Thesis (M.Sc. (Computer Science))--North-West University, Potchefstroom Campus, 2008.

Computational intelligence techniques, such as artificial neural networks (ANNs), have been widely used to improve the performance of power system monitoring and control. Although inspired by the neurons in the brain, ANNs are largely different from living neuron networks (LNNs) in many aspects. Due to the oversimplification, the huge computational potential of LNNs cannot be realized by ANNs. Therefore, a more brain-like artificial neural network is highly desired to bridge the gap between ANNs and LNNs. The focus of this research is to develop a biologically inspired artificial neural network (BIANN), which is not only biologically meaningful, but also computationally powerful. The BIANN can serve as a novel computational intelligence tool in monitoring, modeling and control of the power systems. A comprehensive survey of ANNs applications in power system is presented. It is shown that novel types of reservoir-computing-based ANNs, such as echo state networks (ESNs) and liquid state machines (LSMs), have stronger modeling capability than conventional ANNs. The feasibility of using ESNs as modeling and control tools is further investigated in two specific power system applications, namely, power system nonlinear load modeling for true load harmonic prediction and the closed-loop control of active filters for power quality assessment and enhancement. It is shown that in both applications, ESNs are capable of providing satisfactory performances with low computational requirements. A novel, more brain-like artificial neural network, i.e. biologically inspired artificial neural network (BIANN), is proposed in this dissertation to bridge the gap between ANNs and LNNs and provide a novel tool for monitoring and control in power systems. A comprehensive survey of the spiking models of living neurons as well as the coding approaches is presented to review the state-of-the-art in BIANN research. The proposed BIANNs are based on spiking models of living neurons with adoption of reservoir-computing approaches. It is shown that the proposed BIANNs have strong modeling capability and low computational requirements, which makes it a perfect candidate for online monitoring and control applications in power systems. BIANN-based modeling and control techniques are also proposed for power system applications. The proposed modeling and control schemes are validated for the modeling and control of a generator in a single-machine infinite-bus system under various operating conditions and disturbances. It is shown that the proposed BIANN-based technique can provide better control of the power system to enhance its reliability and tolerance to disturbances. To sum up, a novel, more brain-like artificial neural network, i.e. biologically inspired artificial neural network (BIANN), is proposed in this dissertation to bridge the gap between ANNs and LNNs and provide a novel tool for monitoring and control in power systems. It is clearly shown that the proposed BIANN-based modeling and control schemes can provide faster and more accurate control for power system applications. The conclusions, the recommendations for future research, as well as the major contributions of this research are presented at the end.

As the distributed systems becomes popular, efficient load balancing systems taking better decisions must be designed. The most important reasons that necessitate load balancing in a distributed system are the heterogeneous hosts having different com- puting powers, external loads and the tasks running on different hosts but communi- cating with each other. In this thesis, a load balancing approach, called RALBANN, developed using graph partitioning and artificial neural networks (ANNs) is de- scribed. The aim of RALBANN is to integrate the successful load balancing deci- sions of graph partitioning algorithms with the efficient decision making mechanism of ANNs. The results showed that using ANNs to make efficient load balancing can be very beneficial. If trained enough, ANNs may load the balance as good as graph partitioning algorithms more efficiently.

Modeling and inferential control studies are carried out on a reactive batch distillation system for the esterification reaction of ethanol with acetic acid to produce ethyl acetate. A dynamic model is developed based on a previous study done on a batch distillation column. The column is modified for a reactive system where Artificial Neural Network Estimator is used instead of Extended Kalman Filter for the estimation of compositions of polar compounds for control purposes. The results of the developed dynamic model of the column is verified theoretically with the results of a similar study. Also, in order to check the model experimentally, a lab scale column (40 cm height, 5 cm inner diameter with 8 trays) is used and it is found that experimental data is not in good agreement with the models&rsquo / . Therefore, the model developed is improved by using different rate expressions and thermodynamic models (fi-fi, combination of equations of state (EOS) and excess Gibbs free energy (EOS-Gex), gama-fi) with different equations of states (Peng Robinson (PR) / Peng Robinson - Stryjek-Vera (PRSV)), mixing rules (van der Waals / Huron Vidal (HV) / Huron Vidal Original (HVO) / Orbey Sandler Modification of HVO (HVOS)) and activity coefficient models (NRTL / Wilson / UNIQUAC). The gama-fi method with PR-EOS together with van der Waals mixing rule and NRTL activity coefficient model is selected as the best relationships which fits the experimental data. The thermodynamic models / EOS, mixing rules and activity coefficient models, all are found to have very crucial roles in modeling studies. A nonlinear optimization problem is also carried out to find the optimal operation of the distillation column for an optimal reflux ratio profile where the maximization of the capacity factor is selected as the objective function. In control studies, to operate the distillation system with the optimal reflux ratio profile, a control system is designed with an Artificial Neural Network (ANN) Estimator which is used to predict the product composition values of the system from temperature measurements. The network used is an Elman network with two hidden layers. The performance of the designed network is tested first in open-loop and then in closed-loop in a feedback inferential control algorithm. It is found that, the control of the product compositions with the help of an ANN estimator with error refinement can be done considering optimal reflux ratio profile.

In the domain of the complex control problems for agents, neuroevolution, i.e. artificial evolution of neural networks, methods have been continuously shown to offer high performance solutions which may be unpredictable by external controller design. Recent studies have proved that these methods can also be successfully applied for cooperative multi-agent systems to evolve the desired team behavior. For a given task which may benefit from both cooperation and behavioral specialization, the genetic diversity of the team members may have important effects on the team performance. In this thesis, the single prey hunting problem is chosen as the case, where the performance of the evolved predator teams with distinct genetic similarity levels are systematically examined. For this purpose, three similarity levels, namely homogeneous, partially heterogeneous and heterogeneous, are adopted and analyzed in various problem-specific and algorithmic settings. Our similarity levels differ from each other in terms of the number of groups of identical agents in a single predator team, where identicalness of two agents refers to the fact that both have the same synaptic weight vector in their neural network controllers. On the other hand, the problem-specific conditions comprise three different fields of vision for predators, whereas algorithmic settings refer to varying number of individuals in the populations, as well as two different selection levels such as team and group levels. According to the experimental results within a simulated grid environment, we show that different genetic similarity level-field of vision-algorithmic setting combinations beget different performance results.

Energy is increasingly becoming more important in today&rsquo / s world. Ascending of energy consumption due to development of technology and dense population of earth causes greenhouse effect. One of the most valuable energy sources is hydro energy. Because of limited energy sources and excessive energy usage, cost of energy is rising. There are many ways to generate electricity. Among the electricity generation units, hydroelectric power plants are very important, since they are renewable energy sources and they have no fuel cost. Electricity is one of the most expensive input in production. Every hydro energy potential should be considered when making investment on this hydro energy potential. To decide whether a hydroelectric power plant investment is feasible or not, project cost and amount of electricity generation of the investment should be precisely estimated. This study is about cost estimation of hydroelectric power plant projects. Many design parameters and complexity of construction affect the cost of hydroelectric power plant projects. In this thesis fifty four hydroelectric power plant projects are analyzed. The data set is analyzed by using regression analysis and artificial neural network tools. As a result, two cost estimation models have been developed to determine the hydroelectric power plant project cost in early stage of the project.

The economic evaluation of a new airport investment requires the use of estimated future air passenger demand.Today it is well known that air passenger demand is basicly dependent on various socioeconomic factors of the country and the region where the planned airport would serve. This study is focused on estimating the future air passenger demand for planned airports in Turkey where the historical air passsenger data is not available.For these purposses, neural networks and multi-linear regression were used to develop forecasting models. As independent variables,twelve socioeconomic parameters are found to be significant and used in models. The available data for the selected indicators are statistically analysed and it is observed that most of the data is highly volatile, heteroscedastic and show no definite patterns. In order to develop more reliable models, various methods like data transformation, outlier elimination and categorization are applied to the data.Only seven of total twelve indicators are used as the most significant in the regression model whereas in neural network approach the best model is achieved when all the twelve indicators are included. Both models can be used to predict air passenger demand for any future year for Or-Gi and Zafer Airports and future air passenger demand for similar airports. Regression and neural models are tested by using various statistical test methods and it is found that neural network model is superior to regression model for the data used in this study.

The estimation of effort is at the heart of project tasks, since it is used for many purposes such as cost estimation, budgeting, monitoring, project planning, control and software investments. Researchers analyze problems of the estimation, propose new models and use new techniques to improve accuracy. However up to now, there is no comprehensive estimation methodology to guide companies in their effort estimation tasks. Effort estimation problem is not only a computational but also a managerial problem. It requires estimation goals, execution steps, applied measurement methods and updating mechanisms to be properly defined. Besides project teams should have motivation and responsibilities to build a reliable database. If such methodology is not defined, common interpretation will not be constituted among software teams of the company, and variances in measurements and divergences in collected information prevents to collect sufficient historical information for building accurate models. This thesis proposes a methodology for organizations to manage and execute effort estimation processes. The approach is based on the reported best practices, v empirical results of previous studies and solutions to problems &amp / conflicts described in literature. Five integrated processes: Data Collection, Size Measurement, Data Analysis, Calibration, Effort Estimation processes are developed with their artifacts, procedures, checklists and templates. The validation and applicability of the methodology is checked in a middle-size software company. During the validation of methodology we also evaluated some concepts such as Functional Similarity (FS) and usage of Base Functional Components (BFC) in effort model on a reliable dataset. By this way we evaluated whether these subjects should be a part of methodology or not. Besides in this study it is the first time that the COSMIC has been used for Artificial Neural Network models.

With the advent of Advanced Traveler Information Systems (ATIS), short-term travel time prediction is becoming increasingly important. Travel time can be obtained directly from instrumented test vehicles, license plate matching, probe vehicles etc., or from indirect methods such as loop detectors. Because of their wide spread deployment, travel time estimation from loop detector data is one of the most widely used methods. However, the major criticism about loop detector data is the high probability of error due to the prevalence of equipment malfunctions. This dissertation presents methodologies for estimating and predicting travel time from the loop detector data after correcting for errors. The methodology is a multi-stage process, and includes the correction of data, estimation of travel time and prediction of travel time, and each stage involves the judicious use of suitable techniques. The various techniques selected for each of these stages are detailed below. The test sites are from the freeways in San Antonio, Texas, which are equipped with dual inductance loop detectors and AVI. ?? Constrained non-linear optimization approach by Generalized Reduced Gradient (GRG) method for data reduction and quality control, which included a check for the accuracy of data from a series of detectors for conservation of vehicles, in addition to the commonly adopted checks. ?? A theoretical model based on traffic flow theory for travel time estimation for both off-peak and peak traffic conditions using flow, occupancy and speed values obtained from detectors. ?? Application of a recently developed technique called Support Vector Machines (SVM) for travel time prediction. An Artificial Neural Network (ANN) method is also developed for comparison. Thus, a complete system for the estimation and prediction of travel time from loop detector data is detailed in this dissertation. Simulated data from CORSIM simulation software is used for the validation of the results.

Two new models to calculate the species instantaneous and filtered reaction rates for multi-step, multi-species chemical kinetics mechanisms are developed based on the artificial neural networks (ANN) approach. The proposed methodologies depend on training the ANNs off-line on a thermo-chemical database representative of the actual composition and turbulence level of interest. The thermo-chemical database is constructed by stand-alone linear eddy mixing (LEM) model simulations under both premixed and non-premixed conditions, where the unsteady interaction of turbulence with chemical kinetics is included as a part of the training database. In this approach, the information regarding the actual geometry of interest is not needed within the LEM computations. The developed models are validated extensively on the large eddy simulations (LES) of (i) premixed laminar-flame-vortex-turbulence interaction, (ii) temporally mixing non-premixed flame with extinction-reignition characteristics, and (iii) stagnation point reverse flow combustor, which utilizes exhaust gas re-circulation technique. Results in general are satisfactory, and it is shown that the ANN provides considerable amount of memory saving and speed-up with reasonable and reliable accuracy. The speed-up is strongly affected by the stiffness of the reduced mechanism used for the computations, whereas the memory saving is considerable regardless.