over the rainy season (June-September), between
the years 2003 and 2010. The work focuses on the ability of the
various products to reproduce salient features of the rainfall regime
that impact the hydrological response. The products are analysed on a
multi-criteria basis, focusing in particular on the way they
distribute the rainfall within the season and by rain rate
class. Standard statistical diagnoses such as the correlation
coefficient, bias, root mean square error and Nash skill score are
computed and the inter-annual variability is documented. Two
simplified hydrological models are used to illustrate how the nature
and structure of the product error impact the model output in terms of
runoff (calculated using the Soil Conservation Service method, SCS, in
Niger) or outflow (calculated with the modele du Genie Rural a 4
parametres Journalier', GR4J model, in Benin). Copyright (c) 2013
Royal Meteorological Society

2013


* 215(<-503): Regionalisation of hydrological model parameters under parameter uncertainty: A case study involving TOPMODEL and basins across the globe

In this paper, we present a method to account for modeling uncertainties while regionalising model parameters. Linking model. parameters to physical catchment attributes is a popular approach that enables the application of a conceptual model to an ungauged site. The functional relationship can be derived either from the calibrated model. parameters (direct calibration method) or by calibrating the functional. function (regional. calibration method). Both of these approaches are explored through a case study involving TOPMODEL and a number of small- to medium-sized humid basins located in various geographic and climatic regions around the globe. The predictive performance of the functional relationship derived using the direct calibration method (e.g., multiple regression, artificial neural. network and partial least square regression) varied among the different schemes. However, the average of the model. parameters estimated from regionatisation schemes based on direct calibration is found to be a better surrogate. Even with the use of a parsimonious hydrological model and with posing model calibration as a multi-objective problem, the model. parameter uncertainty and its effect on model. prediction were observed to be high and varied among the basins. Therefore, to avoid the effect of model parameter uncertainty on regionalization results, a regional calibration method that skips direct calibration of the hydrological model was implemented. This method was improved in order to take into account multiple objective criteria white calibrating regional parameters. The predictive performance of the improved regional calibration method was found to be superior to the direct calibration method, indicating that the identifiability of model parameters has an apparent effect on deriving predictive models for regionalisation. However, the regional calibration method was unable to uniquely identify the regional relationship, and the modeling uncertainties quantified using Pareto optimal regional relationships were considerable. Regionalisation schemes that are based on direct calibration do not explicitly account for the modeling uncertainties. Therefore, to account for these uncertainties in model parameters and regionalisation schemes, methods based on regionalisation of vectors of model parameters (i.e. regionalizing the vectors of equally likely values of model parameters) and posterior probability distribution of model parameters (i.e. estimating the posterior probability distribution of model parameters at ungauged sites by linking the entries of model parameters' covariance matrix and the posterior mean of model parameter to the catchment attributes) are introduced. The uncertainties in model prediction as quantified from both methods closely followed the prediction uncertainties quantified from calibrated posterior probability distributions of model parameters. Moreover, though the prediction uncertainties associated with the regional calibration method as quantified from the Pareto optimal regional relationship were comparatively higher than those obtained from the direct calibration schemes, they were in close agreement with the prediction uncertainties quantified from the calibrated posterior probability distribution. The ensemble of simulated flows realized from the model parameters sampled from regionalized posterior probability distributions for five ungauged basins are also presented as validation of the proposed methodology. (c) 2008 Elsevier B.V. All rights reserved.

2008


* 216(<- 19): Neural network river forecasting through baseflow separation and binary-coded swarm optimization

The inclusion of expert knowledge in data-driven streamflow modeling is expected to yield more accurate estimates of river quantities. Modular models (MMs) designed to work on different parts of the hydrograph are preferred ways to implement such approach. Previous studies have suggested that better predictions of total streamflow could be obtained via modular Artificial Neural Networks (ANNs) trained to perform an implicit baseflow separation. These MMs fit separately the baseflow and excess flow components as produced by a digital filter, and reconstruct the total flow by adding these two signals at the output. The optimization of the filter parameters and ANN architectures is carried out through global search techniques. Despite the favorable premises, the real effectiveness of such MMs has been tested only on a few case studies, and the quality of the baseflow separation they perform has never been thoroughly assessed. In this work, we compare the performance of MM against global models (GMs) for nine different gaging stations in the northern United States. Binary-coded swarm optimization is employed for the identification of filter parameters and model structure, while Extreme Learning Machines, instead of ANN, are used to drastically reduce the large computational times required to perform the experiments. The results show that there is no evidence that MM outperform global GM for predicting the total flow. In addition, the baseflow produced by the MM largely underestimates the actual baseflow component expected for most of the considered gages. This occurs because the values of the filter parameters maximizing overall accuracy do not reflect the geological characteristics of the river basins. The results indeed show that setting the filter parameters according to expert knowledge results in accurate baseflow separation but lower accuracy of total flow predictions, suggesting that these two objectives are intrinsically conflicting rather than compatible. (C) 2015 Elsevier B.V. All rights reserved.

2015




* 219(<-483): Improved irrigation water demand forecasting using a soft-computing hybrid model

Recently, Computational Neural Networks (CNNs) and fuzzy inference
systems have been successfully applied to time series forecasting. In
this study the performance of a hybrid methodology combining feed
forward CNN, fuzzy logic and genetic algorithm to forecast one-day
ahead daily water demands at irrigation districts considering that
only flows in previous days are available for the calibration of the
models were analysed. Individual forecasting models were developed
using historical time series data from the Fuente Palmera irrigation
district located in Andalucia, southern Spain. These models included
univariate autoregressive CNNs trained with the Levenberg-Marquardt
algorithm (LM). The individual models forecasting were then corrected
via a fuzzy logic approach whose parameters were adjusted using a
genetic algorithm in order to improve the forecasting accuracy. For
the purpose of comparison, this hybrid methodology was also applied
with univariate autoregressive CNN models trained with the
Extended-Delta-Bar-Delta algorithm (EDBD) and calibrated in a previous
study in the same irrigation district. A multicriteria evaluation with
several statistics and absolute error measures showed that the hybrid
model performed significantly better than univariate and multivariate
autoregressive CNN's. (C) 2008 IAgrE. Published by Elsevier Ltd. All
rights reserved.

2009




* 223(<-225): Comparison of Artificial Neural Network Methods with L-moments for Estimating Flood Flow at Ungauged Sites: the Case of East Mediterranean River Basin, Turkey

A regional flood frequency analysis based on the index flood method is
applied using probability distributions commonly utilized for this
purpose. The distribution parameters are calculated by the method of
L-moments with the data of the annual flood peaks series recorded at
gauging sections of 13 unregulated natural streams in the East
Mediterranean River Basin in Turkey. The artificial neural networks
(ANNs) models of (1) the multi-layer perceptrons (MLP) neural
networks, (2) radial basis function based neural networks (RBNN), and
(3) generalized regression neural networks (GRNN) are developed as
alternatives to the L-moments method. Multiple-linear and
multiple-nonlinear regression models (MLR and MNLR) are also used in
the study. The L-moments analysis on these 13 annual flood peaks
series indicates that the East Mediterranean River Basin is
hydrologically homogeneous as a whole. Among the tried distributions
which are the Generalized Logistic, Generalized Extreme Vaules,
Generalized Normal, Pearson Type III, Wakeby, and Generalized Pareto,
the Generalized Logistic and Generalized Extreme Values distributions
pass the Z statistic goodness-of-fit test of the L-moments method for
the East Mediterranean River Basin, the former performing yet better
than the latter. Hence, as the outcome of the L-moments method applied
by the Generalized Logistic distribution, two equations are developed
to estimate flood peaks of any return periods for any un-gauged site
in the study region. The ANNs, MLR and MNLR models are trained and
tested using the data of these 13 gauged sites. The results show that
the predicting performance of the MLP model is superior to the
others. The application of the MLP model is performed by a special
Matlab code, which yields logarithm of the flood peak, Ln(Q(T)),
versus a desired return period, T.

2013



* 224(<-240): Prediction of the baseline toxicity of non-polar narcotic chemical mixtures by QSAR approach

Environmental contaminants are frequently encountered as mixtures, and
research on mixture toxicity is a hot topic until now. In the present
study, the mixture toxicity of non-polar narcotic chemical was modeled
by linear and nonlinear statistical methods, that is to say, by
forward stepwise multilinear regression (MLR) and radial basis
function neural networks (RBFNNs) from molecular descriptors that are
calculated and be defined as composite descriptors according to the
fractional concentrations of the mixture components. The statistical
parameters provided by the MLR model were R-2 = 0.9512, RMS = 0.3792,
F = 1402.214 and LOOq(2) = 0.9462 for the training set, and R-2 =
0.9453, RMS = 03458, F = 276.671 and q(ext)(2) = 0.9450 for the
external test set. The RBFNN model gave the following statistical
results, namely: R-2 = 0.9779, RMS = 0.2498, F = 3188.202 and LOOq(2)
= 0.9746 for the training set, and R-2 = 0.9763, RMS = 0.2358, F =
660.631 and q(ext)(2) = 0.9745, for the external test set. Overall,
these results suggest that the QSAR MLR-based model is a simple,
reliable, credible and fast tool for the prediction mixture toxicity
of non-polar narcotic chemicals. The RBFNN model gave even improved
results. In addition, epsilon(LUMO+1) (the energy of the second lowest
unoccupied molecular orbital) and PPSA (total charge weighted partial
positively surface area) were found to have high correlation with the
mixture toxicity. (c) 2012 Elsevier Ltd. All rights reserved.

2013


* 225(<-249): Landslide susceptibility estimation by random forests technique: sensitivity and scaling issues

Despite the large number of recent advances and developments in landslide susceptibility mapping (LSM) there is still a lack of studies focusing on specific aspects of LSM model sensitivity. For example, the influence of factors such as the survey scale of the landslide conditioning variables (LCVs), the resolution of the mapping unit (MUR) and the optimal number and ranking of LCVs have never been investigated analytically, especially on large data sets. In this paper we attempt this experimentation concentrating on the impact of model tuning choice on the final result, rather than on the comparison of methodologies. To this end, we adopt a simple implementation of the random forest (RF), a machine learning technique, to produce an ensemble of landslide susceptibility maps for a set of different model settings, input data types and scales. Random forest is a combination of Bayesian trees that relates a set of predictors to the actual landslide occurrence. Being it a nonparametric model, it is possible to incorporate a range of numerical or categorical data layers and there is no need to select unimodal training data as for example in linear discriminant analysis. Many widely acknowledged landslide predisposing factors are taken into account as mainly related to the lithology, the land use, the geomorphology, the structural and anthropogenic constraints. In addition, for each factor we also include in the predictors set a measure of the standard deviation (for numerical variables) or the variety (for categorical ones) over the map unit. As in other systems, the use of RF enables one to estimate the relative importance of the single input parameters and to select the optimal configuration of the classification model. The model is initially applied using the complete set of input variables, then an iterative process is implemented and progressively smaller subsets of the parameter space are considered. The impact of scale and accuracy of input variables, as well as the effect of the random component of the RF model on the susceptibility results, are also examined. The model is tested in the Arno River basin (central Italy). We find that the dimension of parameter space, the mapping unit (scale) and the training process strongly influence the classification accuracy and the prediction process. This, in turn, implies that a careful sensitivity analysis making use of traditional and new tools should always be performed before producing final susceptibility maps at all levels and scales.

2013





* 226(<-423): Neural Networks for the Prediction of Species-Specific Plot Volumes Using Airborne Laser Scanning and Aerial Photographs

Parametric and nonparametric modeling methods have been widely used
for the estimation of forest attributes from airborne laser-scanning
data and aerial photographs. However, the methods adopted suffered
from complex remote-sensed data structures involving high dimensions,
nonlinear relationships, different statistical distributions, and
outliers. In this context, artificial neural networks (ANNs) are of
interest as they have many clear benefits over conventional modeling
methods and could then enhance the accuracy of current
forest-inventory methods. This paper examines the ability of common
ANN modeling techniques for the prediction of species-specific forest
attributes, as exemplified here with the prediction stem volumes
(cubic meters per hectare) at the field plot and forest stand
levels. Three modeling methods were evaluated, namely, the multilayer
perceptron (MLP), support vector regression (SVR), and self-organizing
map, and intercompared with the corresponding nonparametric k most
similar neighbor method using cross-validated statistical performance
indexes. To decrease the number of model-input variables, a
multiobjective input-selection method based on genetic algorithm is
adopted. The numerical results obtained in the study suggest that ANNs
are appropriate and accurate methods for the assessment of
species-specific forest attributes, which can be used as alternatives
to multivariate linear regression and nonparametric nearest neighbor
models. Among the ANN models, SVR and MLP provide the best choices for
prediction purposes as they yielded high prediction accuracies for
species-specific tree volumes throughout.

2010






* 228(<-523): Incorporating anthropogenic variables into a species distribution model to map gypsy moth risk

This paper presents a novel methodology for multi-scale and multi-type
spatial data integration in support of insect pest risk/vulnerability
assessment in the contiguous United States. Probability of gypsy moth
(Lymantria dispar L.) establishment is used as a case study. A neural
network facilitates the integration of variables representing dynamic
anthropogenic interaction and ecological characteristics. Neural
network model (back-propagation network [BPN]) results are compared to
logistic regression and multi-criteria evaluation via weighted linear
combination, using the receiver operating characteristic area under
the curve (AUC) and a simple threshold assessment. The BPN provided
the most accurate infestation-forecast predictions producing an AUC of
0.93, followed by multi-criteria evaluation (AUC = 0.92) and logistic
regression (AUC = 0.86) when independently validating using post model
infestation data. Results suggest that BPN can provide valuable
insight into factors contributing to introduction for invasive species
whose propagation and establishment requirements are not fully
understood. The integration of anthropogenic and ecological variables
allowed production of an accurate risk model and provided insight into
the impact of human activities. (C) 2007 Elsevier B.V. All rights
reserved.

2008



* 229(<-525): Study of the potential of alternative crops by integration of multisource data using a neuro-fuzzy technique

This work proposes a neuro-fuzzy method for suggesting alternative
crop production over a region using integrated data obtained from
land-survey maps as well as satellite imagery. The methodology
proposed here uses an artificial neural network (multilayer
perceptron, MLP) to predict alternative crop production. For each
pixel, the MLP takes vector input comprising elevation, rainfall and
goodness values of different existing crops. The first two components
of the aforementioned input, that is, elevation and rainfall, are
determined from contour information of land-survey maps. The other
components, such as goodness values of different existing crops, are
based on the productivity estimates of soil determined by
fuzzyfication and expert opinion (on soil) along with production
quality by the Normalized Difference Vegetation Index (NDVI) obtained
from satellite imagery. The methodology attempts to ensure that the
suggested crop will also be a high productivity crop for that region.

2008





* 230(<-595): Evaluation of an integrated modelling system containing a multi-layer perceptron model and the numerical weather prediction model HIRLAM for the forecasting of urban airborne pollutant concentrations

In this paper, a multi-layer perceptron (MLP) model and the Finnish variant of the numerical weather prediction model HIRLAM (High Resolution Limited Area Model) were integrated and evaluated for the forecasting in time of urban pollutant concentrations. The forecasts of the combination of the MLP and HIRLAM models are compared with the corresponding forecasts of the MLP models that utilise meteorologically pre-processed input data. A. novel input selection method based on the use of a multi-objective genetic algorithm (MOGA) is applied in conjunction with the sensitivity analysis to reduce the excessively large number of potential meteorological input variables; its use improves the performance of the MLP model. The computed air quality forecasts contain the sequential hourly time series of the concentrations of nitrogen dioxide (NO2) and fine particulate matter (PM2.5) from May 2000 to April 2003; the corresponding concentrations have also been measured at two urban air quality stations in Helsinki. The results obtained with the MLP models that use HIRLAM forecasts show fairly good overall agreement for both pollutants. The model performance is substantially better, when the HIRLAM forecasts are used, compared with those obtained both using either HIRLAM analysis data or meteorological pre-processor, for both pollutants. The performance of the currently widely used statistical forecasting methods (such as those based on neural networks) could therefore be significantly improved by using the forecasts of NWP models, instead of the conventionally utilised directly measured or meteorological pre-processed input data. However, the performance of all operational models considered is relatively worse in the course of air pollution episodes. (c) 2005 Elsevier Ltd. All rights reserved.

2005



* 231(<- 18): Data-driven input variable selection for rainfall-runoff modeling using binary-coded particle swarm optimization and Extreme Learning Machines

Selecting an adequate set of inputs is a critical step for successful data-driven streamflow prediction. In this study, we present a novel approach for Input Variable Selection (IVS) that employs Binary-coded discrete Fully Informed Particle Swarm optimization (BFIPS) and Extreme Learning Machines (ELM) to develop fast and accurate IVS algorithms. A scheme is employed to encode the subset of selected inputs and ELM specifications into the binary particles, which are evolved using single objective and multi-objective BFIPS optimization (MBFIPS). The performances of these ELM-based methods are assessed using the evaluation criteria and the datasets included in the comprehensive IVS evaluation framework proposed by Galelli et al. (2014). From a comparison with 4 major IVS techniques used in their original study it emerges that the proposed methods compare very well in terms of selection accuracy. The best performers were found to be (1) a MBFIPS-ELM algorithm based on the concurrent minimization of an error function and the number of selected inputs, and (2) a MBFIPS-ELM algorithm based on the minimization of a variant of the Akaike Information Criterion (AIC). The first technique is arguably the most accurate overall, and is able to reach an almost perfect specification of the optimal input subset for a partially synthetic rainfall-runoff experiment devised for the Kentucky River basin. In addition, MBFIPS-ELM allows for the determination of the relative importance of the selected inputs. On the other hand, the BFIPS-ELM is found to consistently reach high accuracy scores while being considerably faster. By extrapolating the results obtained on the IVS test-bed, it can be concluded that the proposed techniques are particularly suited for rainfall-runoff modeling applications characterized by high nonlinearity in the catchment dynamics. (C) 2015 Elsevier B.V. All rights reserved.

2015


* 232(<- 34): Improving event-based rainfall-runoff simulation using an ensemble artificial neural network based hybrid data-driven model

An ensemble artificial neural network (ENN) based hybrid function approximator (named PEK), integrating the partial mutual information (PMI) based separate input variable selection (IVS) scheme, ENN-based output estimation, and K-nearest neighbor regression based output error estimation, has been proposed to improve event-based rainfall-runoff (RR) simulation. A hybrid data-driven RR model, named non-updating PEK (NU-PEK), is also developed on the basis of the PEK approximator. The rainfall and simulated antecedent discharges input variables for the NU-PEK model are selected separately by using a PMI-based IVS algorithm. A newly proposed candidate rainfall input set, sliding window cumulative rainfall is also proposed. These two methods are integrated to make a good compromise between the adequacy and parsimony of the input information and make contribution to the understandings of the hydrologic responses to the regional precipitation. The number of component networks and the topology and parameter settings of each component network are optimized simultaneously by using the multi-objective NSGA-II optimization algorithm and the early stopping Levenberg-Marquardt algorithm. The optimal combination weights of the ENN are obtained according to the Akaike information criterions of component networks. By combining all these methods, the simulation accuracy and generalization property of the PEK approximator are much better than traditional artificial neural network. The NU-PEK model is constructed by combining the PEK approximator with a newly proposed non-updating modeling approach to improve event-based RR simulation. The NU-PEK model was applied to three Chinese catchments for RR simulation and compared with two popular RR models, including the conceptual Xinanjiang model and the conceptual-data-driven IHACRES model. The results of simulation and sensitivity analysis indicate that the developed model generally outperforms the other two models. The NU-PEK model is capable of producing high accuracy non-updating RR simulation without the use of the real-time information, e.g. the observed discharges at previous time steps.

2015


* 233(<-103): Improved Neural Network Model and Its Application in Hydrological Simulation

When applying a back-propagation neural network (BPNN) model in hydrological simulation, researchers generally face three problems. The first one is that real-time correction mode must be adopted when forecasting basin outlet flow, i.e., observed antecedent outlet flows must be utilized as part of the inputs of the BPNN model. Under this mode, outlet flow can only be forecasted one time step ahead, i.e., continuous simulation cannot be implemented. The second one is that topology, weights, and biases of BPNN cannot be optimized simultaneously by traditional training methods. Topology designed by the trial-and-error method and weights and biases trained by back-propagation (BP) algorithm are not always global optimal and the optimizations are experience-based. The third one is that simulation accuracy for the validation period is usually much lower than that for the calibration period, i.e., generalization property of BPNN is not good. To solve these problems, a novel coupled black-box model named BK (BP-KNN) and a new methodology of calibration are proposed in this paper. The BK model was developed by coupling BPNN model with K-nearest neighbor (KNN) algorithm. Unlike the traditional BPNN model previously reported, the BK model implemented continuous simulation under nonreal-time correction mode. Observed antecedent outlet flows were substituted by simulated values. The simulated values were calculated by the BPNN model first and then corrected based on the KNN algorithm, historical simulation error, and other relevant factors. According to the calculation process, parameters of the BK model were divided into three hierarchies and each hierarchy was calibrated respectively by the NSGA-II multiobjective optimization algorithm. This new methodology of calibration ensured higher accuracy and efficiency, and enhanced the generalization property of the BPNN. The accuracy of flow concentration module of Xinanjiang model is not always high enough, in order to combine advantages of conceptual and black-box models, XBK and XSBK models were proposed. The XBK model was constituted by coupling runoff generation module of Xinanjiang model with BK flow concentration model and the XSBK model was constituted by coupling runoff generation and separation module of Xinanjiang model with BK flow concentration model. BK, XBK, XSBK, and Xinanjiang models were applied in Chengcun, Dongwan, and Dage watersheds. The simulation results indicated that improved models obtained higher accuracies than Xinanjiang model and can overcame limitations of traditional BPNN model. (C) 2014 American Society of Civil Engineers.

2014




* 234(<-349): Evaluation of modelling techniques for forest site productivity prediction in contrasting ecoregions using stochastic multicriteria acceptability analysis (SMAA)

Accurate estimation of site productivity is crucial for sustainable forest resource management. In recent years, a variety of modelling approaches have been developed and applied to predict site index from a wide range of environmental variables, with varying success. The selection, application and comparison of suitable modelling techniques remains therefore a meticulous task, subject to ongoing research and debate. In this study, the performance of five modelling techniques was compared for the prediction of forest site index in two contrasting ecoregions: the temperate lowland of Flanders, Belgium, and the Mediterranean mountains in SW Turkey. The modelling techniques include statistical (multiple linear regression - MLR, classification and regression trees - CART, generalized additive models - GAM), as well as machine-learning (artificial neural networks - ANN) and hybrid techniques (boosted regression trees - BRT). Although the selected predictor variables differed largely, with mainly topographic predictor variables in the mountain area versus soil and humus variables in the lowland area, the techniques performed comparatively similar in both ecoregions. Stochastic multicriteria acceptability analysis (SMAA) was found a well-suited multicriteria evaluation method to evaluate the performance of the modelling techniques. It has been applied on the individual species models of Flanders, as well as a species-independent evaluation, combining all developed models from the two contrasting ecoregions. We came to the conclusion that non-parametric models are better suited for predicting site index than traditional MLR. GAM and BRT are the preferred alternatives for a wide range of weight preferences. CART is preferred when very high weight is given to user-friendliness, whereas ANN is recommended when most weight is given to pure predictive performance. (C) 2011 Elsevier Ltd. All rights reserved.

2011



* 235(<-416): Comparison and ranking of different modelling techniques for prediction of site index in Mediterranean mountain forests

Forestry science has a long tradition of studying the relationship
between stand productivity and abiotic and biotic site
characteristics, such as climate, topography, soil and
vegetation. Many of the early site quality modelling studies related
site index to environmental variables using basic statistical methods
such as linear regression. Because most ecological variables show a
typical non-linear course and a non-constant variance distribution, a
large fraction of the variation remained unexplained by these linear
models. More recently, the development of more advanced non-parametric
and machine learning methods provided opportunities to overcome these
limitations. Nevertheless, these methods also have drawbacks. Due to
their increasing complexity they are not only more difficult to
implement and interpret, but also more vulnerable to
overfitting. Especially in a context of regionalisation, this may
prove to be problematic. Although many non-parametric and machine
learning methods are increasingly used in applications related to
forest site quality assessment, their predictive performance has only
been assessed for a limited number of methods and ecosystems. In this
study, five different modelling techniques are compared and evaluated,
i.e. multiple linear regression (MLR), classification and regression
trees (CART), boosted regression trees (BRT), generalized additive
models (GAM), and artificial neural networks (ANN). Each method is
used to model site index of homogeneous stands of three important tree
species of the Taurus Mountains (Turkey): Pinus brutia, Pinus nigra
and Cedrus libani. Site index is related to soil, vegetation and
topographical variables, which are available for 167 sample plots
covering all important environmental gradients in the research
area. The five techniques are compared in a multi-criteria decision
analysis in which different model performance measures, ecological
interpretability and user-friendliness are considered as
criteria. When combining these criteria, in most cases GAM is found to
outperform all other techniques for modelling site index for the three
species. BRT is a good alternative in case the ecological
interpretability of the technique is of higher importance. When
user-friendliness is more important MLR and CART are the preferred
alternatives. Despite its good predictive performance, ANN is
penalized for its complex, non-transparent models and big training
effort. (C) 2010 Elsevier B.V. All rights reserved.

2010




* 238(<-568): Authentication of vegetable oils on the basis of their physico-chemical properties with the aid of chemometrics

In food production, reliable analytical methods for confirmation of
purity or degree of spoilage are required by growers, food quality
assessors, processors, and consumers. Seven parameters of
physico-chemical properties, such as acid number, colority, density,
refractive index, moisture and volatility, saponification value and
peroxide value, were measured for quality and adulterated soybean, as
well as quality and rancid rapeseed oils. Chemometrics methods were
then applied for qualitative and quantitative discrimination and
prediction of the oils by methods such exploratory principal component
analysis (PCA), partial least squares (PLS), radial basis
function-artificial neural networks (RBF-ANN), and multi-criteria
decision making methods (MCDM), PROMETHEE and GAIA. In general, the
soybean and rapeseed oils were discriminated by PCA, and the two
spoilt oils behaved differently with the rancid rapeseed samples
exhibiting more object scatter on the PC-scores plot, than the
adulterated soybean oil. For the PLS and RBF-ANN prediction methods,
suitable training models were devised, which were able to predict
satisfactorily the category of the four different oil samples in the
verification set. Rank ordering with the use of MCDM models indicated
that the oil types can be discriminated on the PROMETHEE 11 scale. For
the first time, it was demonstrated how ranking of oil objects with
the use of PROMETHEE and GAIA could be utilized as a versatile
indicator of quality performance of products on the basis of a
standard selected by the stakeholder. In principle, this approach
provides a very flexible method for assessment of product quality
directly from the measured data. (c) 2006 Elsevier B.V. All rights
reserved.

2006




* 250(<- 14): An Interval-Valued Neural Network Approach for Uncertainty Quantification in Short-Term Wind Speed Prediction

We consider the task of performing prediction with neural networks
(NNs) on the basis of uncertain input data expressed in the form of
intervals. We aim at quantifying the uncertainty in the prediction
arising from both the input data and the prediction model. A
multilayer perceptron NN is trained to map interval-valued input data
onto interval outputs, representing the prediction intervals (PIs) of
the real target values. The NN training is performed by nondominated
sorting genetic algorithm-II, so that the PIs are optimized both in
terms of accuracy (coverage probability) and dimension
(width). Demonstration of the proposed method is given in two case
studies: 1) a synthetic case study, in which the data have been
generated with a 5-min time frequency from an autoregressive moving
average model with either Gaussian or Chi-squared innovation
distribution and 2) a real case study, in which experimental data
consist of wind speed measurements with a time step of 1
h. Comparisons are given with a crisp (single-valued) approach. The
results show that the crisp approach is less reliable than the
interval-valued input approach in terms of capturing the variability
in input.

2015





* 253(<-502): Multiobjective training of artificial neural networks for rainfall-runoff modeling

This paper presents results on the application of various optimization
algorithms for the training of artificial neural network
rainfall-runoff models. Multilayered feed-forward networks for
forecasting discharge from two mesoscale catchments in different
climatic regions have been developed for this purpose. The
performances of the multiobjective algorithms Multi Objective Shuffled
Complex Evolution Metropolis-University of Arizona (MOSCEM-UA) and
Nondominated Sorting Genetic Algorithm II (NSGA-II) have been compared
to the single-objective Levenberg-Marquardt and Genetic Algorithm for
training of these models. Performance has been evaluated by means of a
number of commonly applied objective functions and also by
investigating the internal weights of the networks. Additionally, the
effectiveness of a new objective function called mean squared
derivative error, which penalizes models for timing errors and noisy
signals, has been explored. The results show that the multiobjective
algorithms give competitive results compared to the single-objective
ones. Performance measures and posterior weight distributions of the
various algorithms suggest that multiobjective algorithms are more
consistent in finding good optima than are single-objective
algorithms. However, results also show that it is difficult to
conclude if any of the algorithms is superior in terms of accuracy,
consistency, and reliability. Besides the training algorithm, network
performance is also shown to be sensitive to the choice of objective
function(s), and including more than one objective function proves to
be helpful in constraining the neural network training.

2008




* 258(<-214): Modeling and optimization of biodiesel engine performance using advanced machine learning methods

This study aims to determine optimal biodiesel ratio that can achieve
the goals of fewer emissions, reasonable fuel economy and wide engine
operating range. Different advanced machine learning techniques,
namely ELM (extreme learning machine), LS-SVM (least-squares support
vector machine) and RBFNN (radial-basis function neural network), are
used to create engine models based on experimental data. Logarithmic
transformation of dependent variables is used to alleviate the
problems of data scarcity and data exponentiality
simultaneously. Based on the engine models, two optimization methods,
namely SA (simulated annealing) and PSO (particle swarm optimization),
are employed and a flexible objective function is designed to
determine the optimal biodiesel ratio subject to various user-defined
constraints. A case study is presented to verify the modeling and
optimization framework. Moreover, two comparisons are conducted, where
one is among the modeling techniques and the other is among the
optimization techniques. Experimental results show that, in terms of
the model accuracy and training time, ELM with the logarithmic
transformation is better than LS-SVM and RBFNN with/without the
logarithmic transformation. The results also show that PSO outperforms
SA in terms of fitness and standard deviation, with an acceptable
computational time. (C) 2013 Elsevier Ltd. All rights reserved.

2013




* 297(<-136): Improving the Accuracy of Urban Land Cover Classification Using Radarsat-2 PolSAR Data

Land cover classification is one of the most important applications of polarimetric SAR images, especially in urban areas. There are numerous features that can be extracted from these images, hence feature selection plays an important role in PolSAR image classification. In this study, three main steps are used to address this task: 1) feature extraction in the form of three categories, namely original data features, decomposition features, and SAR discriminators; 2) feature selection in the framework of the single and multi-objective optimization; and 3) image classification using the best subset of features. In single objective methods, we employ genetic algorithms (GAs) and support vector machines (SVMs) or multi-layer perceptron (MLP) neural network in order to maximize classification accuracy. Then a new method is proposed to perform an efficient land cover classification of the San Francisco Bay urban area based on the multi-objective optimization approach. The objectives are to minimize the error of classification and the number of selected PolSAR parameters. The experimental results on Radarsat-2 fine-quad data show that the proposed method outperforms the single objective approaches tested against it, while saving computational complexity. Finally, we show that the our method has a better performance than the SVM with full set of features and the Wishart classifier which is based on the covariance matrix.

2014




* 317(<-250): Evaluation of environmental impacts using backpropagation neural network

Purpose - The purpose of this present study is to find a scientific
method for the evaluation of environmental impacts according to the
requirement 4.3.1. Design/methodology/approach - To realize the
objectives, the authors worked with a representative sample from
certified ISO 14001 organizations. The data aim to identify and
evaluate (according to the organization's methodology) significant
environmental impacts. In this study, the authors created two models
for the evaluation of environmental impacts based on an artificial
neural network applied in the pilot organization and compared the
results obtained from these models with those obtained by applying an
analytic hierarchy process (AHP) method. AHP is part of an
multi-criteria decision making method and provides good multi-criteria
support for decision making for problems that can be structured
hierarchically. Findings - This paper presents a new approach that
uses a backpropagation neural network to evaluate environmental
impacts regardless of the organization type. Originality/value - This
paper presents a unique approach for the reliable and objective
evaluation of environmental impacts.

2013




* 335(<-154): Optimal Management of a Freshwater Lens in a Small Island Using Surrogate Models and Evolutionary Algorithms

This paper examines a linked simulation-optimization procedure based
on the combined application of an artificial neural network (ANN) and
genetic algorithm (GA) with the aim of developing an efficient model
for the multiobjective management of groundwater lenses in small
islands. The simulation-optimization methodology is applied to a real
aquifer in Kish Island of the Persian Gulf to determine the optimal
groundwater-extraction while protecting the freshwater lens from
seawater intrusion. The initial simulations are based on the
application of SUTRA, a variable-density groundwater numerical
model. The numerical model parameters are calibrated through automated
parameter estimation. To make the optimization process computationally
feasible, the numerical model is subsequently replaced by a trained
ANN model as an approximate simulator. Even with a moderate number of
input data sets based on the numerical simulations, the ANN metamodel
can be efficiently trained. The ANN model is subsequently linked with
GA to identify the nondominated or Pareto-optimal solutions. To
provide flexibility in the implementation of the management plan, the
model is built upon optimizing extraction from a number of zones
instead of point-well locations. Two issues are of particular interest
to the research reported in this paper are: (1) how the general idea
of minimizing seawater intrusion can be effectively represented by
objective functions within the framework of the
simulation-optimization paradigm, and (2) the implications of applying
the methodology to a real-world small-island groundwater lens. Four
different models have been compared within the framework of
multiobjective optimization, including (1) minimization of maximum
salinity at observation wells, (2) minimization of the root mean
square (RMS) change in concentrations over the planning period, (3)
minimization of the arithmetic mean, and (4) minimization of the
trimmed arithmetic mean of concentration in the observation wells. The
latter model can provide a more effective framework to incorporate the
general objective of minimizing seawater intrusion. This paper shows
that integration of the latest innovative tools can provide the
ability to solve complex real-world optimization problems in an
effective way. (C) 2014 American Society of Civil Engineers.

2014




* 339(<-619): Joint application of artificial neural networks and evolutionary algorithms to watershed management

Artificial neural networks (ANNs) have become common data driven tools for modeling complex, nonlinear problems in science and engineering. Many previous applications have relied on gradient-based search techniques, such as the back propagation ( BP) algorithm, for ANN training. Such techniques, however, are highly susceptible to premature convergence to local optima and require a trial-and-error process for effective design of ANN architecture and connection weights. This paper investigates the use of evolutionary programming ( EP), a robust search technique, and a hybrid EP-BP training algorithm for improved ANN design. Application results indicate that the EP-BP algorithm may limit the drawbacks of using local search algorithms alone and that the hybrid performs better than EP from the perspective of both training accuracy and efficiency. In addition, the resulting ANN is used to replace the hydrologic simulation component of a previously developed multiobjective decision support model for watershed management. Due to the efficiency of the trained ANN with respect to the traditional simulation model, the replacement reduced the overall computational time required to generate preferred watershed management policies by 75%. The reduction is likely to improve the practical utility of the management model from a typical user perspective. Moreover, the results reveal the potential role of properly trained ANNs in addressing computational demands of various problems without sacrificing the accuracy of solutions.

2004




* 349(<- 54): Coupled Data-Driven Evolutionary Algorithm for Toxic Cyanobacteria (Blue-Green Algae) Forecasting in Lake Kinneret

Cyanobacteria blooming in surface waters have become a major concern
worldwide, as they are unsightly, and cause a variety of toxins,
undesirable tastes, and odors. Approaches of mathematical
process-based (deterministic), statistically based, rule-based
(heuristic), and artificial neural networks have been the subject of
extensive research for cyanobacteria forecasting. This study suggests
a new framework of linking an evolutionary computational method (a
genetic algorithm) with a data driven modeling engine (model trees)
for external loading, physical, chemical, and biological parameters
selection, all coupled with their associated time lags as decision
variables for cyanobacteria prediction in surface waters. The
methodology is demonstrated through trial runs and sensitivity
analyses on Lake Kinneret (the Sea of Galilee), Israel. Model trials
produced good matching as depicted through the results correlation
coefficient on verification data sets. Temperature was reconfirmed as
a predominant parameter for cyanobacteria prediction. Model optimal
input variables and forecast horizons differed in various
solutions. Those in turn raised the problem of best variables
selection, pointing towards the need of a multiobjective optimization
model in future extensions of the proposed methodology. (C) 2014
American Society of Civil Engineers.

2015




* 354(<-512): Prediction of the solar radiation evolution using computational intelligence techniques and cloudiness indices

In this paper, Artificial Neural Networks are applied for multi-step
long term solar radiation prediction. The input-output structure of
the neural network models is selected using evolutionary computation
methods. The networks are trained as one-step-ahead predictors and
iterated over time to obtain multi-step longer term
predictions. Auto-regressive and auto-regressive with exogenous inputs
models are compared, considering cloudiness indices as inputs in the
latter case. These indices are obtained through pixel classification
of ground-to-sky images, captured by a CCD camera.

2008




* 357(<-231): Manganese waste mud immobilization in cement natural zeolite lime blend: Process optimization using artificial neural networks and multi-criteria functions

In this study, stabilization/solidification process of manganese
contaminated mud using portland cement was optimized. For that
purpose, immobilization process was modeled applying artificial neural
networks with radial basis activation function. The optimal model
presented satisfactory prediction characteristics (R2 value for
manganese leaching was 0.9615 while and for concrete flexural strength
0.8748). Therefore, it was used in combination with seven in-house
developed multi-criteria optimization functions, separately, in order
to optimize concrete formulation. The used approach proved itself as
efficient and cost effective alternative in ecological material
formulation process. The best properties (i.e. high flexural strength
and lowest manganese leaching) manifested stabilization/solidification
matrix consisted of 350 g of portland cement, 20 g of lime, 70 g of
natural zeolite, 10 g of manganese waste mud and 180 g of water.

2013




* 358(<-318): OPTIMIZATION OF ARSENIC SLUDGE IMMOBILIZATION PROCESS IN CEMENT - NATURAL ZEOLITE - LIME BLENDS USING ARTIFICIAL NEURAL NETWORKS AND MULTI-OBJECTIVE CRITERIA FUNCTIONS

This work focuses on optimization of arsenic sludge immobilization
process in cement natural zeolite lime blends using artificial neural
networks and multi-objective criteria functions. The developed
artificial neural network model describes relations between
solidified/stabilized cement formulation and its mechanical
(compressive strength) and ecological properties (arsenic and iron
release). It is proven that developed artificial neural network
solidified/stabilized model has satisfactory performance
characteristics (R-2>0.9031 without presence of systematic error;
based on an external validation experimental data set). Four
multi-objective optimization criteria functions, different in terms of
mathematical formulation and ecological interpretation, were
developed. The developed criteria functions were used in combination
with the artificial neural network solidified/stabilized model,
providing optimal cement formulation. Finally, this study describes an
efficient and cost-effective alternative in ecological material
formulation process.

2012




* 370(<-603): Application of genetic algorithms for process integration and optimization considering environmental factors

A systematic methodology for pollution prevention based on process
integration is presented in this report. In this methodology, process
simulation was carried out to provide mass and energy information of
the chemical process. An artificial neural network (ANN) was used to
replace rigorous process simulation in the optimization process to
improve computational efficiency. Multiobjective optimization was
performed to obtain coordinate optimization of process performance for
both economics and environmental aspects. A multiobjective genetic
algorithm was used to solve the multiobjective optimization
problems. Mass and energy use were considered simultaneously in this
program. A case study of a wastewater recovery system in ammonia
production process is discussed to illustrate the effectiveness of
this pollution prevention methodology. (c) 2004 American Institute of
Chemical Engineers Environ Prog,

2005




* 371(<-611): Decision support for watershed management using evolutionary algorithms

An integrative computational methodology is developed for the
management of nonpoint source pollution from watersheds. The
associated decision support system is based on an interface between
evolutionary algorithms (EAs) and a comprehensive watershed simulation
model, and is capable of identifying optimal or near-optimal land use
patterns to satisfy objectives. Specifically, a genetic algorithm (GA)
is linked with the U.S. Department of Agriculture's Soil and Water
Assessment Tool (SWAT) for single objective evaluations, and a
Strength Pareto Evolutionary Algorithm has been integrated with SWAT
for multiobjective optimization. The model can be operated at a small
spatial scale, such as a farm field, or on a larger watershed scale. A
secondary model that also uses a GA is developed for calibration of
the simulation model. Sensitivity analysis and parameterization are
carried out in a preliminary step to identify model parameters that
need to be calibrated. Application to a demonstration watershed
located in Southern Illinois reveals the capability of the model in
achieving its intended goals. However, the model is found to be
computationally demanding as a direct consequence of repeated SWAT
simulations during the search for favorable solutions. An artificial
neural network (ANN) has been developed to mimic SWAT outputs and
ultimately replace it during the search process. Replacement of SWAT
by the ANN results in an 84% reduction in computational time required
to identify final land use patterns. The ANN model is trained using a
hybrid of evolutionary programming (EP) and the back propagation (BP)
algorithms. The hybrid algorithm was found to be more effective and
efficient than either EP or BP alone. Overall, this study demonstrates
the powerful and multifaceted role that EAs and artificial
intelligence techniques could play in solving the complex and
realistic problems of environmental and water resources systems.

2005



* 408(<-277): Defining a nonlinear control problem to reduce particulate matter population exposure

In this paper a multi-objective nonlinear approach to control air
quality at a regional scale is presented. Both economic and air
quality sides of the problem are modeled through artificial neural
network models. Simulating the complex nonlinear atmospheric
phenomena, they can be used in an optimization routine to identify the
efficient solutions of a decision problem for air quality
planning. The methodology is applied over Northern Italy, an area in
Europe known for its high concentrations of particulate
matter. Results illustrate the effectiveness of the approach assessing
the nonlinear chemical reactions in an air quality decision
problem. (C) 2012 Elsevier Ltd. All rights reserved.

2012




* 409(<-288): Surrogate models to compute optimal air quality planning policies at a regional scale

Secondary pollutants (such as PM10) derives from complex non-linear reactions involving precursor emissions, namely VOC, NOx, NH3, primary PM and SO2. Due to difficulty to cope with this complexity, Decision Support Systems (DSSs) are essential tools to help Environmental Authorities to plan air quality policies that fulfill EU Directive 2008/50 requirements in a cost-efficient way. To implement these DSSs the common approach is to describe the air quality indices using linear models, derived through model reduction techniques starting from deterministic Chemical Transport Model simulations. This linear approach limits the applicability of these surrogate models, and while these may work properly at coarse spatial resolutions (continental/national), where average values over large areas are of interest, they often prove inadequate at sub national scales, where the impact of non linearities on air quality are usually higher. The objective of this work is to identify air quality models able to properly describe the relation between emissions and air quality indices, at a sub national scale. In this context, artificial neural networks, identified processing long-term simulation output of a 3D deterministic multi-phase modelling system, are used to describe the non-linear relations between the control variables (precursor emissions reduction) and a pollution index. These models can then be used with a reasonable computing effort to solve a multi-objective (air quality and emission reduction costs) optimization problem, that requires thousands of model runs and thus would be unfeasible using the original process-based model. A case study of Northern Italy is presented. (C) 2011 Elsevier Ltd. All rights reserved.

2012


* 410(<-508): A multi-objective nonlinear optimization approach to designing effective air quality control policies

This paper presents the implementation of a two-objective optimization methodology to select effective tropospheric ozone pollution control strategies on a mesoscale domain. The objectives considered are (a) the emission reduction cost and (b) the Air Quality Index. The control variables are the precursor emission reductions due to available technologies. The nonlinear relationship linking air quality objective and precursor emissions is described by artificial neural networks, identified by processing deterministic Chemical Transport Modeling system simulations. Pareto optimal solutions are calculated with the Weighted Sum Strategy. The two-objective problem has been applied to a complex domain in Northern Italy, including the Milan metropolitan area, a region characterized by frequent and persistent ozone episodes. (c) 2008 Elsevier Ltd. All rights reserved.

2008


* 411(<- 11): AI techniques for optimizing multi-objective reservoir operation upon human and riverine ecosystem demands

Flow regime is the key driver of the riverine ecology. This study proposes a novel hybrid methodology based on artificial intelligence (Al) techniques for quantifying riverine ecosystems requirements and delivering suitable flow regimes that sustain river and floodplain ecology through optimizing reservoir operation. This approach addresses issues to better fit riverine ecosystem requirements with existing human demands. We first explored and characterized the relationship between flow regimes and fish communities through a hybrid artificial neural network (ANN). Then the non-dominated sorting genetic algorithm II (NSGA-II) was established for river flow management over the Shihmen Reservoir in northern Taiwan. The ecosystem requirement took the form of maximizing fish diversity, which could be estimated by the hybrid ANN. The human requirement was to provide a higher satisfaction degree of water supply. The results demonstrated that the proposed methodology could offer a number of diversified alternative strategies for reservoir operation and improve reservoir operational strategies producing downstream flows that could meet both human and ecosystem needs. Applications that make this methodology attractive to water resources managers benefit from the wide spread of Pareto-front (optimal) solutions allowing decision makers to easily determine the best compromise through the trade-off between reservoir operational strategies for human and ecosystem needs. (C) 2015 Elsevier B.V. All rights reserved.

2015



* 412(<-140): Simulation and Optimization Modeling for the Management of Groundwater Resources. II: Combined Applications

The world population is increasing continuously and is expected to reach the 9.5billion mark in 2050 from the current 7.1billion. The importance of the groundwater resources is also increasing with the increase in population, because the quality and quantity of water resources are continuously declining because of urbanization, contamination, and climate change impact. Thus, under the current environment, the conservation and management of groundwater resources is a critical challenge for fulfilling the rising water demand for agricultural, industrial, and domestic uses. Various simulation and optimization approaches have been used to solve the groundwater management problems. Optimization methods have proved to be of great importance when used with simulation models, and the combined use of these two approaches gives the best results. The main objective of this review is to analyze combined uses of simulation and optimization modeling approaches and to provide an impression of their applications reported in literature. In addition to traditional optimization techniques, this paper highlights the application of computational intelligence techniques, such as artificial neural networks, response matrix approach, and multiobjective approach. Conclusions are drawn based on this review, which could be useful for system managers and planners for selecting the best suitable technique for their specific uses.

2014


* 413(<-160): Optimization modelling for seawater intrusion management

The coastal aquifers of the world are facing environmental problem of seawater intrusion. This problem is the results of indiscriminate and unplanned groundwater exploitation for fulfilling the growing need of freshwater for the burgeoning global population. There is a need to develop appropriate management models for assessing the maximum feasible pumping rates which protects seawater intrusion in coastal aquifers. The comprehensive reviews on the use of various programming techniques for the solution of seawater intrusion management problem of coastal aquifers have been provided in this paper. The literature review revealed that the management models used in the past mainly considered the objectives of maximization of pumping rate, minimization of drawdown, minimization of pumped water, minimization of seawater volume into the aquifer, and/or minimization of pumping cost. The past reviews are grouped into five sections based on the programming techniques adopted. The sections include: linear programming, nonlinear programming, genetic algorithms, artificial neural networks, and multiobjective optimization models. Conclusions are drawn where gaps exist and more research needs to be focused. This review provides the basis for the selection of appropriate methodology for the management of seawater intrusion problems of coastal aquifers. (C) 2013 Elsevier B.V. All rights reserved.

2014


* 414(<-171): Multi-Objective Quantity-Quality Reservoir Operation in Sudden Pollution

Damage caused by entered pollution in reservoirs can affect a water resource system in two ways: (1) Damages that are caused due to consumption of polluted water and (2) damages that are caused due to insufficient water allocation. Those damages conflict with each other. Thus, the crisis should be managed in a way that the least damage occurs in the water resource system. This paper investigates crisis management due to the sudden entrance of a 30 m(3) methyl tert-butyl ether (MTBE) load to the Karaj dam in Iran, which supplies municipal water to the cities of Tehran and Karaj. To simulate MTBE advection, dispersion, and vaporization, the latter process is added to the CE-QUAL-W2 model. After that, the multi-objective NSGAII-ALANN algorithm, which is a combination of the NSGAII optimization method along with a multi layer perceptron (MLP), which is one of the most widely used artificial neural network (ANN) structures, is employed to extract the best set of decisions in which the two aforementioned damages are minimized. By assigning a specific importance to each objective function, after extracting the optimal solutions, it is possible to choose one of the solutions with the least damage. Four scenarios of entering pollution to the Karaj reservoir the first day of each season are considered, resulting in a Pareto set of operation policies for each scenario. Results of the proposed methodology indicate that if the pollution enters the reservoir in summer, by using one of the optimal policies extracted from the Pareto set of the 2nd Scenario, by a 36% reduction in meeting the demand, allocated pollution decreases to about 60%. In other seasons, there is a significant decrease in allocated pollution with a smaller reduction in the met demand.

2014


* 415(<-347): Characterizing the Socio-Economic Driving Forces of Groundwater Abstraction with Artificial Neural Networks and Multivariate Techniques

Integrated groundwater quantity management in the Middle East region must consider appropriate control measures of the socio economic needs. Hence, there is a need for a better knowledge and understanding of the socio economic variables influencing the groundwater quantity. Gaza Strip was chosen as the study area and real data were collected from twenty five municipalities for the reference year 2001. In this paper, the effective variables have been characterized and prioritized using multi-criteria analysis with artificial neural networks (ANN) and expert opinion and judgment. The selected variables were classified and organized using the multivariate techniques of cluster analysis, factor analysis, principal components and classification analysis. There are significant discrepancies between the results of ANN analysis and expert opinion and judgment in terms of ranking and prioritizing the socio-economic variables. Characterization of the priority effective socio-economic driving forces indicates that water managers and planners can introduce demand-based groundwater management in place of the existing supply-based groundwater management. This ensures the success of undertaking responsive technical, managerial and regulatory measures. Income per capita has the highest priority. Efficiency of revenue collection is not a significant socio-economic factor. The models strengthen the integration of preventive approach into groundwater quantity management. In addition to that, they assist decision makers to better assess the socio economic needs and undertake proactive measures to protect the coastal aquifer.

2011


* 416(<- 25): Multi-Objective Operations of Multi-Wetland Ecosystem: iModel Applied to the Everglades Restoration

The Everglades is a complex, multiwetland ecosystem that is heavily managed to meet often-competing flood control, water supply, and environmental demands. Using objective measures to balance these demands through operational protocols has always been a challenge in the multibillion-dollar restoration plans for the ecosystem. Physically based models have been the primary tools for planning efforts but for such a complex system, they are laborious and computationally intensive. Development of optimal operations based on iterative runs of these models is a great challenge. This paper presents an inverse modeling framework for formal optimization suited for wetland system operations that helps overcome such limitations. Labor-intensive and computation-intensive physically representative models are emulated in each individual wetland area using an autoregressive artificial neural network with exogenous variables. Using prescribed inflow, outflow, and meteorological input data, such hydrologic model emulators aided by a dimension-reduction technique provide targeted spatial and temporal predictions for water level (stage) within each area of the Everglades, while excluding computation processes that are intensive but insignificant to the predictions. This computer software uses the augmented Lagrangian genetic algorithm technique (subject to linear and nonlinear constraints) to steer predictions of stage spatial variability within individual wetlands towards corresponding desired goals (including restoration targets). In the augmented Lagrangian genetic algorithm, flow releases are coded as the decision variables to be optimized subject to budget, intrahydraulic conveyance, flow capacity, and upstream storage constraints. Optimization is performed by dividing and solving a sequence of subproblems using the genetic algorithm procedures of initialization, selection, elitism, crossover, and mutation. As part of the process, Lagrangian and penalty parameters are updated, and optimization terminates when certain stopping criteria are met. Applying the technique reported in this paper to a specific Everglades restoration plan (the River of Grass Project) showed a sound hydrologic model emulator prediction when compared to the physical model for all wetland areas. Feeding optimal releases predicted by the computer software into a physical model showed equal or better matching of the restoration target with different release patterns compared to that of the physical model base run scenario. Results show that hydraulic conveyance limitations play a significant role in Everglades restoration. Also, results show that employing an adversity tradeoff matrix presented multiple so-called optimal solutions with different optimization weights and a powerful negotiation matrix. (C) 2015 American Society of Civil Engineers.

2015


* 417(<- 42): Multi-Objective Optimal Operation Model of Cascade Reservoirs and Its Application on Water and Sediment Regulation

Recently suspended river, where formed both in tributary and in main stream of Ningxia-Inner Mongolia reaches in the Upper Yellow River, severely threatens people's lives and property safety downstream. In this paper, taking Ningxia-Inner Mongolia reaches for the example, water and sediment are regulated by cascade reservoirs upstream, which can form artificial controlled flood, to improve the relationship of water-sediment and slow down the speed of sedimentation rate. Then, multi-objective optimal operation model of cascade reservoirs is established with four objectives: ice and flood control, power generation, water supply, water and sediment regulation. Based on optimization technique of feasible search space, the non-dominated sorting genetic algorithm (NSGA-II) is improved and a new multi-objective algorithm, Feasible Search Space Optimization-Non-dominated Sorting Genetic Algorithm (FSSO-NSGA-II) is innovatively proposed in this paper. The best time of water and sediment regulation is discussed and the regulation index system and scenarios are constructed in three level years of 2010, 2020 and 2030. After that regulation efforts and contribution to sediment transportation are quantified under three scenarios. Compared with history data in 2010, the accuracy and superiority of multi-objective model and FSSO-NSGA-II are verified. Even more, four-dimensional vector coordinate systems are proposed innovatively to represent each objective and sensitivity of three scenarios are analyzed to clarify the impact on regulation objectives by regulation indexes. At last, relationships between four objectives are revealed. The research findings provide optimal solutions of multi-objectives optimal operation by FSSO-NSGA-II, which have an important theoretical significance to enrich the methods of water and sediment optimal operation by cascade reservoirs, guiding significance to water and sediment regulation implementation and construct water and sediment control system in the whole Yellow River basin. Research Highlights We establish an multi-objective optimal operation model with four regulation objectives. We proposed a improved multi-objective algorithm (FSSO-NSGA-II) based on feasible search space optimization. Regulation index system and three scenarios are constructed. Four-dimensional vector coordinate systems are proposed to represent each objective and sensitivity of scenarios are analyzed. Relationships between four objectives are revealed.

2015





* 418(<- 99): An adaptive ant colony optimization framework for scheduling environmental flow management alternatives under varied environmental water availability conditions

Human water use is increasing and, as such, water for the environment
is limited and needs to be managed efficiently. One method for
achieving this is the scheduling of environmental flow management
alternatives (EFMAs) (e.g., releases, wetland regulators), with these
schedules generally developed over a number of years. However, the
availability of environmental water changes annually as a result of
natural variability (e.g., drought, wet years). To incorporate this
variation and schedule EFMAs in a operational setting, a previously
formulated multiobjective optimization approach for EFMA schedule
development used for long-term planning has been modified and
incorporated into an adaptive framework. As part of this approach,
optimal schedules are updated at regular intervals during the planning
horizon based on environmental water allocation forecasts, which are
obtained using artificial neural networks. In addition, the changes
between current and updated schedules can be minimized to reduce any
disruptions to long-term planning. The utility of the approach is
assessed by applying it to an 89km section of the River Murray in
South Australia. Results indicate that the approach is beneficial
under a range of hydrological conditions and an improved ecological
response is obtained in a operational setting compared with previous
long-term approaches. Also, it successfully produces trade-offs
between the number of disruptions to schedules and the ecological
response, with results suggesting that ecological response increases
with minimal alterations required to existing schedules. Overall, the
results indicate that the information obtained using the proposed
approach potentially aides managers in the efficient management of
environmental water. Key Points <list id="wrcr21132-list-0001"
list-type="bulleted"> <list-item id="wrcr21132-li-0001">Adaptive
optimization framework developed for use in an operational setting
<list-item id="wrcr21132-li-0002">The framework showing advantages
compared with long-term planning approaches <list-item
id="wrcr21132-li-0003">Ecological response increased with minimal
disruption to existing schedules <doi origin="wiley"
registered="yes">10.1002/(ISSN)1944-7973</doi

2014




* 426(<-486): Multi-criteria validation of artificial neural network rainfall-runoff modeling

In this study we propose a comprehensive multicriteria validation test
for rainfall-runoff modeling by artificial neural networks. This study
applies 17 global statistics and 3 additional non-parametric tests to
evaluate the ANNs. The weakness of global statistics for validation of
ANN is demonstrated by rainfall-runoff modeling of the Plasjan Basin
in the western region of the Zayandehrud watershed, Iran. Although the
global statistics showed that the multi layer perceptron with 4 hidden
layers (MLP4) is the best ANN for the basin comparing with other MLP
networks and empirical regression model, the non-parametric tests
illustrate that neither the ANNs nor the regression model are able to
reproduce the probability distribution of observed runoff in
validation phase. However, the MLP4 network is the best network to
reproduce the mean and variance of the observed runoff based on
non-parametric tests. The performance of ANNs and empirical model was
also demonstrated for low, medium and high flows. Although the MLP4
network gives the best performance among ANNs for low, medium and high
flows based on different statistics, the empirical model shows better
results. However, none of the models is able to simulate the frequency
distribution of low, medium and high flows according to non-parametric
tests. This study illustrates that the modelers should select
appropriate and relevant evaluation measures from the set of existing
metrics based on the particular requirements of each individual
applications.

2009



* 429(<-350): Simulation and optimization of sludge hygienization research irradiator

This paper presents an approach to optimize a computational fluid
dynamics problem coupled with radiation physics for sludge
hygienization research irradiator. This was implemented by mapping
results of computational fluid dynamics using recurrent artificial
neural network with simplified dose rate evaluation using modified
Sievert integral. This methodology led to considerable reduction of
computation time while maintaining enough accuracy to validate the
experimental results as well as suggest an optimum placement of source
pencils using reference point based multi-objective optimization. (C)
2011 Elsevier Ltd. All rights reserved.

2011




* 435(<- 68): Hybrid ACO-ANN-Based Multi-objective Simulation-Optimization Model for Pollutant Load Control at Basin Scale

The systematic and integrative approach to optimum watershed
management couples a watershed simulation model and an efficient
optimization algorithm for evaluating great number of "what if"
scenarios in the search domain. This study integrates a
multi-objective Non-dominated Archiving Ant Colony Optimization
(NA-ACO) algorithm as an optimization tool with Soil and Water
Assessment Tool (SWAT) as the simulation module for optimum management
of total suspended solids (TSS) loading to downstream water
bodies. The resulting NA-ACO-SWAT model is computationally and
experimentally expensive because of the large number of required
function evaluations which demands repetitive execution of SWAT
simulation model. To increase the computational efficiency of the
watershed simulation model, the SWAT model is replaced by a trained
artificial neural network (ANN) model to form a hybrid NA-ACO-SWAT-ANN
model to efficiently develop the set of optimum non-dominated
solutions for configuration and design of detention ponds in basin
scale. The applicability of the proposed method is evaluated at
Gharesou watershed in the northwest of Iran. The outcomes of the
proposed approach is further analyzed and compared in terms of their
quality of solutions and computational efficiencies. Results show that
the proposed hybrid approach may reduce the computational time by 90 %
while keeping the accuracy of the results in the same order.

2015



* 436(<- 93): Developing Strategies for Urban Flood Management of Tehran City Using SMCDM and ANN

Water management in urban areas includes controlling storm water and developing efficient drainage systems. High-intensity rainfall events, reduced permeability attributable to urban development, and the aging of drainage systems are the primary reasons for the occurrence of destructive floods in urban areas. Developing a map of areas with the potential for flood hazard may be an appropriate tool for urban planning and development strategies. The vulnerability analysis of different urban areas is a complex process because it depends on various spatial and temporal parameters and criteria. The purpose of this research is to prepare a tool to make precise decisions in urban flood management by using multicriteria decision making and a geographic information system. The development of an artificial neural network (ANN) model as an alternative to the weighting process of decision makers is presented as a solution to mitigate the disagreement among decision makers on the weighting process of decision-making analysis. The developed spatial multicriteria decision making (SMCDM) tool allows the processing of necessary data and criteria and combining them through the decision-making process. All of the necessary data analysis and processing are automatically run within a developed toolbox. The advantages of using the developed toolbox in generating flood management strategies are discussed in a case study of Tehran, Iran.

2014


* 437(<- 94): Adaptive surrogate modeling for optimization of flood control detention dams

Real world optimization problems commonly require complex computationally demanding models which use in practice could be limited. To address this drawback, different meta-modeling approaches have been proposed in the literature. This study represents an adaptive meta-modeling framework to solve a high dimensional design problem by coupling a hydrodynamic model, a multi objective genetic algorithm and artificial neural network as the surrogate model. The proposed dynamic learning approach includes special space filling techniques and control strategies to periodically update surrogate model and keep the fidelity level of solutions as the optimization progresses. An experimental application on flood management based on a case study in Iran was carried out in order to test the efficiency of the proposed approach. Results showed the approach provides a 74% time saving while maintaining the quality of solutions when compared with those of the original simulation model indicating the viability of the approach. (C) 2014 Elsevier Ltd. All rights reserved.

2014




* 441(<-  5): Optimal Design of Detention Rockfill Dams Using a Simulation-Based Optimization Approach with Mixed Sediment in the Flow

Detention rockfill dams are generally known as economic and efficient
structural methods for flood mitigation and watershed
management. Despite the existing studies in the context of rockfill
dam design, designers still rely on time-consuming and costly
experimental tests which can only lead into practical
guidelines. Moreover, investigation of various rockfill dam design
alternatives in terms of hydraulic performance needs an exhaustive
framework to obtain the optimal design values. With respect to these,
this research study aims to investigate the optimal dam thickness and
mean diameter of coarse aggregates in detention rockfill dams in order
to improve the hydraulic characteristics, e.g., reduction in flood
peak discharge and increase in flood duration while the construction
cost is minimized. Here, a methodology based on data-driven simulation
model, multi-objective optimization model, multi-criterion decision
making model, and social choice method is suggested to find the
optimal design parameters of detention rockfill dam. This methodology
adapts data to model and model to optimum design
variables. Experimental results are converted to data-driven
simulation model based on multilayer perceptron neural network, which
can model the relationship between design variables of detention
rockfill dam and parameters of flood hydrograph. The simulation model
is used in a robust multi-objective optimization algorithm,
non-dominated sorting genetic algorithm-I (TM) I (TM) (NSGA-I (TM) I
(TM)), to establish a trade-off between the conflicting
objectives. Eventually, PROMETHEE decision making model and Borda
count social choice method are used to find the best agreed-upon
design optimal point on the trade-off curve. Results indicate that the
best non-dominated solution can be considered as 10.01 and 5 cm for
the optimum thickness of detention rockfill dam and optimum mean
diameter of coarse aggregates in the porous media, respectively. The
similar optimum characteristics of detention rockfill dam using
PROMETHEE and Borda count social choice method depict the stable
performance of the proposed methodology.

2015



* 450(<-538): Multi-objective performance comparison of an artificial neural network and a conceptual rainfall-runoff model

A multi-objective comparison between an artificial neural network and
the conceptual HBV rainfall-runoff model has been performed. The
popular NSGA-11 algorithm was used for calibration of both models. A
combination of three objective functions was used to evaluate model
performance. The results show that, for a small forecast lead time,
the artificial neural network outperformed the HBV model on the
objective functions for low and high flows, but the former was
outperformed on a novel objective function related to the shape of the
hydrograph. As the forecast horizon increases, the HBV model starts to
outperform the ANN model on all objective functions. The main
conclusion of this study is that, although the differences between the
two model approaches make a straightforward and unequivocal comparison
difficult, the multi-objective approach enables a more reliable
evaluation of the two models than the single-objective approach.

2007



* 451(<-572): Optimal design of artificial neural networks by a multiobjective strategy: groundwater level predictions

Currently, environmental modelling is frequently conducted with the aid of artificial neural networks (ANNs) in an effort to achieve greater accuracy in simulation and forecasting beyond that typically obtained when using solely linear models. For the design of an ANN, modellers must contend with two key issues: (a) the selection of model input and (b) the determination of the number of hidden neurons. A novel approach is introduced to address the optimal design of ANNs based on a multi-objective strategy that enables the user to find a set of feasible ANNs, determined as optimal trade-off solutions between model simplicity and accuracy. This is achieved in a multi-objective fashion by simultaneously minimizing three different cost functions: the model input dimension, the hidden neuron number and the generalization error computed on a validation set of data. The multi-objective approach is based on the Paretc, dominance criterion and an evolutionary strategy has been employed to solve the combinatorial optimization problem. From a theoretical perspective, the choice of a multi-objective approach marks an attempt to account for, and overcome, the "curse of dimensionality" and to circumvent the drawbacks of "overfitting" that are inherent in ANNs. Moreover, it is demonstrated that the strategy renders the choice of the ANN more robust, as is evident by "unseen data" in the testing stage, since structure determination is not merely based on the statistical evaluation of the generalization performance. The methodology is tested and the results are reported in a case study relating groundwater level predictions to total monthly rainfall.

2006


* 452(<-575): Artificial neural networks and multicriterion analysis for sustainable irrigation planning

The objective of the present paper is to select the best compromise irrigation planning strategy for the case study of Jayakwadi irrigation project, Maharashtra, India. Four-phase methodology is employed. In phase 1, separate linear programming (LP) models are formulated for the three objectives, namely. net economic benefits, agricultural production and labour employment. In phase 2, nondominated (compromise) irrigation planning strategies are generated using the constraint method of multiobjective optimisation. In phase 3, Kohonen neural networks (KNN) based classification algorithm is employed to sort nondominated irrigation planning strategies into smaller groups. In phase 4, multicriterion analysis (MCA) technique, namely, Compromise Programming is applied to rank strategies obtained from phase 3. It is concluded that the above integrated methodology is effective for modeling multiobjective irrigation planning problems and the present approach can be extended to situations where number of irrigation planning strategies are even large in number. (c) 2004 Elsevier Ltd. All rights reserved.

2006





* 455(<-669): An integrated approach to economic emission load dispatching using neural network and goal-attainment methods

This paper presents an integrated approach comprising artificial
neural network (ANN) and goal-attainment (GA) methods to economic
emission load dispatching (EELD) in power system operation and
scheduling phases. The GA method, which is one of the most powerful
tools for multiobjective optimization problems, is quantitatively
performed to grasp trade-off relations between the two conflicting
objectives (economy and emission impact) in the training set creation
phase. Finally, a radial basis function ANN is trained by the
orthogonal least squares learning algorithm to reach the optimal
generations. The ANN models so developed have been tested to solve
EELD problem on B-bus and 71-bus power systems. The test results
demonstrate that the proposed approach is capable of obtaining
well-coordinated optimal solutions suitable both in accuracy and speed
while allowing flexibility in the operation of the generators.

1999



* 456(<- 24): Review: Simulation-optimization models for the management and monitoring of coastal aquifers

The literature on the application of simulation-optimization
approaches for management and monitoring of coastal aquifers is
reviewed. Both sharp- and dispersive-interface modeling approaches
have been applied in conjunction with optimization algorithms in the
past to develop management solutions for saltwater
intrusion. Simulation-optimization models based on sharp-interface
approximation are often based on the Ghyben-Herzberg relationship and
provide an efficient framework for preliminary designs of
saltwater-intrusion management schemes. Models based on
dispersive-interface numerical models have wider applicability but are
challenged by the computational burden involved when applied in the
simulation-optimization framework. The use of surrogate models to
substitute the physically based model during optimization has been
found to be successful in many cases. Scalability is still a challenge
for the surrogate modeling approach as the computational advantage
accrued is traded-off with the training time required for the
surrogate models as the problem size increases. Few studies have
attempted to solve stochastic coastal-aquifer management problems
considering model prediction uncertainty. Approaches that have been
reported in the wider groundwater management literature need to be
extended and adapted to address the challenges posed by the stochastic
coastal-aquifer management problem. Similarly, while abundant
literature is available on simulation-optimization methods for the
optimal design of groundwater monitoring networks, applications
targeting coastal aquifer systems are rare. Methods to optimize
compliance monitoring strategies for coastal aquifers need to be
developed considering the importance of monitoring feedback
information in improving the management strategies.

2015


* 457(<-246): Sustainable management of a coupled groundwater-agriculture hydrosystem using multi-criteria simulation based optimisation

In this paper we present a new simulation-based integrated water management tool for sustainable water resources management in arid coastal environments. This tool delivers optimised groundwater withdrawal scenarios considering saltwater intrusion as a result of agricultural and municipal water abstraction. It also yields a substantially improved water use efficiency of irrigated agriculture. To allow for a robust and fast operation we unified process modelling with artificial intelligence tools and evolutionary optimisation techniques. The aquifer behaviour is represented using an artificial neural network (ANN) which emulates a numerical density-dependent groundwater flow model. The impact of agriculture is represented by stochastic crop water production functions (SCWPF). Simulation-based optimisation techniques together with the SCWPF and ANN deliver optimal groundwater abstraction and cropping patterns. To address contradicting objectives, e. g. profit-oriented agriculture vs. sustainable abstraction scenarios, we performed multi-objective optimisations using a multi-criteria optimisation algorithm.

2013


* 458(<-444): Saltwater Intrusion Management of Coastal Aquifers. I: Linked Simulation-Optimization

A set of different methodologies are developed for multiple objective management of coastal aquifers. The coastal aquifer management models are developed using a numerical simulation model, meta-model, and the multiple objective optimization algorithm NSGA-II. The NSGA-II algorithm is also modified to accommodate initial solution generation using the Latin hypercube sampling for uniform sampling in bound space. These initial solutions are useful to improve the efficiency of the optimization algorithm. One important issue in developing management models for coastal aquifers incorporating the density dependent flow and transport processes is the computational feasibility. A few variations of the management model are also evaluated to test the potential for improving the computational efficiency. The variations in the formulation of management models include: direct linking of numerical simulation model, introducing meta-models [in this study artificial neural network (ANN)] in place of original numerical simulation model, using partially trained meta-model as a screening model for specifying initial solutions for the numerical simulation model linked optimization model. The developed models are capable of obtaining the nondominated front of the multiobjective management model without solving the modified single objective model iteratively. Also, the meta-model approach is shown to be computationally efficient in generating the nondominated front, especially when a partially trained (ANN-based) meta-model is used as an initial screening model for the nondominated front search process. These developed methodologies are tested for an illustrative coastal aquifer study area. The performance evaluations show potential applicability of developed methodologies for multiobjective management of saltwater intrusion in coastal aquifers.

2009




* 459(<-555): Multiobjective analysis of a public wellfield using artificial neural networks

As competition for increasingly scarce ground water resources grows,
many decision makers may come to rely upon rigorous multiobjective
techniques to help identify appropriate and defensible policies,
particularly when disparate stakeholder groups are involved. In this
study, decision analysis was conducted on a public water supply
wellfield to balance water supply needs with well vulnerability to
contamination from a nearby ground water contaminant plume. With few
alternative water sources, decision makers must balance the
conflicting objectives of maximizing water supply volume from
noncontaminated wells while minimizing their vulnerability to
contamination from the plume. Artificial neural networks (ANNs) were
developed with simulation data from a numerical ground water flow
model developed for the study area. The ANN-derived state transition
equations were embedded into a multiobjective optimization model, from
which the Pareto frontier or trade-off curve between water supply and
wellfield vulnerability was identified. Relative preference values and
power factors were assigned to the three stakeholders, namely the
company whose waste contaminated the aquifer, the community supplied
by the wells, and the water utility company that owns and operates the
wells. A compromise pumping policy that effectively balances the two
conflicting objectives in accordance with the preferences of the three
stakeholder groups was then identified using various distance-based
methods.

2007


* 460(<-631): Planning groundwater development in coastal aquifers

This study examines an approach for planning groundwater development in coastal aquifers. The seawater intrusion is controlled through a series of barrier extraction wells. The multi-objective management problem is cast as a nonlinear, nonconvex combinatorial model and is solved using a coupled simulation-optimization approach. A density-dependent groundwater flow and transport model, SEAWAT is used for simulating the dynamics of seawater intrusion. The Simulated Annealing algorithm is used for solving the optimization problem. The idea of replacing the SEAWAT model with a trained artificial neural network (ANN) to manage the computational burden within practical time frames on a desktop computer is explored. The utility of the study is demonstrated through a trade-off curve between prioritizing groundwater development and controlling seawater intrusion at desired levels.

2004





* 464(<-176): Comparison of multi-objective evolutionary neural network, adaptive neuro-fuzzy inference system and bootstrap-based neural network for flood forecasting

Accurate flood forecasting is of utmost importance in mitigating flood disasters. Flood causes severe public and economic loss especially in large river basins. In this study, multi-objective evolutionary neural network (MOENN) model is developed for accurate and reliable hourly water level forecasting at Naraj gauging site in Mahanadi river basin, India. The performance of the developed model is compared with adaptive neuro-fuzzy inference system (ANFIS) and bootstrap-based neural network (BNN) models. The performance of the models is compared in terms of Nash-Sutcliffe efficiency, root mean square error, mean absolute error and percentage deviation in peak (D). The performance of the models in forecasting floods is also evaluated using existing performance evaluation criterion of Central Water Commission, India as well as a multiple linear regression model. A partitioning analysis in conjunction with threshold statistics is carried out to evaluate the performance of the developed models in forecasting floods for low, medium and high water levels. It is found that the performance of MOENN and BNN models is more stable and consistent compared to ANFIS model. For longer lead times, the performance of MOENN model is found to be the best, with its performance in forecasting higher water levels being significantly better compared to ANFIS and BNN models. Overall, it is found that MOENN model has great potential to be applied in flood forecasting.

2013



* 518(<-269): Multi objective calibration of large scaled water quality model using a hybrid particle swarm optimization and neural network algorithm

Large scaled simulation models, especially the water quality
simulation models, are so complicated that makes calibration processes
huge tasks; in order to attain optimum solution, lots of parameters
must be calibrated, simultaneously. Methods based on evolutionary
algorithm developed new horizons in calibration procedure. Hybrid
algorithms are of the newest. In hybrid algorithms, one of the modules
is applied as a simulator and the other one takes role as an
optimization module. In this article, overcoming these challenges,
hybrid ANN-PSO algorithm is applied in calibration process of water
quality model CE-QUAL-W2. Here, Particle Swarm Optimization (PSO)
provides simulation (CE-QUAL-W2) model with sets of parameters to
simulate model. Using these results, Neural Network (estimator) is
trained. In the next step, simulator would be replaced with estimator
and Artificial Neural Network (ANN) would estimate simulator's
behavior in a way less time. The first goal is to calibrate thermal
parameter; going forward through this process needs water surface
elevation parameter to be calibrated, too. As a result, the proposed
model will become multi-objective one, applied in Karkheh reservoir in
Iran during 6 month simulation period. The proposed approach overcomes
the high computational efforts required if a conventional calibration
search technique was used, while retaining the quality of the final
calibration results. Estimator (ANN) embedded in optimization
algorithm (PSO) in calibration process, undoubtedly, reduced run time
while the answers have reliable quality.

2012




* 523(<-143): Neuro-evolutionary optimization methodology applied to the synthesis process of ash based adsorbents

Ash and modified ash were investigated as alternative adsorbents for
copper ions. Our aim was to establish optimal working conditions for
obtaining the new adsorbents, using a neuro-evolutionary optimization
methodology. The materials were characterized by SEM, FT-IR, EDAX,
XRD, and by the removal percentage. Three multilayer perceptron neural
networks were developed and aggregated into a stack to form the model
of the process. The neural model was integrated into an optimization
procedure solved with a genetic algorithm to obtain the optimum values
for the percentage of adsorption. The new adsorbents provide two
benefits: environmental protection and energy recovery. (C) 2013 The
Korean Society of Industrial and Engineering Chemistry. Published by
Elsevier B.V. All rights reserved.

2014




* 554(<-295): Modeling and genetic algorithm-based multi-objective optimization of the MED-TVC desalination system

This study proposes a systematic approach of analysis and optimization
of the multi-effect distillation-thermal vapor compression (MED-TVC)
desalination system. The effect of input variables, such as
temperature difference, motive steam mass flow rate, and preheated
feed water temperature was investigated using response surface
methodology (RSM) and partial least squares (PLS)
technique. Mathematical and economical models with exergy analysis
were used for total annual cost (TAC), gain output ratio (CUR) and
fresh water flow rate (Q). Multi-objective optimization (MOO) to
minimize TAC and maximize CUR and Q was performed using a genetic
algorithm (GA) based on an artificial neural network (ANN) model. Best
Pareto optimal solution selected from the Pareto sets showed that the
MED-TVC system with 6 effects is the best system among the systems
with 3, 4, 5 and 6 effects, which has a minimum value of unit product
cost (UPC) and maximum values of CUR and Q. The system with 6 effects
under the optimum operation conditions can save 14%, 12.5%, 2% in cost
and reduces the amount of steam used for the production of 1 m(3) of
fresh water by 50%, 34% and 18% as compared to systems with 3.4 and 5
effects, respectively. (C) 2012 Elsevier B.V. All rights reserved.

2012



* 557(<-134): Electrocoagulation efficiency and energy consumption probing by artificial intelligent approaches

Color removal efficiency (CR%) and energy consumption (EnC) of
Electrocoagulation (EC) were investigated using synthetic wastewater,
containing disperses like orange25 dye (DO25). Five operational
parameters including initial pH (pH(0)) (2, 5.5, and 9), initial dye
concentration (C-0) (20, 60, and 100mgL(-1)), applied voltage (V-EC)
(10, 20, and 30V), initial electrolyte concentration (C-S) (0, 1.5,
and 3gL(-1)), and treatment time (t(EC)) (0, 0.5, 5, 10, 15, 25, 35,
and 50min) were probed as more effective operational parameters of
EC. Combined design of experiments (DOE) was designed and experiments
were conducted in accordance with it. The experimental data were
obtained in a laboratory through a handmade batch reactor. The
achieved CR% (0-99.9) and EnC (0-69.4wh) were gained under
experimental conditions. The optimum value of C-0 was almost 20ppm
(minimum range). Two optimum clusters could be discriminated for other
four parameters. First group was corresponded to conditions with
pH(0)=9 (maximum value of range), C-S=0.7-1.1(g/lit), V-EC=10V
(minimum of range), and t(EC)=1min. Second group was corresponded to
the conditions with pH(0)=6.8 (except two cases), C-S=1.1-2(g/lit),
V-EC=10-15.2V, and t(EC)=49.4-50min. The data was used for model
building by employing two more popular models in this study: reduced
quadratic multiple regression model (SMLR) and artificial neural
network (ANN). Further statistical tests were applied to exhibit
models' goodness and to compare the models. Based on statistical
comparison, ANN models obviously outperformed SMLR models. Finally,
multi objective optimization of CR% and EnC was carried out using
genetic algorithm (GA) over the outperformed ANN models. The
optimization procedure causes nondominated optimal points, which gave
an insight into optimal operating conditions of the EC.

2014



* 560(<-348): RSM and ANN modeling for electrocoagulation of copper from simulated wastewater: Multi objective optimization using genetic algorithm approach

Performance of electrocoagulation system for the removal of copper
from synthetic wastewater was investigated using aluminum electrode
pair at four operational parameters: copper concentration (2.5-32.5 mg
L(-1)), pH (5-9), voltage (6-18 V) and treatment time (5-25
min). Metal removal efficiency and energy consumption were monitored
as responses. Experiments were conducted as per center composite
design, and the data was used for model building employing response
surface methodology (RSM) and artificial neural network approach
(ANN). Multi objective optimization for maximizing the copper removal
efficiency and minimizing the energy consumption was carried out using
genetic algorithm (GA) over the ANN model. The optimization procedure
resulted in the creation of nondominated optimal points which gave an
insight regarding the optimal operating conditions of the process. (C)
2011 Elsevier B.V. All rights reserved.

2011




* 561(<-308): Optimization studies for catalytic ozonation of cephalexin antibiotic in a batch reactor

This study examines the effect of pH of solution, cephalexin (CEX)
concentration, and O-3 dosage on removal of CEX by catalytic
ozonation. All three parameters were found to exert a significant
effect on the removal of cephalexin and on the enhancement in
biodegradability of solution. The operating conditions were optimized
using response surface methodology (RSM) and artificial neural network
(ANN). Both RSM and ANN models were found capable of predicting
removal of CEX during catalytic ozonation. Simultaneous optimization
of two responses (chemical oxygen demand (COD) removal and CEX
removal) was carried out using genetic algorithm based multiobjective
optimization. Non-dominated Pareto optimal solutions provided the
range of optimum conditions for the catalytic ozonation process. The
optimized values of pH (9.7), ozone supply (34.5 mg/L), and CEX
concentration (33.6 mg/L) using GA multiobjective optimization
corresponded to complete conversion of CEX with 72% removal of COD.

2012



* 566(<- 75): Application of Artificial Neural Network and Genetic Programming in Modeling and Optimization of Ultraviolet Water Disinfection Reactors

Ultraviolet (UV) disinfection is an environmental-friendly technology
for water treatment. However, design and operation of UV disinfection
reactors are very difficult without a good model. In this work, two
modeling methods, Artificial Neural Network (ANN) and Genetic
Programming (GP), were applied to model UV water disinfection
reactors. The model training data were obtained from simulation using
Computational Fluid Dynamics (CFD) software. The accuracy of these two
modeling methods was compared based on modeling error as well as
generalization ability for new inputs. The ANN and GP models were then
used to determine optimal design and operating variables of UV
disinfection process, using multi-objective optimization. Selected
Pareto-optimal solutions were compared using CFD simulations, and the
results are presented and discussed in this paper.

2015




* 587(<- 21): ANN-based interval forecasting of streamflow discharges using the LUBE method and MOFIPS

The estimation of prediction intervals (PIs) is a major issue limiting
the use of Artificial Neural Networks (ANN) solutions for operational
streamflow forecasting. Recently, a Lower Upper Bound Estimation
(LUBE) method has been proposed that outperforms traditional
techniques for ANN-based PI estimation. This method construct ANNs
with two output neurons that directly approximate the lower and upper
bounds of the PIs. The training is performed by minimizing a coverage
width-based criterion (CWC), which is a compound, highly nonlinear and
discontinuous function. In this work, we test the suitability of the
LUBE approach in producing Pis at different confidence levels (CL) for
the 6 h ahead streamflow discharges of the Susquehanna and Nehalem
Rivers, US. Due to the success of Particle Swarm Optimization (PSO) in
LUBE applications, variants of this algorithm have been employed for
CWC minimization. The results obtained are found to vary substantially
depending on the chosen PSO paradigm. While the returned PIs are poor
when single-objective swarm optimization is employed, substantial
improvements are recorded when a multi-objective framework is
considered for ANN development. In particular, the Multi-Objective
Fully Informed Particle Swarm (MOFIPS) optimization algorithm is found
to return valid PIs for both rivers and for the three CL considered of
90%, 95% and 99%. With average PI widths ranging from a minimum of 7%
to a maximum of 15% of the range of the streamflow data in the test
datasets, MOFIPS-based LUBE represents a viable option for
straightforward design of more reliable interval-based streamflow
forecasting models. (C) 2015 Elsevier Ltd. All rights reserved.

2015




* 588(<-100): Estimates of energy consumption in Turkey using neural networks with the teaching-learning-based optimization algorithm

The main objective of the present study was to apply the ANN
(artificial neural network) model with the TLBO
(teaching-learning-based optimization) algorithm to estimate energy
consumption in Turkey. Gross domestic product, population, import, and
export data were selected as independent variables in the
model. Performances of the ANN-TLBO model and the classical back
propagation-trained ANN model (ANN-BP (teaching learning-based
optimization) model) were compared by using various error criteria to
evaluate the model accuracy. Errors of the training and testing
datasets showed that the ANN-TLBO model better predicted the energy
consumption compared to the ANN-BP model. After determining the best
configuration for the ANN-TLBO model, the energy consumption values
for Turkey were predicted under three scenarios. The forecasted
results were compared between scenarios and with projections by the
MENR (Ministry of Energy and Natural Resources). Compared to the MENR
projections, all of the analyzed scenarios gave lower estimates of
energy consumption and predicted that Turkey's energy consumption
would vary between 142.7 and 158.0 Mtoe (million tons of oil
equivalent) in 2020. (C) 2014 Elsevier Ltd. All rights reserved.

2014




* 589(<-224): Experimental investigation, modeling and optimization of membrane separation using artificial neural network and multi-objective optimization using genetic algorithm

In this work, treatment of oily wastewaters with commercial
polyacrylonitrile (PAN) ultrafiltration (UF) membranes was
investigated. In order to do these experiments, the outlet wastewater
of the API (American Petroleum Institute) unit of Tehran refinery, is
used as the feed. The purpose of this paper was to predict the
permeation flux and fouling resistance, by applying artificial neural
networks (ANNs), and then to optimize the operating conditions in
separation of oil from industrial oily wastewaters, including
trans-membrane pressure (TMP), cross-flow velocity (CFV), feed
temperature and pH, so that a maximum permeation flux accompanied by a
minimum fouling resistance, was acquired by applying genetic algorithm
as a powerful soft computing technique. The experimental input data,
including TMP, CFV, feed temperature and pH, permeation flux and
fouling resistance as outputs, were used to create ANN models. This
fact that there is an excellent agreement between the experimental
data and the predicted values was shown by the modeling
results. Eventually, by multi-objective optimization, using genetic
algorithm (GA), an optimization tool was created to predict the
optimum operating parameters for desired permeation flux (i.e. maximum
flux) and fouling resistance (i.e. minimum fouling) behavior. The
accuracy of the model is confirmed by the comparison between the
predicted and experimental data. (C) 2012 The Institution of Chemical
Engineers. Published by Elsevier B.V. All rights reserved.

2013





* 591(<-373): Estimating Water Retention with Pedotransfer Functions Using Multi-Objective Group Method of Data Handling and ANNs

Pedotransfer functions (PTFs) have been developed to estimate soil
water retention curves (SWRC) by various techniques. In this study
PTFs were developed to estimate the parameters (theta(s), theta(r),
alpha and lambda) of the Brooks and Corey model from a data set of 148
samples. Particle and aggregate size distribution fractal parameters
(PSDFPs and ASDFPs, respectively) were computed from three fractal
models for either particle or aggregate size distribution. The most
effective model in each group was determined by sensitivity
analysis. Along with the other variables, the selected fractal
parameters were employed to estimate SWRC using multi-objective group
method of data handling (mGMDH) and different topologies of artificial
neural networks (ANNs). The architecture of ANNs for parametric PTFs
was different regarding the type of ANN, output layer transfer
functions and the number of hidden neurons. Each parameter was
estimated using four PTFs by the hierarchical entering of input
variables in the PTFs. The inclusion of PSDFPs in the list of inputs
improved the accuracy and reliability of parametric PTFs with the
exception of theta(s). The textural fraction variables in PTF1 for the
estimation of alpha were replaced with PSDFPs in PTF3. The use of
ASDFPs as inputs significantly improved alpha estimates in the
model. This result highlights the importance of ASDFPs in developing
parametric PTFs. The mGMDH technique performed significantly better
than ANNs in most PTFs.

2011




* 594(<-492): Comparisons of grey and neural network prediction of industrial park wastewater effluent using influent quality and online monitoring parameters

In this study, Grey model (GM) and artificial neural network (ANN)
were employed to predict suspended solids (SSeff) and chemical oxygen
demand (CODeff) in the effluent from a wastewater treatment plant in
industrial park of Taiwan. When constructing model or predicting, the
influent quality or online monitoring parameters were adopted as the
input variables. ANN was also adopted for comparison. The results
indicated that the minimum MAPEs of 16.13 and 9.85% for SSeff and
CODeff could be achieved using GMs when online monitoring parameters
were taken as the input variables. Although a good fitness could be
achieved using ANN, they required a large quantity of
data. Contrarily, GM only required a small amount of data (at least
four data) and the prediction results were even better than those of
ANN. Therefore, GM could be applied successfully in predicting
effluent when the information was not sufficient. The results also
indicated that these simple online monitoring parameters could be
applied on prediction of effluent quality well.

2008


* 595(<-550): Optimizing the monitoring strategy of wastewater treatment plants by multiobjective neural networks approach

This paper applies artificial neural network ( ANN) to model the observed effluent quality data. The ANN's structure, involving the number of hidden layer and node and their connection, is determined endogenously by resorting to the compromise of data cost minimization and prediction accuracy maximization. To obtain the best compromise possible, the model introduces an aspiration variable ( mu) that represents the level of aspiration achieved in one objective and the conjugate of mu, ( 1-mu), represents level of aspiration achieved in the other objective. Because a massive amount of calculation is required, the model applies genetic algorithm ( GA) for its computational flexibility and capability to ensure global solution. Feasibility and practicality of the model is tested by a case study with a set of 150 daily observations on 17 operational variables and quality parameters at an industrial wastewater treatment plant ( WTP) located in southern Taiwan. Of these 17 variables open to selection, only 6 variables, wastewater flow rate ( Q), CN-, SS, MLSS, pH and COD are selected by the model to achieve the maximum accuracy of prediction, 0.94, with a total cost of 5,950 NT$. By constraining budget availability, the variables included in the model are reduced in number, causing a concomitant reduction in prediction accuracy, that is, by varying mu ( aspiration level of accuracy), a trajectory of cost and accuracy is generated. The calculation results a cost of 3,650 NT$ and 0.54 accuracy for the case with variables including flow rate, SCN- and SS in equalization basin; aeration tank hydraulic retention time ( HRT) and percentage of returned sludge ( R%) are selected for building the prediction model when the importance of required budget is equal to the accuracy of prediction model. In addition, when required cost for building ANN model is between 3,650 NT$ and 3,900 NT$, the marginal return of budget input is highest in the entire range of calculation.

2007





* 609(<-217): Modelling and optimization of Mn/activate carbon nanocatalysts for NO reduction: comparison of RSM and ANN techniques

A response surface methodology (RSM) involving a central composite
design was applied to the modelling and optimization of a preparation
of Mn/active carbon nanocatalysts in NH3-SCR of NO at 250 degrees C
and the results were compared with the artificial neural network (ANN)
predicted values. The catalyst preparation parameters, including metal
loading (wt%), calcination temperature and pre-oxidization degree
(v/v% HNO3) were selected as influence factors on catalyst
efficiency. In the RSM model, the predicted values of NO conversion
were found to be in good agreement with the experimental
values. Pareto graphic analysis showed that all the chosen parameters
and some of the interactions were effective on response. The
optimization results showed that maximum NO conversion was achieved at
the optimum conditions: 10.2v/v% HNO3, 6.1wt% Mn loading and
calcination at 480 degrees C. The ANN model was developed by a
feed-forward back propagation network with the topology 3, 8 and 1 and
a Levenberg-Marquardt training algorithm. The mean square error for
the ANN and RSM models were 0.339 and 1.176, respectively, and the R-2
values were 0.991 and 0.972, respectively, indicating the superiority
of ANN in capturing the nonlinear behaviour of the system and being
accurate in estimating the values of the NO conversion.

2013


* 610(<-519): Combining support vector regression and ant colony optimization to reduce NOx emissions in coal-fired utility boilers

Combustion optimization has recently demonstrated its potential to reduce NOx emissions in high capacity coal-fired utility boilers. In the present study, support vector regression (SVR), as well as artificial neural networks (ANN), was proposed to model the relationship between NOx emissions and operating parameters of a 300 MW coal-fired utility boiler. The predicted NOx emissions from the SVR model, by comparing with that of the ANN-based model, showed better agreement with the values obtained in the experimental tests on this boiler operated at different loads and various other operating parameters. The mean modeling error and the correlation factor were 1.58% and 0.94, respectively. Then, the combination of the SVR model with ant colony optimization (ACO) to reduce NOx emissions was presented in detail. The experimental results showed that the proposed approach can effectively reduce NOx emissions from the coal-fired utility boiler by about 18.69% (65 ppm). A time period of less than 6 min was required for NOx emissions modeling, and 2 min was required for a run of optimization under a PC system. The computing times are suitable for the online application of the proposed method to actual power plants.

2008


* 611(<-601): Multi-objective optimization of the coal combustion performance with artificial neural networks and genetic algorithms

The present work introduces an approach to predict the nitrogen oxides
(NOx) emissions and carbon burnout characteristics of a large capacity
pulverized coal-fired boiler with an artificial neural network
(ANN). The NOx emissions and carbon burnout characteristics are
investigated by parametric field experiments. The effects of
over-fire-air (OFA) flow rates, coal properties, boiler load, air
distribution scheme and nozzle tilt are studied. An ANN is used to
model the NOx emissions characteristics and the carbon burnout
characteristics. A genetic algorithm (GA) is employed to perform a
multi-objective search to determine the optimum solution of the ANN
model, finding the optimal setpoints, which can suggest operators'
correct actions to decrease NOx emissions and the carbon content in
the flyash simultaneously, namely, get a good boiler combustion
performance with high boiler efficiency while keeping the NOx emission
concentration meet the requirement. Copyright &COPY; 2005 John Wiley &
Sons, Ltd.

2005





* 648(<-110): A Methodological Study for Optimizing Material Selection in Sustainable Product Design

A computational intelligence-based identification of the properties of
maximally sustainable materials for a given application, as derived
from key properties of existing candidate materials, is put
forward. The correlation surface between material properties (input)
and environmental impact (EI) values (output) of the candidate
materials is initially created using general regression (GR)
artificial neural networks (ANNs). Genetic algorithms (GAs) are
subsequently employed for swiftly identifying the minimum point of the
correlation surface, thus exposing the properties of the maximally
sustainable material. The ANN is compared to and found to be more
accurate than classic polynomial regression (PR)
interpolation/prediction, with sensitivity and multicriteria analyses
further confirming the stability of the proposed methodology under
variations in the properties of the materials as well as the relative
importance values assigned to the input properties. A nominal
demonstration concerning material selection for manufacturing
maximally sustainable liquid containers is presented, showing that by
appropriately picking the pertinent input properties and the desired
material selection criteria, the proposed methodology can be applied
to a wide range of material selection tasks.

2014




* 652(<-551): Multi-step-ahead neural networks for flood forecasting

A reliable flood warning system depends on efficient and accurate
forecasting technology. A systematic investigation of three common
types of artificial neural networks (ANNs) for multi-step-ahead (MSA)
flood forecasting is presented. The operating mechanisms and
principles of the three types of MSA neural networks are explored:
multi-input multi-output (MIMO), multi-input single-output (MISO) and
serial-propagated structure. The most commonly used multi-layer
feed-forward networks with conjugate gradient algorithm are adopted
for application. Rainfall-runoff data sets from two watersheds in
Taiwan are used separately to investigate the effectiveness and
stability of the neural networks for MSA flood forecasting. The
results indicate consistently that, even though the MIMO is the most
common architecture presented in ANNs, it is less accurate because its
multi-objectives (predicted many time steps) must be optimized
simultaneously. Both MISO and serial-propagated neural networks are
capable of performing accurate short-term (one- or two-step-ahead)
forecasting. For long-term (more than two steps) forecasts, only the
serial-propagated neural network could provide satisfactory results in
both watersheds. The results suggest that the serial-propagated
structure can help in improving the accuracy of MSA flood forecasts.

2007




* 661(<- 15): Analyzing the effect of various soil properties on the estimation of soil specific surface area by different methods

Depending on the method used, measuring the specific surface area
(SSA) can be expensive and time consuming and limited numbers of
studies have been conducted to predict SSA from soil properties. In
this study, 127 soil sample data were gathered from the available
literature. The data set included SSA values and some of the soil
physical and chemical index properties. At the first step, linear
regression, non-linear regression, regression trees, artificial neural
networks, and a multi-objective group method of data handling were
used to develop seven pedotransfer functions (PTFs) for the purpose of
finding the best method in predicting SSA. Results showed that the
artificial neural networks performed better than the other methods
used in the development and validation of PTFs. At the second step, to
find the best set of SSA for predicting input variables and to
investigate the importance of the input parameters, the artificial
neural networks were further used and 25 models were developed. The
results showed that the PTF, containing the input variables of sand%,
clay%, plastic limit, liquid limit, and free swelling index performed
better than the other PTEs. This can be attributed to the close
relation between the free swelling index and Atterberg limits with the
soil clay mineralogy, which is one of the most important factors
controlling SSA. The sensitivity analysis showed that the greatest
sensitivity coefficients were found for the cation exchange capacity,
clay content, liquid limit, and plasticity index in different
models. Overall, the artificial neural networks method was proper to
predict SSA from soil variables. (C) 2015 Elsevier B.V. All rights
reserved.

2015