Sukumar Kamalasadan (ETA987083)

Special Study Report

Advisor Dr. D. Thukaram

Electric Power Systems Management, Energy Program, SERD, Asian Institute of Technology, Bangkok, Thailand November 1998

Application of Artificial Intelligence techniques in Power System

ABSTRACT A reliable, continuos supply of electrical energy is essential for the functioning of today's modern complex and advanced society. Electricity is one of the prime factors for the growth and determines the value of the society. Manual calculation, technical analysis and conclusions initially adopted the power system design, operation and control. As the power system grew it became more complex due to the technical advancements, variety and dynamic requirements. Conventional Power System analysis become more difficult due to 1. Complex versatile and large amounts of data that are used in calculation, diagnosis and learning. 2. The increase in the computational time period and the accuracy due to extensive system data handling. The modern power system operates close to their limits due to the increasing energy consumption and impediments of various kinds, and the extension of existing electric transmission networks. This situation requires a significantly less conservative power system operation and control regime which, in turn, is possible only by monitoring the system states in much more detail than was necessary previously. Sophisticated computer tools have become predominant in solving the difficult problems that arise in the areas of Power System planning, operation, diagnosis and design of the systems. Among these computer tools Artificial Intelligence has grown extensively in recent years and has been applied in the areas of the power systems. The most widely used and important ones of Artificial Intelligent tools, applied in the field of Electrical Power Systems are the Artificial Neural networks and the so-called Fuzzy systems. This special study gives a review of the Artificial Intelligence (Both artificial Neural Network and Fuzzy systems) basic principles and the concepts, along with the application of these tools in the power systems areas. A survey of the applications of ANN and Fuzzy systems in the field of power systems is complied and presented and the details of the important application are discussed. Finally the major achievements of this soft computing technique in power system areas are commented and the future scopes of these methods in the modern power system are analyzed.

Special Study Report

Page i

Application of Artificial Intelligence techniques in Power System

Table of Contents



Pag e

Title Page Table of Contents Abstract List of Figures 1 Introduction 1.1 Back Ground 1.2 Neural network and its application 1.3 Fuzzy sets/logic and its application 1.4 Structure of the Study 2 Artificial Neural Network 2.1 Definition of the Neural Network 2.2 Fundamentals of artificial Neural Network 2.3 Neural Network Design 2.4 Learning, Recall and Memory in ANN 2.5 When and why using Neural Network 2.6 An Overview of the well known ANN Models Fuzzy Logic and Fuzzy Systems 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 4 Importance of Fuzzy Systems Basic Concepts Fuzzy Sets and Rules Classical Operations of Fuzzy Sets Membership function and membership values Fuzzy Relations Properties of Fuzzy Sets Fuzzy Truth Value Learning in Fuzzy Systems Fuzzy Logic Controllers (FLC) Pattern Recognition in Fuzzy Systems Relational Data Adaptivity features and Adaptive Controllers

i,ii iii iv 1 1 1 2 2 4 4 4 5 6 8 9 17 17 17 18 18 19 19 19 20 20 21 21 22 23 24
Page ii


Application of Artificial Neural Networks in Power Systems

Special Study Report

7 Load Forecasting 5.Application of Artificial Intelligence techniques in Power System 4.2 Advantages of Fuzzy Logic Application Conclusion Bibliography 7 Special Study Report Page iii .2.4 State Estimation 5.5 Security Assessment 5.8 Protection 4.3 Generator Operation and Control 5.1 Reactive Power Control 5.1 Requirements of Fuzzy based Application Knowledge Consistency and Interaction with the User Introduction onFuzzy logic application 5.3 Fault Diagnosis Voltage Stability Enhancement 6 Analysis of the Techniques 6.3 Hopfield Network 6.4 Security Assessment 4.4 Training the Inputs 6.2 Fuzzy Logic based Application 6.1 Power System Stabilizer Practical Implementation 6.2 Transient Stability 5.2.6 Contingency Screening Introduction on ANN application 24 25 25 26 27 30 31 31 32 32 33 34 34 34 34 38 38 40 40 41 41 42 44 44 44 44 45 45 45 45 46 46 46 48 49 4.2.5 State Estimation 4.1 Design of Network 6.1.6 Fault Diagnosis and Restoration Training Set Generation 6.1 Neural Network based Application 6.9 Load Modeling 5 Application of Fuzzy Logic in the Power System 5.2 Load Forecasting 4.2 Major applications Major Applications 4.2.7 Voltage Stability Assessment 4.2.

2 Figure 5.6 Figure 3.1 Schematic Diagram of the Neuron Ways of Implementing a Solution to a Specific Problem Overview of the Main ANN models Three Layer Feedforward Neural Network Back Propagation Algorithm/Network Typical RBF Network Truth Values in Fuzzy Logic The Characterization of Pattern Recognition An Adaptive Fuzzy Controller Modular Neural Network Feedforward Architecture Unsupervised/Supervised forecasting Fault Diagnosis process Procedure Adopted for Load 4 9 10 11 13 14 20 22 23 26 28 29 36 The membership function of controlling ability of controlling devices The membership function of Voltage violation Level Computation Procedure for the solution for Voltage Profile Enhancement Figure 5.4 Figure 2.2 Figure 3.1 Figure 3.5 Figure 2.3 Figure 2.2 Figure 4.1 Figure 2.3 37 37 Special Study Report Page iv .Application of Artificial Intelligence techniques in Power System List of Figures Figure 2.1 Figure 4.3 Figure 4.3 Figure 5.2 Figure 2.

functional reasoning hybrid application along with Neural nets. The Neural network theory was first adopted in 1940 where the starting point was the learning law proposed by ITEBB in 1949. Fuzzy Systems or logic’s as introduced by Zadeh [LAZ 65] in 1965 has basically introduced to solve inexact and vague concepts by relating those using multi-valued ness in a logical way.. Power system related applications became a useful tool for the online researchers in this field. to mimic the computations of the brain [VVR 93]. which demonstrated how neurons could exhibit learning behavior. optimal control. traction control. Further as a part of developing it as mathematics the applications of these theories were considered in different areas. signal processing. Artificial Neural Networks (referred as ANN here on) and the Fuzzy logic (systems) that are considered as the so called soft computing methods are now a days becoming predominant tools in the area of Artificial Intelligence linked application oriented methods. Among the major applications viz. They embody computational paradigms. The application further waxed and waned away because of the lack of powerful technological advancement. First stream among this is concerned with modeling the brain and thereby explains its cognitive behavior. Special Study Report Page 1 of 56 . speech recognition. semiconductors. The application of ANN's in the power systems belongs to this category and is one of the recent interesting topics in the Power System Engineering.2 Neural Network and its Applications ANN is biologically inspired and represented as a major extension of computation. The application of fuzzy systems were mainly in the field of modal interface. Main applications of the ANN’s can be divided into two principal streams.1 Back Ground The increasing prominence of the computers has led to a new way of looking at the world. business other than in almost all the areas of the power systems. learning. The primary aim of researchers in the second stream is to construct useful ‘computers’ for real world problems of classification or Pattern Recognition by drawing on these principles. 1. based on biological metaphor. analysis. robotics. information. Further to the technological advancement in the field of ANN. researchers were attracted on their important applications where logical and relational thinking is required. database.Application of Artificial Intelligence techniques in Power Systems CHAPTER 1 INTRODUCTION 1. Earlier research in this field was based on mathematical understanding of set theory and probability. The resurgence occurred recently due to the new methods that are emerging as well as the computational power suitable for simulation of interconnected neural networks. The improved understanding of the functioning of neuron and the pattern of its interconnection has enabled researchers to produce the necessary mathematical modes for testing their theories and developing practical applications.

Application of Artificial Intelligence techniques in Power Systems 1. Since the literatures covering the wide range of topics are extensive. and is used synonymously with fuzzy set theory. They are model free systems. Special Study Report Page 2 of 56 . The basic rules. Many promising applications have been reported in the broad fields of system control. the use of Fuzzy logic / set is increasing in the power systems problems. Assessment. Subsequently critical evaluations and the potential and scope of further areas of work in the related fields are summarized for the benefit of the researchers interested in these areas. in which all things are matters of degree. optimization. Diagnosis. Due to these reasons. Fuzzy logic is a logical system for formalization of approximate reasoning. Fuzzification offers superior expressive power. the importance and the technical details are discussed the Chapter three. The basic ANN applications mainly cover the areas like control.3 Fuzzy sets / logic and its Applications Fuzzy set theory systems provide tools for representing and manipulating inexact concepts and the ambiguity prevalent in the human interpretations and thought process. Basic concepts of Neural network including the learning features are explained in the Chapter two. Moreover an overview of the well-known ANN models and the comparison between them highlighting the main advantages is reviewed. This theory devices from the fact that almost all natural classes and concepts are fuzzy rather than crisp in nature. decision support. Screening. The concept of Fuzzy Rules and systems. the concept of the socalled Fuzzy Logic Controllers (FLC). and the adaptive features of Fuzzy Sets are analyzed. the purpose and use of the ANN were further detailed. The purpose of this study is to focus attention on the most significant works as a part of the application of AI in power systems involving typical power systems problems. as it is in all intelligent processing. forecast. Chapter four mainly deals with the application of ANN in the field of Power Systems. system analysis and planning. information processing. The various research works on ANN application in the various areas in the Power Systems were reviewed. The structure of the Neural network. It can be considered as super set of classical (Boolean) logic which users multiple truth-values to handle the concepts of partial truth. Modeling. in the field of power systems. diagnosis. Moreover Pattern Recognition technique. greater generality and an improved capability to model complex problems at a low solution cost. its design and construction were discussed. the properties and definitions of this theory are and the operations are seen.4 Structure of the Study This study reviews basics of both ANN and fuzzy logic along with the recent works reported on these tools. The training of ANN. 1. They provide an excellent framework to more completely and effectively model uncertainty and the imperious in human reasoning with the use of linguistic variables with membership functions. the main consideration is to the important works in the different field of power systems.

Application of Artificial Intelligence techniques in Power Systems Similar in line. Diagnosis. Forecasting. Assessment. Concluding the Strengths of these techniques and the abilities are illustrated. Chapter four details the application of Fuzzy Logic’s in Power Systems. Main applications cover Stability Control. Planning and Estimation. Special Study Report Page 3 of 56 . Further the analysis of these techniques is done in chapter six with a view to importance of various applications and the further scope of research.

The fixed weighted inputs may be either from the previous layers of ANN or from the output of neurons. Special Study Report Page 4 of 56 . it forms the feedback architecture it has feedforward architecture.Application of Artificial Intelligence techniques in Power Systems CHAPTER 2 ARTIFICIAL NEURAL NETWORK 2. If these inputs are derived from neuron outputs. As basic element the neurons are not powerful but their interconnections allow encoding relationship between variables of the problems to which it is applied and providing very powerful processing capabilities. x3… xn)) times the corresponding weights Wij = (wi1.1 Schematic Diagram of the Neuron General model of the processing unit of ANN can be considered to have the following three elements. This output signal is usually a non-linear function of the input vector and a weight vector. called Neurons. The main differences between various existing models of ANN are mainly in their architectures or the way their basic processing elements (neurons) are interconnected. 2.1 Definition of the Neural Network Neural networks are systems that typically consist of a large number of simples processing units.……. x2. The weight vector is adjusted in a training phase by using a large set of examples and the learning rate. The learning rule adapts the weight of all neurons in networks in order to learn an underlying relation in the training example. win). Weighted Summing Unit The weighted summing unit consists of external or internal inputs (Xi (x1. Incoming Weighted Connections Neuron Output = F ( Σ Inputs ) Outgoing Weighted Connections Figure 2. wi2. The function to be performed on the Input vector is hence defined by the non-linear function and the weight vector of the neuron.2 Fundamentals of a Artificial Neural Network Elementary processing unit of ANN’s is neuron. Generally it contains several inputs but has only one output. A neuron has generally a high-dimensional Input vector and one single output signal.

2. which do not process information but distribute information to other units. For example each output unit represents a class of objects.3 Neural Network Design A neural network element is a smallest processing unit of the whole network essentially forming a weighted sum and transforming it by the activation function to obtain the output. The manner in which actually the neurons are connected together depends on the different classes of the neural networks. several neurons are interconnected together. They can be confined to a range and are adjustable during network training. The number of layers and the number of nodes per layer often specify the framework. 2. The types of layer include Input Layer where the nodes are called input units. Hidden Layer(s) where the nodes are called hidden units. 1. sigmoid. It is a static nonlinear function which may be pulse type or step type. a proportional. which are not directly observable.3. ANNs have parallel distributed architecture with a large number of nodes and connections. (i) Determination of network topology (ii) Determination of system (activation & synaptic) dynamics Determination of the Network Topology The topology of the neural network refers to its framework as well as its interconnection scheme.1 ANN Architecture Construction of neural Network involves the following tasks. a time delay or a combination of these. Perception models consist of weighted summing unit having no feedback inputs. In order to gain sufficient computing power. 2. They provide into the networks the capability to map or classify nonlinear problems. Some of the examples of such functions are threshold. Other main important concept is the weightage for the connected unit. Example: Following two general functions can be used to relate input Pi with output Qi as (a1. Tan hyperbolic or Gaussian functions. no dynamic function and signal as non-linear function. differentiable (smooth) or non-identification (sharp) and having positive mean or zero mean. This block may exist for time varying signals and introduces a function that is an integral. The Output Layer where the nodes are called output units. It can be real or integer numbers. all of them attain fixed values. which encode possible concepts (or values) to be assigned to the instance under consideration. Different characteristics of neurons can be evolved using different type and combination of the above three of its basic components. Basically neurons are arranged in layers. Special Study Report Page 5 of 56 . Feedback or dynamic networks utilize the dynamic function block. When training is completed.a2)Qi (t) = Pi (t) Qi (t) = Pi (t-T) Non linear function This decides the firing of neuron for a given input values.Application of Artificial Intelligence techniques in Power Systems Linear Dynamical Function It is essentially a single input or single output function block.

Temporal learning encodes a sequence of patterns necessary to achieve final outcome. The synaptic weights form Long Term Memory (LTM) where as the activation's state forms Short Term Memory (STM) of the network. The mechanism for learning alters the weights associated with the various interconnections and thus leads to a modification in the strength of interconnection. a training pair consists of an input vector and a desired target vector. The development of any ANN involves two phases: Learning or Training phase and Recall or testing phase. 2. how long and how often to present each association for training and supplying performance error information. while computing the activation dynamics. Training patterns with examples carried out training in the network. Supervised learning is further classified as Structural learning / Temporal learning.4 Learning. These techniques include deciding. The synaptic dynamics dictates the learning process. Memory.Application of Artificial Intelligence techniques in Power Systems Determination of Systems (Activation & Synaptic) Dynamics The dynamics of the network determines its operation. ANN’s can be trainable nonlinear dynamical systems. This is called testing phase. Recall and Memory in ANN Learning in a neural network essentially consists of modifying in some systematic manner the interconnection strengths between the neural units. Synaptic activation determines the change in the synaptic weights. In the supervised learning. it may be presented with new unknown patterns and its efficiency can be checked. The memory in ANN can be a Content Addressable Memory (CAM). The difference constitutes an error that is used to modify network weights in a manner that reduces the error in subsequent training cycles. Special Study Report Page 6 of 56 . is in form of values of weights of the interconnecting links. Therefore. Neural dynamics consists of two parts one which corresponding to the dynamics of activation states and the other corresponding to the dynamics of synaptic weights. The activation dynamics determines the time evolution of the neural activation’s. when to turn off the learning. Learning methods can be classified into two categories Supervised learning Unsupervised learning Supervised learning is the process that incorporates an external guidance. in ANN. Synaptic weights change gradually. where it stores the data at stable state in memory (or weight) matrix W or an Associate Memory which provides output response from input stimuli. Once the network has learnt the problem. whereas the neuron's activation's fluctuate rapidly. the system weights are assumed to be constant. ANN uses memory to learn and adapt. This is achieved by observing the system in question to see how the process evolves with time or in response to additional external actions. Structural learning encodes the proper auto associate (single pattern vector) or heteroassociate vector of patterns pair mapping into weight matrix W.

The learning algorithm specifies how the network adapts its weights for all given vectors. During the training phase the weights of ANN stabilize and while testing for an unknown pattern gives the output without a time-delay of learning phase. a) Approximation Suppose a nonlinear input/output mapping is given described by the functional relationship d = g(x) where x is the input vector and the scalar d is the output. given a set of the input/output pairs (x1. b) Pattern Classification In the pattern classification there are fixed number of categories into which activation's are classified. which is new but belongs to the same kind of the patterns used in the network.d2)…. Further to that the neural network has to classify this new pattern correctly.Application of Artificial Intelligence techniques in Power Systems In the Unsupervised learning no target vector exists. Some of the most popularly used rules for learning includes Hebb's rule and Delta rule for single layer (perception) ANN. Some of the learning tasks that benefit the use of neural networks are as follows.4.d1). To resolve this activation classification neural network undergoes training. Backpropagation algorithm for multilayer (perception) ANN. especially Special Study Report Page 7 of 56 . 2.(x2. the network provides output in just one pass and allows flow of signal in only one direction from input to hidden and to output layers. The processing algorithm specifies how the neural network with a given set of weights calculates the output vector for any input vector. The supervised learning can also be viewed as functional mapping problem. which a neural network is required to perform. The input vector is applied to the network and the system “self organizes” so that a consistent output (possibly unpredicted before training) is produced. The architecture specifies the way the neurons are connected. signals can flow amongst neurons in either direction and /or recursively. In feedforward network. In feedback network. The requirement is to design a neural network that approximates the non-linear function g(x). The approximation problem is the main example for supervised learning. The function g(x ) is assumed to be unknown. its processing algorithm and its learning algorithm characterize a neural network. The advantage of using the neural network to perform pattern classification is that ANN can construct non-linear decision boundaries between the different classes in a nonparametric fashion and thereby offer a practical method of solving otherwise highly complex pattern classification problems. After that a new pattern is presented to the network. Thus its architecture.dn).(xn . There is also the unsupervised learning in pattern classification.1 Learning Tasks The choice of a particular learning procedure is very much influenced by the learning task. The pattern recognition can be classified as a supervised learning problem. In the training the network is repeatedly presented a set of patterns along with the categories where the pattern belongs. The recall or testing depends on the interconnection of the network.

The task of the neural net is to group the set of training vectors into clusters based on some kind of similarities. The main type of learning can be supervised and unsupervised learning. Kohonen net uses unsupervised learning. d) Association The two types of associations are Auto association and Hetero association. Auto association involves the use of unsupervised learning whereas the type of learning involved in hetero-association is of a supervised nature. These output values are then used as a basis for the correction of the weights. Hetero Association differs from Auto association in that an arbitrary set of input patterns are paired with another arbitrary set of output patterns. …. are used to predict the present sample x (n). Supervised learning is done through a set of examples where each example consists of the input values and target output values. In others no assumption is made with respect to the number and the nature of the clusters. where in the set of m past samples that are uniformly spaced in time. we may compute the prediction error e(n) = x(n) . x (n-2). In auto association a neural network is required to store a set of patterns by repeatedly presenting them to the network. Sample x (n) serves the purpose of the desired response.5 When and why using Neural Network Neural set is basically a new way of solving the problems. c) Prediction Prediction is most basic task. The main difference between different classes of the network can be based on the learning approach. Based on the previous samples x (n-1). In some of the cases the number of clusters or their diameter is determined before training. which way can successfully be followed for a number of problems.x(n | n-1. The single layer feed-forward net and the Backpropagation nets use supervised learning Unsupervised learning has a set of examples where the input conditions are known but the associated target output conditions are not given. Here unsupervised learning is used to perform the role of adaptive feature extraction or clustering prior to Pattern Recognition. For some problem neural network is not however Special Study Report Page 8 of 56 . Prediction may be viewed as the form of the model building in the sense that smaller the prediction error in a statistical sense the better will the network serve as the physical model of the underlying stochastic process responsible for the generation of the time-series. N-m) and thus the error-correction learning is used to modify the weights of the network. It’s a signal processing problem.…. 2. However when simulated with a particular input. Also network is presented a partial and distorted version of an original pattern stored in it. x(n-m) .. it is not known beforehand to which cluster the output obtained from the net belongs. Now the network is asked to recall that particular pattern.Application of Artificial Intelligence techniques in Power Systems when there is no prior knowledge of the categories into which the activation patterns are to be classified. When the process is of nonlinear in nature then the use of ANN provides a powerful method for solving the prediction problem by virtue of the non-linear processing units built into its construction.

For number of problems the implementation of the solution in Neural Network is useful. Main difference of using the Neural Network and conventional method of solving problems are. McCulloh_Pitts Special Study Report Page 9 of 56 . training examples are presented to the networks and the weights of the neural networks are adapted by a learning rate. or consume too much power. Neural Network is trained to perform satisfactory. “Too complicated” means that implemented (conventional) algorithms are either too large. Conventional methods typically use an (analytical or empirical) model of the task. Its not useful to train neural network on problems for which the solution can easily be implemented in an algorithm. while for other problems the solution should not use neural networks. The adaptability of the neural network will then be used to adapt the implemented solution whenever the problems changes • Its useful to apply Neural Networks to problems for which only too complicated algorithms can be derived.2 Ways of Implementing a Solution to a Specific Problem Useful Functions of the Neural Network Useful Function to be performed by the Neural network can be subdivided into few categories. In a training phase.6 Overview of the well known ANN Models In 1943 McCullah and Pitts discussed for the first time the role of mathematical logic in neural activity. 2. The ways of implementing the solution to specific problems can be divided as Problem Problem Level Solution Level Algorithm Neural Network Implementation Level Software hardware Figure 2. It was then the McCulloh_pitts neuron was first described. • It is useful to apply Neural Networks for the problems that change over the time.Application of Artificial Intelligence techniques in Power Systems useful. which are distinguished by the nature of the problem • Its useful to apply the neural networks on problems for which no direct algorithmic solutions exists but for which problem examples of the desired responses are availed. Neural Network can also learn these simple algorithms but neural implementation is generally larger and less accurate than the direct algorithmic implementation of the solution.

is the most widely studied. but the least applied model of all ANNs.1 Perceptron The perceptron is a single layer adaptive feedforward network of threshold logic Units.Application of Artificial Intelligence techniques in Power Systems neuron has fixed threshold. incidentally. The main types of the Neural networks are 2. It was found that the single layer perceptron works well for problems. It forms the basis of most of the further advances made in this field. has identical weights of excitatory synapses and the inhibitory synapses are absolute in nature. Single layer perpectron. which are linearly separable. Minkey and Papert in 1969 studied perceptrons intensively. Rumelharl proposed a multilayer perceptron with an error back propagation learning algorithm using a differential sigmoid activation function to facilitate learning rather than Special Study Report Page 10 of 56 . which are non-separable. ANN MODELS FEED BACK FEED FORWARD CONSTRUCTED TRAINED LINEAR NON LINEAR UNSUPERVISED SUPERVISED HOPFIELD ADAPTIVE RESONANCE KOHONON BACK PROPOGATION Figure 2. Block in 1964. Rosenblatt in 1958 invented perceptron. An Overview of main types of ANN models are as in figure. a pioneer in the field of design and development of digital computers made comparisons between the computers and the brain in 1962. Neumann. and gave the famous learning rule named after him. but fails to solve even simple problems. This is because they lacked an internal representation of stimuli.6. which possess some learning capability.3 Over View of the Main ANN models Hebb in 1949 introduced the fundamental concepts of learning in his classical text Organizational Behavior. which was proposed as a model for the organization of neural activity in the brain.

Training in this type of Neural nets are based on a limited number of training samples and it possess good generalization capability. X1 X2 X3 Input Layer Hidden Layer U1 U2 Un Output Layer Figure 2. Unidirectional connections exist from the input layer to the hidden layer and from the hidden layer to the output. 2.6. Varying the weights can change the input mapping. Backpropagation Special Study Report Page 11 of 56 .2 Multilayer Feedforward Neural Network In the feedforward neural network all the connections are unidirectional in a feedforward way. It consists of input layer of input variable. Therefore a multilayer perceptron possess a better learning capability.Application of Artificial Intelligence techniques in Power Systems using a threshold logic units or linear functions for activation. output layer of output variable and at least one hidden layer of hidden neuron.4 Three Layer Feedforward Neural Network 2. They are used as representational models trained using a learning rule based on set of Input / output data. Further progress was made with Amari in 1967 propounding the gradient-descent rule and designing of Backpropogation learning algorithm by Werbos in 1974.3 Backpropagation Networks It was demonstrated that the ANNs with hidden nodes and nonlinear activation's are able to simulate non-linear and linearly non-separable functions effectively.6. which was utilized in the multilayer perceptron model. The main learning rule used is the popular Back propagation algorithm (also known as a generalized Delta Rule). It has been proved that they are Universal Approximators. A multilayer perceptron is the typical example of feedforward neural network. The output variables are real-valued functions of input variables and weights. Major application of feedforward neural network is in large-scale systems that contain a large number of variable and complex systems where little analytical knowledge is available. There is no connection between any neurons in the same layer.

They classify an exemplar by calculating the Hamming distance for each class and selecting that one with the minimum Special Study Report Page 12 of 56 .4 Hopfield Network Hopfield Network invented by John Hopfield in 1982. and then the error is fed to backward to change the weights in the previous layer and this goes on for all the proceeding layers. Next. Stable configurations achieve a permanent state after a finite number of changes. In the Back Propagation Algorithms. a three-layer Back Propagation network can encode any arbitrary input or output relationship. A major limitation of Hopfield network is that not more than 0. the network begins with a random set of weights. and the squared difference between these two vectors determines the system error. They give better results when the input is perfectly represented as a string of binary bits. Backpropagation algorithm employs gradient descent search in weight space over the error surface to find the point resulting in minimum error. and the network converges to some of the other exemplars.5 Hamming Sets Hamming sets are similar to Hopfield networks. has lateral and recurrent connections. It is said to be unstable if it keeps on oscillating from one state to another. An input vector is presented and fed forward through the network.Application of Artificial Intelligence techniques in Power Systems networks are essentially multilayer perceptron networks. the output of a neuron are fed back to itself and intra-layer connections are present. 2. Hopfield network is used as associative memories. Training of the weights is carried out by Generalized delta rule (GDR) also known as Backpropagation algorithm (BPA). Here an exemplar is said to be suitable if it applies at time zero. Each node of the network is McCulloch.6. The state of Hopfield network is the set of stable states of all its neurons. This error is used to readjust weights between the last two layers and feedback to the penultimate layer over the weights connecting it with output layer. in which the network weights are adjusted in the direction of decreasing error. They can also be used to solve optimization problems. which are nonlinear. the calculated output is compared to the measured output data. The difference is that while perceptron uses hard-limiting threshold functions. Secondly it has got exemplar patterns. The implementation of Backpropagation algorithm. that is. Backpropagation network uses sigmoid functions. thus involves a forward pass through the layers to estimate the error at the output.15 N numbers of patterns can be stored on a network. and non-decreasing in nature. and the output is calculated by using this initial weighted matrix. It has been demonstrated that if a sufficient number of hidden neurons are present. In the learning phase of Backpropagation network a pattern is presented at the inputs and weights are assigned arbitrary small values. 2. The accumulated error for all the input / output pairs is defined as the Euclidean distance in the weight space.Pits neuron as used in perceptron. which the network attempts to minimize. Minimization is accomplished via the gradient descent approach. N being the number of needs in it. The learning is unsupervised and takes place offline. The corresponding actual and target outputs are compared and error is computed.6.

m being the number of classes stored on the network. which do not match the corresponding exemplar bias.k) Out( I. Special Study Report Page 13 of 56 . The topology of the network is similar to Hopfield Network. The hidden layer is the Kohonen layer and the output layer is Grossberg outside layer. It implements a clustering algorithm.6. Such ANN. Input Layer Hidden Layer (K-1) Layer In ( j.6 Adaptive Resonant Theory The binary Adaptive resonance theory (ART-1) introduced by Carpentar and Grossberg in 1968 is a two layer nearest neighbor classifier and trained without supervision which can be used only for binary inputs. Onelayer is the inputlayer.Application of Artificial Intelligence techniques in Power Systems Hamming distance. A simple representation of the counterpropagation network consists of three layer. and therefore their performance is better than or equal to that of Hopfield network. Input layer revises input from the input layer and has recurrent connection. Thus it has got feedback paradigm. which selects the first input as the exemplar for the first cluster. The input layer is a simple fan-out layer. Otherwise the example for a new cluster is performed. it is simple. The advantages of the CPN are that.j) Kth Layer Output Layer Figure 2. They also require less number of nodes than Hopfield network. This process is iterated for all inputs. The Hamming distance is the number of bits in the inputs. The counter propagation networks (CPN) have been recently used because of various advantages offered. having m nodes.5 Back Propagation Algorithm / Network 2. implements optimum minimum error classifier when bit errors are random and independent. The next input compares to the first cluster exemplar and clustered with it if the distance is less than a threshold.

Output Units multiplies pattern activation by a weight. The TDNN can be used for prediction problems. Xi Xj Xp Input Units Pattern Units Wj Wi Wn +1 W Bias Bias Output Units Figure 2.6. It’s started from Bashkriov and Aizerman at which time the networks are referred to as the method of potential functions. In both the cases the localized basic functions falls of rapidly to the distance between the centers of the basic function as the input gets large. The time-delays are introduced into the synaptic structure of the network and their values are adjusted during the training phase. Training in RBF consists of iteratively adapting the parameters of the network until the output approach the desired output over the whole range of training patterns. Classification of new patterns is done in much the same way in RBFs as in PNNs. It functions as a look-up table capable of generalization The Time-delay neural network (TDNN) is non-recurrent dynamic neural network which copes with time alignment by explicitly delaying the signal waveform by a fixed time span. and adds a bias.6 Typical RBF Network Special Study Report Page 14 of 56 .Application of Artificial Intelligence techniques in Power Systems easy to train and prevents a good statistical model of its input vector environment.7 Radial Basic Function (RBF) Neural networks based on localized basic functions and iterative function approximations are usually referred to as RBF networks. 2. RBF network is generally a regression network and so estimates the value of a customer variable. In simplest case the output of the network is a linear combination of all the basic function response. sums them.

8 Probabilistic Neural Network (PNN) and General Regression Neural Network (GRNN) PNN and GRNN are feedforward neural networks. PNN's are classifier version obtained when decision making is combined with a nonparametric estimator for probability density functions where as GRNN is a function approximated version. Adaptive PNN and GRNN are characterized by adapting separate widths for the basis function for each dimension. Since the sigmoidal functions have non-negligible values throughout all measurements space. much iteration are required to find a combination that has acceptable error in all parts of measurement space for which training data are available. The second types. a) Probabilistic Neural Network There are four variations for implementation of the pattern units in PNN network. The function that is to be approximated is defined to be a combination of these sigmoidal functions. Estimation of probability density functions and 2. Back propagation error method is training method applied to global basis function which is defined as nonlinear functions of the distance of the pattern vector from a hyperplane. The function to be approximated is approximated as a linear combination of these basis functions. Feedback may or may not be used in the training of networks. The training may be in terms of global -. and multivariable interpolation. Iterative functions approximation PNN's and GRNN's used for estimation of values of continuous variables are based on first type i. the topology of PNN is similar in structure to back propagation.It is possible to train a network of local basis functions in one pass through the data by straightforwardly applying the principles of statistics. future values. PNNs are ideal for exploration of new databases and preprocessing Special Study Report Page 15 of 56 . which is useful for estimating the values of continuous variables such as future position. based on iterative function approximation. differing primarily in that the sigmoidal activation function is replaces by an exponential activation function.Application of Artificial Intelligence techniques in Power Systems 2. They respond to the input pattern by processing the input data from one layer to the next with no feedback path. These networks use functions that have a maximum at some center location and fall off to zero as functions of distance from that center.e. estimation of probability density function. An obvious advantage of these networks is that training a network to have the proper response in one part of the measurement space does not disturb the trained response in other distant parts of the measurement space. In one variation. are usually referred to as Radial Basis Function (RBF) networks. Two main types of localized basis function networks are based on 1.or local basis functions.6. Due to this. Basic forms of PNN and GRNN are characterized by one pass learning and use of same width for the basic function for all dimension of the measurement space. These networks learn pattern statistics from a training set.

Special Study Report Page 16 of 56 . Regression is the least-mean-square estimation of the value of a variables based on examples. the GRNN provides smooth transitions from one observed value to another. If the variable to be estimated is future values. Thus GRNN can be used in these applications. Even with sparse data in a multidimensional measurement space. The remaining three implementations of the pattern units are optimized for implementation of the pattern units are optimized for implementation on multiply/accumulate digital signal processors or on special-purpose integer arithmetic processors. with relatively short test sets. b) General Regression Neural Network (GRNN) GRNN provides estimates of continuous variables and converges smoothly to the underlying (linear or nonlinear) regression surface. The term General Regression implies that being linear does not restrict the regression surface. because this use of the neural network typically requires frequent retraining and evaluation. plant or system. Like PNN. the GRNN is a predictor. If they are dependent variables related to input variables in a process.Application of Artificial Intelligence techniques in Power Systems techniques. GRNN features instant learning and a highly parallel structure.

Fuzzy systems are model free systems in which all things are matters of degree. Fuzzification offers superior expressive power.or non-statistical . Thus fuzzy systems reason with when we say. fuzzy systems fire each fuzzy rule in parallel. They systems "reason” with parallel associate's interference. to infer a conclusion or output. It matters the degree. Fuzzy logic systems provide an excellent framework to more completely and effectively model uncertainty and imprecision in human reasoning with the use of linguistic variables with membership functions. and an improved capability to model complex problems at a low solution cost. Would U say " begin braking 14 feet from the cross walk " or shall we say “apply brakes pretty soon. to which an event occurs. and in a wider sense. When asked a question or given an input. These systems use an inferential approach oriented towards system analysis and decision support.1 Importance of Fuzzy Systems Fuzzy set theory derives from the fact that almost all-natural classes and concepts are fuzzy rather than crisp in nature. Fuzzy programming admits degrees. They estimate sampled functions from input to output. 3. greater generality. and is propagated in every day’s life. the possibility of fuzzy event emerges. a coin toss) the outcome is a matter of chance . It is an extension of multi valued logic. Fuzzy systems store benefits of fuzzy associates or common sense "rules". not whether it occurs or occurs in random to what degree it occurs is fuzzy. They may use linguistic or numeric samples for example they may use HEAVY. Unlike fuzziness the probability dissipates with increasing information. LONGER or number (relative) for the degree of fuzziveness.Application of Artificial Intelligence techniques in Power Systems CHAPTER 3 FUZZY LOGIC AND FUZZY SYSTEMS 3. “fuzzy" or multivalued sets.g. is used. but to a different degree. Imprecision in data and information gathered from and about our environment is either statistical (e. Special Study Report Page 17 of 56 . We will say the latter and so the natural language is one example of ways vagueness arises. Fuzzy interpretations of data are a natural and intuitively plausible way to formulate and solve various problems in pattern recognition. used anonymously with Fuzzy set theory. "there is 20 percent chance of light rain tomorrow" .involves compound uncertainties. Fuzziness describes event ambiguity.This latter type of uncertainty is called fuzziness. Whether an ambiguous event occurs . instead of bivalent propositions.2 Basic Concepts Suppose your are approaching a red light and must advise a driving student when to apply brakes. Fuzzy logic is a logical system for formalization of approximate reasoning.

mB F (x). It is not necessary that either u ∈ F or u (F) exist. Special Study Report Page 18 of 56 .4 Classical Operations of Fuzzy Sets Zadeh [LAZ 65] defined classical operations for fuzzy sets Let f (X) = all fuzzy subsets of X (that is.1] instead from the two-element set {0. the rule is described as follows. A fuzzy rule is composed of a premise and a conclusion. given as a capital letter here X. In the fuzzy sets theory the generalization of the membership properties are as follows. defined on a referential called the universe of discourse. either u (C) or U (C) occurs.1). m f (X)  The fuzzy sets mA. function __ theoretic operations) Definition: A is a subset of B (A ⊆ B) if and only if ∀X ∈ X: (⊂) Containment A ⊂ B  m A (x) ≤ m B (x) The other operations are ∀X ∈ X: (~) Compliment mA (x) = 1-mA (x) ∀X ∈ X: (∩) Intersection m A ∩B (x) = min {mA (x). in order to distinguish them from fuzzy sets. If <x is A> then < conclusion>. In fuzzy set theory this property is generalized. The classical structure of a rule is “If < premise> then <conclusion>” When the premise is an elementary fuzzy proposal. For any crisp set C it is possible to define a characteristic function µC: U [0. The set that is defined on the basis of such an extended membership function is called as fuzzy set. Basic concept of fuzzy logic's is fuzzy " If then Rule " or Fuzzy Rule. mB (x)} ∀X ∈ X: (∪) Union mA∪B (x) = min {mA (x). then for any element of u of U.3 Fuzzy Sets and Rules In fuzzy set theory ' normal 'sets are called crisp sets. mB(x)} 3. generally real. The fuzzy rules are Definition: Two fuzzy sets are equal (A = B) if and only if ∀X ∈ X: (=) Equality A = B  m A (x) = m B (x) (∀X where x: pointwise. The x is a variable.Application of Artificial Intelligence techniques in Power Systems 3. Fuzzy rules are elementary or composed proposals. therefore in a fuzzy set F. A is a linguistic term. Let C be a crisp set defined on the universe U. They result from a conjunction between elementary fuzzy proposals. taken in a set of terms noted as TX.5 Membership Function and Membership Values m: X | (0.1}. 3.

Application of Artificial Intelligence techniques in Power Systems Membership function is the basis idea in fuzzy set theory. v) /(u. Its definition is Let U and V be uncountable (continuous) universe and µR : U X V R= [ 0. v ) 3.7 Properties of Fuzzy Sets Let A and B be the fuzzy sets. v) /(u.6 Fuzzy Relations The fuzzy relation can be considered as a fuzzy set of tuples. and let R be a fuzzy relation defined on XxY. then UxV ∫ µR (u. (u) ∈ {0. This is denoted by S (A). v ) /( u .1}. defined respectively on the universes X and Y.Inf (S (A)) where Sup and Inf denote the mathematical Special Study Report Page 19 of 56 .1] . The width of the convex fuzzy set A with support set S (A) is defined by Width (A) which is equal to Sup (S (A)) . Fuzziness represents similarities of objects to imprecisely defined properties and probabilities which convey information about value frequencies. then R= UxV ∑ µR (u. Its values measure degrees to which objects satisfy imprecisely defined properties. Therefore in fuzzy control theory the term width of a fuzzy set is used additionally to the term support. v) This is a binary fuzzy relation on U x V. every element u of U has a membership degree µF determined by the set of tuples F = {(u. formally defined as S (A) = {u ∈X | µA (u) >0} When one deals with convex fuzzy sets as it is the case in fuzzy control theory the support of a fuzzy set is an interval. F is completely So. v The integral symbol denoted the set of all tuples on U x V denoted by µ R ( u . µF (u)) | u ∈ U} 3. That means each tuples has membership degree between 0 and 1. The member ship function µF of the fuzzy set F is a function µF: U [0.1]. If U and V are controllable (discrete) universes. The support of fuzzy set A is the crisp that contains all element of A with non-zero membership degree.

then this point is called the peak value of A. Fuzzy systems learn associative rules to estimate functions or control systems through unknown probability (sub set hood) function p (x). An unsupervised learning systems process each sample X but does not “know " that X belongs to class Di and not to Special Study Report Page 20 of 56 . But one cannot learn without changing. y). how the system encodes information.1] and L be the set of all fuzzy values.8 Fuzzy Truth Value A fuzzy truth-value is defined to be a fuzzy set on the closed interval V = [0. Then solutions points (X.1]} The same can be graphically written as follows -1 0 0 1 0 a b 1 0 (a) Numerical Truth Values (b) Interval Truth Values (c) Fuzzy Truth Values Figure 3. f (x)) should reside in highprobability regions of the input/ output product space X x Y. A is a fuzzy truth-value if and only if A is a fuzzy set on [0. The nucleous of a fuzzy set A is defined by Nucleus ( A) = { µ ∈X |µ A ( u) = 1 } If there is only one point with membership degree equal to 1.Application of Artificial Intelligence techniques in Power Systems operations supremum and infimum. If the support set S (A) is bounded as is usual in fuzzy control. a few of which the neural or fuzzy systems sample. When a neural or fuzzy system estimates a function f: X Y.9 Learning in Fuzzy Systems 1 Generally learning can be well or can be bad. and we cannot change without learning. The probability density function p (x) describes a distribution of vector patterns or signals X.1 Truth Values in Fuzzy Logic 3. Max and Min can replace Sup and Inf. it in effect estimates the joint probability density P (x. They determine how the synaptic web process unfolds in time as the system samples new information. Learning laws describe the synaptic dynamical system.1] as follows. This is one way neural network compute with dynamical systems. that is L = {a | a is fuzzy set on [0. 3.

Numerical PR is characterized in four major areas as shown in the figure 3. proximity's). pattern vectors) and relational data (similarities. the successful Pattern recognition is developed by iteratively revisiting each of the four modules until the system satisfies a given set of performance requirements and economic constraints.. Generally two data structures are used in numerical PR systems. Pattern Recognition is the search for structure in data.…. 3. 3. Object data are represented in the sequel as X= {x1. supply algorithm to perform fuzzy computations those will generally lead to fuzzy outputs. Object data vectors (feature vectors. 3) FLC inputs are normally observations of technological systems and their outputs control statements. that is. 2) The designer formulates rules of FLC systems. (Defuzzification). xn} a set of n feature vectors in feature space Rp . Main approach to PR is the structural (Synatic) approach. which is. 4) Design the computational unit. 2) Define the condition interface. utilizing neuristic knowledge. Supervised learning use class-membership information and unsupervised learning used unlabelled samples. 5) Determine rules according to which fuzzy control statements can be transformed into crisp control actions. fixed the way in which observations of the process are expressed as fuzzy sets.2. Design Problems of FLC are 1) Define Input and Output variables that are determined which status of the process shall be observed and which control actions are to be considered. 3) Design the rule base. This branch of PR is the less well developed in terms of fuzzy and neural models. that is. The difference between expert systems and the fuzzy logic controllers (FLC) are 1) FLC models are rule-based systems.11 Pattern Recognition in Fuzzy Systems Pattern Recognition is a fixed concerned with machine recognition of meaningful regularities in noisy or complex environments. x3. fix the way in which observations of the process are expressed as fuzzy sets.Application of Artificial Intelligence techniques in Power Systems class Dj.x2. In practice.10 Fuzzy Logic Controllers (FLC) Fuzzy systems. have been employed very effectively as controllers popularly known as Intelligence Control. Special Study Report Page 21 of 56 . the jth object observed in the process has vector Xj as its numerical representation: Xjk is the kth characteristic associated with the object j.

Fuzzy models for PR associated with relational data are fairly developed now a day. For example every metric d or Rp X Rp produces a dis-similarity relation matrix R (X: d) as in figure. Where we take ρ = d. we have access to a set of n2 numerical relationships say {rjk} between pairs of objects Oj and Ok.1} then it is hard (or clip) binary relation function. Special Study Report Page 22 of 56 . rjk represents the extent to which objects j and k are related in the sense of some binary relation ρ. If every rjk is in {0.Application of Artificial Intelligence techniques in Power Systems Humans Process Description Feature Nomination X= Numerical Object Data D : Xx X R R= Pair-Relation Data Design Data Test Data Sensors Feature Analysis Preprocessing Extraction 2-D Display Cluster Analysis Classifier Design Classification Estimation Prediction Control Exploration Validity Figure 3. That is.12 Relational Data It may happen that. ok)). Its is convenient to array the relational values as an n X n matrix R = (rjk) = (ρ (oj.2 Characterization of Pattern Recognition 3. Many functions convert X x X to relational data. If 0<rjk<1 for any j and k we call R as fuzzy relation. instead of an object data set X.

It begins with a few crude values of thumb that describes the process.3 Adaptive Fuzzy Controllers Basic Model of Adaptive Fuzzy Controller is as shown. Static Fuzzy logic controllers (FLC) have already been widely used in engineering applications. Neural parameter estimators embed directly in an overall fuzzy architecture. The emphasis in such problems is to approximate multiple pattern classes in a joint input output space. Neural networks “blindly " generate and refine fuzzy rules from training data. Fuzzy Logic (FL) has been used in areas like pattern recognition problems and processing inexact ideas. Adaptive controllers are important for good performance in non-stationary applications. Process Model Performance Measure Identifier Decision Maker Process Model Based Controller Figure 3. Adaptive fuzzy systems learn to control complex process very much as we do.13 Adaptivity Features and Adaptive Controllers One of the main topics of high interest to researchers in fuzzy logic (FL) field is the development of automotive-data-driven adaptive controllers. Expert may give them the rules or may extract the rules from the observed expert behavior. Successive experience refined the rules and usually improves performances.Application of Artificial Intelligence techniques in Power Systems 3. Special Study Report Page 23 of 56 .

Vision. The major advantage of the neural networks is its self-learning capability. and security assessment. The development of the input and output data is done several thousand times. Most of the applications make use of the conventional multilayer Perception (MLP) model based on back propagation algorithm. The network function is determined largely by the connections between elements. power system stabilizer design. unit commitment. ANNs are composed of many simple elements operating in parallel. and the field is still developing rapidly. On the field of AI. control systems and EMS. First. They have been trained to perform complex functions in various fields of application including Pattern Recognition. building on experience to provide better answers.Application of Artificial Intelligence techniques in Power Systems CHAPTER 4 APPLICATION OF ARTIFICIAL NEURAL NETWORKS IN POWER SYSTEMS 4. and then it produces a set of output values. In recent years. ANNs have attracted much attention due to their computational speed and robustness. the network is presented with a set of correct input and output values. Therefore they reaches the desired solution efficiently. They can match stored examples against a new one. many interesting applications of ANNs have been reported in the power system areas like load forecasting. A major advantage of the ANN approach is that the domain knowledge is distributed in manner. Economic load Dispatch and fault analysis. Classification. They have become an alternative to modeling of physical systems such as synchronous machine and transmission line. As a branch of Artificial Intelligence. Special Study Report Page 24 of 56 . ANNs take problem-solving one step further. Second the network is presented with only the input data. Such networks are characterized by exceptional pattern recognition and learning capabilities. Then it adjusts the connection strength among the internal network nodes until proper transformation is learned. Neural networks are based on neurophysical models of human brain cells and their interconnection. Absence of full information is not a big as a problem in ANNs as it is in the other methodologies. After proper number of learning cycles or iterations the network will be able to produce accurate output data from input data similar to those used for learning. Identification. ANN computing shows great potential in solving difficult data-interpreting tasks. multilayer perception model suffers from slow learning rate and the need to guess the number of hidden layers and neurons in each hidden layer. Speech.1 Introduction on ANN Application ANNs can play a richly significant potential role in electric power systems. The field of ANNs has a history of nearly five decades but has found solid application only in the past ten years. However. Many improvements are suggested over the conventional MLP to overcome these advantages.

El-Sherbiny et al [ShSaI 96] introduce a novel Power System Stabilizer (PSS) controller based on a multilayer feedforward artificial neural network (ANN). Then the control law is implemented using the inverse dynamics of the controlled plant. H. with a nonlinear transigmoid activation function. The input layer has four nodes.K. inverse dynamics neural network (IDNN). Srinivas Pilutla and Ali keyhani in [SA 97] illustrated the use of the modular neural networks for power system stabilizer modeling. M. an input layer. These networks learn different aspects of a problem by partitioning the data space into several different regions and are less susceptible to interference than backpropogations networks. Input to the ANN was the generator real power output P and the Power Factor. Here an output feedback control law is driven with some conditions satisfied. S. Y. The proposed ANN based PSS consists of three layers. on offline identifies the inverse dynamics of the controlled plant. When trained to respond differently to different operating conditions. modular neural network architectures have been used for problems in system identification and control. Backpropagation neural networks have recently been applied to problems in power system stabilizer modeling. a hidden layer and an output layer. An ANN. In recent years. Hsu and Chen [HC 91] proposed a fourlayer perceptron network for this purpose. namely. Offline simulations generated the training set for this ANN.2. M. Time domain solution with specified state disturbance for a synchronous machine connected to an infinite bus through an external transmission line are employed to prove the effectiveness of the proposed ANN based controller under a wide range of variations of the operating conditions and variety of exciter gains.H Hyun and J. The network consists of two input nodes.1 Power System Stabilizer Real time timing of PSS is a complex task. Lee [PHL 96] suggested enhancing the dynamic performance of power system. two hidden layer of four nodes each and two output nodes. The outputs of ANN were the PSS gain settings. To speed up the learning process an adaptation law was used to dynamically update the learning rate of the backpropagation. Another important application is the stable power system stabilizer based on inverse dynamics of the controlled system using an ANN. Park. The last layer (output layer) has one node whose activation function is transigmoid.Application of Artificial Intelligence techniques in Power Systems 4. these networks tend to produce interference between conflicting solutions. which guarantees the internal stability and robustness against the asymptotically stable external disturbances.2 Major Applications 4. The best number of the nodes for the hidden layer has been found by trial and error to be seven. A feature of the proposed controller is that the ANN parameters can be adapted online in real time according to generator loading conditions. Special Study Report Page 25 of 56 .

This ANN was used for a one-day ahead load forecasting. ANN schemes using perceptron networks and self-organizing feature maps have been successful in short-term as well as long-term load forecasting with impressive accuracy. Hsu et al [HY 91] demonstrated the suitability of combining self-organizing feature maps and multilayer perceptron for short-term load forecasting. An average percent relative error of two % was achieved.Application of Artificial Intelligence techniques in Power Systems Output Layer Gatting Layer Gating Network Local Expert I Local Expert L Fully Connected Input Layer * Figure 4. * Ref [SA 97] Special Study Report Page 26 of 56 . The availability of historical load data on the utility databases makes this area highly suitable for ANN implementation.1 Modular Neural Network FeedForward Architecture 4. for the winter.2 Load Forecasting Load forecasting is perhaps the most important SCADA task and also one of the most popular areas for ANN implementation. Lee et al [LCP 90] used a multi layer perceptron for short-term load forecasting. summer and fall seasons. Park et al [PEM 91] employed a similar approach to compare the performance of multi layer perceptron with a utility’s numerical forecasting methods.2. spring.

However. To obtain the hourly load pattern for a day. using current amplitude and phase angle distribution patterns. at el [VB98]. the hourly load patterns of several days in the past. Special Study Report Page 27 of 56 . three layer feedforward Quasi Optimal neural network for the short term Load forecasting [MCS97] and the window based forecasting procedure using combined Supervised and Unsupervised learning concept [DRSP 95]. However. Y.S. at present the ES implementations outnumber the ANN implementations. D. The ANNs were trained using backpropagation. Many applications for the various fault diagnosis problems have been reported in the literature. Recurrent Neural networks are members of a class of neural network models exhibiting inherent dynamic behavior. which is realized by means of a multi layer perceptron based on the back propagation learning algorithm already used for the ANN implementation. thus to obtain clusters of the similar load profiles. The second stage consists in an actualization process of the information deduced from the previous day type identification. The first stage provides some identification criteria of the characteristics of the days through the classification of historical hourly loads. Prudenzi at el [LPSCO 96] illustrated a new ANN based procedure (SOM + ANNI) in order to enhance the forecasting accuracy in the analysis of the load forecasting. Human operators perform this activity that gives a meaning of the load classes. The classification is performed by means of a Kohenon’s SOM. Vermaak. Success of applying a class of recurrent neural network in short term load forecasting was tested by J.3 Fault Diagnosis ANN’s has recently invaded fault diagnosis. The procedure provides the combined approach (unsupervised + supervised) structured in three subsequent stages.Application of Artificial Intelligence techniques in Power Systems The self-organizing feature maps were used to identify the day types from historical data. The explanatory abilities of ESs and their more powerful user interface make them a more attractive alternative. Here the feedforward neural networks (including those used for the recurrent network training) employed a single hidden layer. R. still there are certain areas. and are still open to ANN implementation.Tzeng [CTH 96] Using SCADA. a multilayer perceptron was used. To predict the daily load. were averaged. Kanoh et al [HMK 88] proposed a cascade structure of three three-layer perceptron networks for the identification of a faulted transmission section. which has been a traditional area for ES (expert system) implementation. 4. and were trained in batch mode according to the error backpropagation algorithm. The other main works in the area of load forecasting are substation load forecasting C. The recurrent network parameters were obtained by training a feedforward network to learn the mapping. The first and the second ANN in the cascade structure identify the candidate’s one and two for fault selection. which are of the same day type. The third stage.2. using the conjugate gradient descent optimization. The most general of these is the fully connected recurrent neural network. Srinivasan et al [DLC 94] for a short term forecaster using multilayer neural network. which require a quick response. Chen. performing the proper forecasting task.M.Lamedia A.

and a current amplitude distribution pattern. P1 DAY(I-1) P24 Cluster Codes Relevent To Days (I-2). Computer simulations using the EMTP generated the training set.Application of Artificial Intelligence techniques in Power Systems The third ANN obtains the final fault location using the above candidates one and two. training the ANN with these data and testing the ANN on new patterns. However. Results of this approach indicates that this method can achieve 98.2 Unsupervised/Supervised Procedure Adopted for Load Forecasting Ebron et al [EL 90] used a three-layer perceptron network to detect high impedance faults on distribution feeders. the ANN caused a false alarm in seventeen cases as mentioned. P1 DAY I ………………………. From the results obtained ANN classified ten of these cases correctly.i DAY TYPE CLASSIFICATON Kohonen's SOM Learning EXTRAPOLATION AND REPRODUCTION OF CLASSIFICATION CRITERIA P1 ……….4 percentage accuracy even when the measured values differed by thirty percentage from the EMTP as mentioned above. * [LPSCO 96] Special Study Report Page 28 of 56 . processed feeder line currents.. P24 FORECASTING Supervised Back-propagation Learning ………… P1 DAY(I-2) P24 …………. Their approach consisted of three parts: collecting sets of sampled. P24 CALENDER TIME CHARACTERISTICS OF FUTURE DAYS * Figure 4.(I-1).

Application of Artificial Intelligence techniques in Power Systems ANNs were also successful in incipient fault detection of induction motors [CY90]. This approach used two ANNs. the neural network-based approach presented here fulfills the scalability and dynamic adaptability requirements of the application. The way in which the neural system is conceived allows full scalability to realsize power systems. 2. Chow and Yee [CY91] used multilayer perceptron networks for incipient fault detection in single. In contrast with the monolithically diagnosis systems. 1. An incipient fault detector ANN to detect faults based on data collected from the motor.Rodriguez at el [RRMLMP 96] presented a modular and neural network-based solution to power systems alarm handling and fault diagnosis described it overcomes the limitations of ‘toy’ alternatives constrained to small and fixed-topology electrical networks. C. Mapping the power grid onto a set of interconnected modules that model the functional behavior of electrical equipment provides the flexibility and speed demanded by the problem.phase squirrel cage induction motors. A disturbance and noise filter ANN to filter out the transient measurements while retaining the steady-state measurements. 1 PREPROCESSING 2 DISTURBANCE DETECTION AND CLASSIFICATION FAULT DIAGNOSIS 3 HYPOTHESIS GENERATION 4 HYPOTHESIS JUSTIFICATION * Figure 4.3 Fault Diagnosis process * [RMAMP 96] Special Study Report Page 29 of 56 .

Transmission losses are also considered in the constraint function. The goal of security assessment is to supply the operating state so that suitable preventive actions can be undertaken. Sobajic and Pao [PS89] synthesized one of the crucial parameters of the system. a fast dynamic security evaluation for a specified wind perturbation is performed using an ANN. Security assessment has been at the forefront of ANN applications from the beginning. from Taylor Series Expansion of the transient energy function in the state space near a certain class of unstable equilibrium point. A three-layer perceptron network with twelve input nodes. Security constrained optimal rescheduling of real power using Hopfield network was analyzed by Soumen Ghosh et al [SC 96]. For a given operating point. labeled with the corresponding CCT values. Special Study Report Page 30 of 56 . The CCT parameters were obtained by numerical integration of the post-disturbance system equations. The proposed method is based on solution of a set of differential equations obtained from transformation of an energy function. D. the worst impending contingency. Results from this work are compared with the results from a method based on dual linear programming formulation of the optimal corrective rescheduling. If insecurity is detected. Inclusion of inequality constraints on active flow limits and equality constraint on real power generation and load balance assures a solution representing a secure system. In this paper a new method for security-constrained corrective rescheduling of real power using the Hopfield network is presented. new alternative stable operating points are suggested. Here a local approximation of the stability boundary is made by tangent hyper surfaces.2. using a hybrid ANN-optimization approach that checks several feasible possibilities.Application of Artificial Intelligence techniques in Power Systems 4. Neural networks are used to determine the unknown coefficients of the hypersurfaces independently of operating conditions. which are developed. The ANN was trained on the results of off-line stability analysis. The training set was a twelve dimensional pattern set. The minimum deviations in real power generations and loads at buses are combined to form the objective function for optimization. without any abnormalities. The CCT parameters output by the ANN matched closely with the actual values using a three-layer perceptron network to assess the dynamic security of the power systems. six hidden-layer nodes and one output node was employed for this purpose.4 Security Assessment Security of a power system is the ability to sustain.N Fidalgo et al [FPV 96] described the ANN based approach for the definition of preventive control strategies of autonomous power systems with a large renewable power penetration. resulting from changes in power produced by diesel and wind generators and other combinations of diesel units in operation.J Sobajic and Pao et al [DSP 94] by a direct method for the multimachine systems. The transient security assessment analysis is done by M. In one of the early approaches. J. the critical clearing time (CCT).Djukanovic.

The results of the ANN based state estimation compared favorably with that of the Kalman filter. The ANN was trained using offline simulation data of a test system. Since a systems operational history is available in most utility databases. An adaptive linear combiner and a multilayer perceptron network were also used [KF 90] for state estimation.2. Bialasiewicz et al [BPW 89] showed that a multilayer perceptron network could be used as a state estimator in a model reference intelligent control system. the following empirical rule was proposed for calculating the number of hidden nodes in the perceptron network H = I log 2 N ± I Where ‘N’ is the number of training patterns. This ANN has a generalized structure that is independent of applications. In this work. The power system topological observability is dealt with [TM 89] using a three-layer perceptron network. the ANN was trained using several Kalman filter solutions for the power network. a huge number of possible contingencies are to be evaluated and ranked. However it is impossible to generate enough training sets to cover the entire range of power system operation. which tackles the power system state estimation problem. and H the number of hidden nodes.2. Hence a Hopfield network was proposed in [FKCRY 90] for contingency screening. OET2 overcomes the major shortcomings of the backpropagation learning rule and can also be very useful for other problems. An adaptive learning scheme was employed. The authors claimed that this empirical rule is valid for certain classes of sensor validation problems.5 State Estimation ANNs have been very successful in system identification. The learning rate of the backpropagation was updated dynamically to speed up the learning process.Application of Artificial Intelligence techniques in Power Systems 4. Power system network decomposition techniques are used to decrease the computational burden of the topology classifier training session. Conventional ranking methods suffer from masking and long computing time. parameter estimation and analysis. This paper used an optimization Special Study Report Page 31 of 56 . Eryurek et al [EU 90] proposed a three-layer perceptron network for sensor validation in a power plant.6 Contingency Screening To assess system security. ‘I’ the size of the input vector. it should be possible to group contingencies into various subclasses [FKCR 89]. A structured ANN was reported in [NA 90]. A P Alvas da Silva and V H Quintana [AQ 95] presented a paper on an ANN topology determination and a supervised learning algorithm for very large training sets using the Optimal Estimate Training 2(OET2). This ANN was trained using simulation results and back-propagation. 4. In this paper Fischl et al showed that a two-layer perceptron network could classify power system security status accurately under different loading and contingency conditions. Performance of this network was shown to be superior to that of a back propagation scheme. In this implementation.

Selection of input/output parameters for training. 4. accuracy and efficiency.8 Protection The application of ANN in this related field too is now days becoming important since the concept of online protection are widely accepted.2.Application of Artificial Intelligence techniques in Power Systems method to find the weights and thresholds of the ANN. The training process involves the following steps: 1.A. 2. The optimization method used linear programming techniques to maximize the probability of correct classification of contingencies. S. Adaptive relaying Special Study Report Page 32 of 56 . 3. the network is exposed to a set of patterns.7 Voltage Stability Assessment ANNs have been recently proposed as an alternative method for solving certain traditional problems in power systems where conventional techniques have not achieved the desired speed. Training of the neural network continues with the updates in weights in V and W. The technique followed in this paper is the backpropagation method. In the training process. each of which consists of an input vector X. In the ANN model for each loading condition various combination of control variables are generated by running many iterations of LP based reactive power optimization algorithm. This implementation classifies contingencies according to the number and type of limit violations. Two new indices – severity index and a margin index for line flow – are defined. L index has been popularly used for assessing voltage stability margin. until the error E reaches a predefined minimum value in a steepest descent manner. Khaparde. on the learning time and test performance of Multi Layer Perceptron (MLP) based ANN model. Warke at el [KWA 96] shows that ANN can be effectively used effectively to achieve adaptive relaying for the above-mentioned problem. in contrast to the learning method of the perceptron networks.2. Investigations are carried out on the influence of information encompassed in input vector and target output vector. Testing of the network with unknown set of data 4. N. The method has interesting applications in combining security monitoring and preventive control. S Gosh and B H Chowdhury [GC 96] modulated a three-layer perceptron artificial neural network with back propagation learning technique that is designed for line flow contingency ranking. Generation of training data. Settings of control variable influences the ANN input feature vectors differently. A regression-based correlation technique is used to select training parameters for the neural network. Normalization of training data 4. and the corresponding desired vector d. Only active power injection of slack bus and reactive power injection of all generator buses vary in input vectors of ANN2 for a given loading condition while variation in input vectors of ANN-1 is observed in most of the critical line flows.

9 Load Modeling The application of the ANN in load modeling is increasing for the past years. P Alves da Silva and C. The main draw back of the functional link net is that the required non-linear transformation can only be found by trial and error.2. However selection of the optimal number of hidden layers and the optimal number of hidden layers. The polynomial network is a nonparametric ANN model i. and the optimal number of neurons in each layer.e. A modified multilayered perceptron model employs an additional node in the input layer. which is based on the following ANN models: functional link net and polynomial network. The desired change in the quadrilateral relay characteristic is achieved by making appropriate changes in the thresholds and weights of the hidden layer neurons. Here the feedforward multilayer neural network was chosen for the study. During further studies and analysis different combinations of the following network training methods were chosen and tested in order to ensure that the model would be continuously refined 4. These additional input facilities changes in the relay characteristics. The network architecture proposed here is the Functional Polynomial Network. Special Study Report Page 33 of 56 . A. The other method used by Q. Accurate dynamic load models allow more precise calculations of power system controls and stability limits.H Song [XSJMW 96] illustrated an adaptive protection technique based on neural networks with special emphasis on analysis of the firstzone performance. Y. Xuan. Y. is still an open issue? The guidelines given for the number of the hidden neurons were adopted as a starting point. The impact of the clustering different load compositions is also investigated. so the philosophy of adaptive relaying must vary accordingly.Application of Artificial Intelligence techniques in Power Systems covers a large number of applications and the characteristics of relays vary widely. it does not require the architecture pre-specification. Ferreira et al [AFZL 97] detailed the performance of a nonparametric load model based on a new constructive artificial neural network (Functional Polynomial Network) (FPN) and it’s compared with the popular “ZIP” model.

Application of Artificial Intelligence techniques in Power Systems

CHAPTER 5 APPLICATI0N OF FUZZY LOGIC IN THE POWER SYSTEM 5.1 Introduction on Fuzzy logic applications Fuzzy logic applications are widely used in all parts of the power system planning, design and operations. The main important applications are 1. Stability Assessment / Enhancement 2. Power System Control 3. Fault Diagnosis 4. Security Assessment 5. Load Forecasting 6. Reactive Power Planning and Control 7. State Estimation 5.2 Major Applications 5.2.1 Reactive Power and Voltage Control The rapid growth in the power system coupled with variations in operating conditions leads to better management in voltage profile and reactive power. Reactive sources which are spread throughout the system should be controlled accurately based on the loading conditions (light load or peak load) to optimize and ensure the security of electric power transmission system. These controls are known as voltage/reactive power or voltage/VAR control. The aim of these controls is to reduce voltage deviations or minimum losses or enhancing voltage[ NU 98]. Main types of voltage/ VAR problems are 1. Planning of system reactive demands and control facilities as well as installation of reactive power control resources 2. The operation of existing voltage/VAR resources and control device. The online planning is much more cumbersome and important in the power system operation. This is because in a day to day operation of power system both under/over voltage occurs and VAR sources need to be adjusted to avoid high/low voltage problem. This can be termed as voltage/VAR scheduling and this is very important in the power system security. There are various algorithms employing linear and non-linear optimization technique used for voltage correction. These algorithms involve numerical computations and hence may not be curtailed and also the amount of controller movement needs to be minimized.

Special Study Report

Page 34 of 56

Application of Artificial Intelligence techniques in Power Systems

Fuzzy set theory has been applied off late for reactive power control with the purpose of improving the voltage profile of power system. Here the voltage deviation and controlling variables are translated into fuzzy set notations to formulate the relations between voltage deviation and controlling ability of controlling device. Main control variables are VAR compensators, transformer taps and generator excitation. A fuzzy rule system is formed to select these controllers, their movement and step size. The controllers are selected based on 1. Local controllability towards a bus having unacceptable voltage. 2. Overall controllability towards the buses having poor voltage profile. K. H. Abdul_Rehman / S. M. Shahidehpur et al [AS 93] presents a mathematical formulation for the optimal reactive power control problem using the fuzzy set theory. The objectives are to minimize real power losses and improving the voltage profile of the given system. Transmission losses are expressed in terms of voltage increments by relating the control variable, i.e. tap positions of transformers and reactive power injections of VAR sources, to the voltage increments in a modified Jacobian matrix. Main advantage of this method illustrated is that the specific formulation of this problem doesn’t require Jacobian Inversion of matrix and hence it will save computation time and memory space. The objective function and the constraints are modeled by the fuzzy sets. Linear membership functions of the fuzzy sets are defined and the fuzzy linear optimization problem is formulated. The solution space here is defined as the intersection of the fuzzy sets describing the constraints and the objective function. Each solution is characterized by a parameter that determines the degree of satisfaction with the solution. The optimal solution is the one with the maximum value for the satisfaction parameter. Multicase VAR planning problem involves the determination of an installation pattern of location and sizes of new compensators for multiple cases. The problem should basically cover the operating limits, complicated security and economic factors. a) Voltages and VAR controllers must be kept within their operating limits for the entire system under both normal and contingency cases. b) The expansion between cases should be coordinated to avoid excessive investment. c) The amount of compensation (by capacitor and reactors) must be descritized. In the area of the Multicase VAR planning R. A. Fernandus et al [FLBHW 83] proposed augmented Lagrangian type objective function and later augmented Lagrangian and generalized benders decomposition methods were applied [GPM 88] to treat both preventive and corrective controls of VAR planning. The drawbacks of traditional approaches were pin pointed by Hong and Liu et al [HL 92]. An expert system (VPES) (VAR planning Expert System) was introduced. It incorporated constraints resulting from considerations of the voltage collapse and able to handle both fixed and the variable cost and discrete device.

Special Study Report

Page 35 of 56

Application of Artificial Intelligence techniques in Power Systems

Fuzzy Set theory has also been applied to solve. Here an extended approach based on VPES is proposed to take fuzzy reasoning rules into account for solving Multicase VAR planning solution. Combination of individual information from each single case is performed by fuzzy relationship the center of gravity algorithm. Thus the coordination of multicase VAR planning is achieved. The other important area is the application of the reactive power compensation in distribution system .The aim is to achieve power and energy loss reduction, voltage regulation, and system capacity release. An approach using fuzzy dynamic programming to decide the optimal capacitor placement and size of compensating shunt capacitor for distribution systems with harmonic distortion is proposed by Hong Chan Chin et al [HC 95]. The problem is formulated as fuzzy dynamic programming of minimization of real power loss and capacitor cost under the constraints of voltage limits and total harmonic distortion. The algorithm proposed greatly reduces the effort of finding optimal location by any exhaustive search. The computational algorithm is narrated in the following steps as given in. 1. Perform the load flow program at the fundamental frequency to calculate the bus voltage. 2. Find the membership functions µP, µV, µH and µD for the fuzzy sets P, V, H and D. 3. Identify the optimal location of shunt capacitor at the bus with the lowest membership Value µp(K) ( bus K ) 4. Try the capacitor placement at bus K with various discrete sizes. Select the optimalsize QC that will result in lowest cost function without violating the constraints. 5. Install the capacitor QC at the bus K and simulate the load flow to calculate the new bus voltage violation. Ching-Tzong Su & Chien_tung Lin [SL 95] illustrated voltage profile enhancement for Power Systems using fuzzy control approach. The voltage violations are transformed to fuzzy set notations to formulate the relation between the voltage violation level and the controlling ability of controlling devices. A feasible solution set is first attained using the min-operation of fuzzy sets, and then the optimal solution is fast determined employing the max- operation. The membership function of the bus voltage violations is represented as in the following figure. Here ∆Vi represents the voltage violation level of bus I, and u∆Vi represents the membership function of ∆Vi The maximum deviation of the bus voltage is given by


Cij min 0 Cij max *Figure 5.1 The membership function of controlling ability of controlling devices * [SL 95]
Special Study Report Page 36 of 56

line Impedance. Bus voltage limits. controlling margin) Perform base Case Load Flow Find the sensitivity coefficient Calculate the Controlling Ability Find the membership value of bus voltage violation level and controlling ability Evaluate the Optimal control Solution Modify the value of the Control Variables Check Voltage level has enhanced to the desired level YES Perform the load Flow and output the Results NO * Figure 5. bus power.01 ∆VImax ∆VImin = Vimin .2 The membership function of Voltage violation Level The computational procudure of the above algorithm was repersented as Input data (Including network configuration.01 0 0.Application of Artificial Intelligence techniques in Power Systems U∆Vi ∆V i ∆VImin -0.3 Computational Procedure for the solution for Voltage Profile Enhancement * [SL 95] Special Study Report Page 37 of 56 .ViNorm *Figure 5.

M. Hence the on-line fuzzy control of the Super-conducting Magnetic Energy Storage (SMES) and the Static Synchronous Series Compensator (SSSC) are suggested. This stabilizer used a fuzzy relation matrix to produce the output based on the fuzzy inputs. In another research transient stability limit in power system transmission lines using the fuzzy control of FACTS Devices was studied. Finally the effect of the control loop time delay on the performance of the controller is presented. The validity of the suggested control strategies is confirmed by simulation tests. which may be constrained by a transient stability limit. Special Study Report Page 38 of 56 . The simulation results show that by the use of the proposed method. Ehsan in et al [SEHFH 98] investigate the application of FACTS devices to increase the maximum loadability of the transmission lines.3 Generator Operation and Control The major application lies in the control of excitation system of the Synchronous Generator. A number of new control theories have been introduced to design high performance excitation controllers.2. Sadehzadeh and M. Application of the fuzzy set theory in transient stability evaluation was first reported by Soulfis et al [SMP 89]. Synchronous Generator excitation control is one of the most important measures to enhance power system stability and to guarantee the quality of the electrical power it provides. [MOG 91]. the line power transfer can be increased via the improvement of the transient stability limit. classified as belonging to one of the six possible states were represented using the fuzzy membership values in fuzzy Pattern recognition (PR) systems. 5. resulting in a simple design.2. The practical implementation and experimental results of this stabilizer using a digital signal processor were reported in [HM 93]. speed deviation and acceleration. Among them the linear optimal control theory [JHA 89]. The system operating states. Hassan et al reported another successful application of a fuzzy logic stabilizer for improving the stability of synchronous machines.2 Transient Stability The most active area of the fuzzy system research in the power systems has been stability assessment and enhancement. configuration or loading condition [AV 89]. Only local measurements from each machine were used for this stabilizer.Application of Artificial Intelligence techniques in Power Systems 5. The stable performances of the synchronous machines under all anticipated conditions of system transients are essential for ensuring overall system stability. The developed method is applicable for any power system irrespective of its size. The fuzzy rule bases are defined and explained. S. An application of Fuzzy set theory for design of stabilizer to improve the dynamic performance of a multimachine power system was first proposed by Hsu and Cheng [HC 90]. the adaptive control theory [CCM 86] the fuzzy logic control theory [HC 90] and the nonlinear control theory [LS 89] are the most commonly used ones.

Here a design technique for the new hydro power plant controller using fuzzy set theory and ANN was developed. can offer better damping effects for generator oscillations over a wider range of operating conditions than conventional regulators. Djukanovic and M. They are simple in structure and relatively easy to realize. This method is time consuming and does not guarantee an optimal controller. the guide vane and the runner blade positions. Mathematical models of the control systems are not required. which determines the shortfall or surplus generation unit that has to be corrected. In this controller the generator terminal voltage and the rotor speed deviation are used as its inputs. Calvoic at et al [DCNS 97]. As a result. This sort of knowledge is sometimes difficult to acquire and represent in the required form. The fuzzy controller was implemented in the control ACE calculation. All of these advantages have enabled this technique to attract more and more attention in recent years. Chown. The developed fuzzy logic controller. based on a set of fuzzy logic operations that are performed on controller inputs. with the aim of improving the generating unit transients by acting through the exciter input. both the voltage profile and the dynamic stability of the generating unit are enhanced.P. The extended formulation is combined with the genetic algorithms and simulated. Special Study Report Page 39 of 56 . O. This formulation is then extended to the entire economic dispatch problem when the fuel consumption is higher than the agreed amount in the take-orpay contract. b) Parameters of the fuzzy logic controller are usually determined by trial and error. The main problem solved by this method is the secondary frequency controller and AGC. Jinyu Wena. The controller is suitable for real time operation. S. Variations of the parameters and operation conditions of the controlled systems do not significantly effect the performance of the controller.A.Application of Artificial Intelligence techniques in Power Systems Fuzzy logic Controllers are advantageous in many respects. Stabilizer control and the exciter and governor loops using fuzzy set theory and the Neural nets was developed by M. R. provides a means of converting linguistic control requirements based on expert knowledge into an efficient control strategy. Short term generation scheduling with take-or-pay fuel contract was developed by Kit Po Wong and Suzannah Yin Wa Wong et al [KSY 96] in which a fuzzy set approach is developed to assist the solution process to find schedules which meet as closely as possible the take-or-pay fuel consumption.C. Hartman et al [CH 98] for Automatic Logic Controller (AGC). The FLC. The main disadvantages of this method are a) Knowledge used to design a fuzzy logical controller mainly comes from the heuristic knowledge or expertise of the human experts. whose control signals are adjusted using the on-line measurements.B. Also FLC design has been carried out by G. Using unsupervised learning of ANN generates a fuzzy associative matrix. Malik et al [JSM 98] suggested a method to design the FLC based on Genetic Algorithm (G A).annealing optimization methods for the establishment of new algorithms for the problem.

Shabani. The conventional approach based on simulation of probable contingencies is not suitable for on-line security assessment because of the large computation time involved. If W is found to be large. Special Study Report Page 40 of 56 . N. This estimator is more robust than the WLS estimator.g. etc). 5. In the past. This not only provides a very reliable security classification but the fuzzy grade membership also provides a quantitative ' level of confidence ' for the security classification. The notable drawback of this method is the poor computational efficiency for large sized problems. The main disadvantage of this method is the presence of the gross errors. the weighted least absolute value (WLAV) has been applied to power system problems. The security status of the input pattern is determined using a fuzzy estimation technique. Pattern Recognition is one of the potential methods.2. some pattern recognition methods have been proposed for power system security assessment. Sinha et al [AKS 95] presented a PR and fuzzy estimation technique. Here the knowledge about the system operating conditions is stored in a structured memory by grouping similar patterns into clusters which are arranged into a hierarchical tree structure.Application of Artificial Intelligence techniques in Power Systems 5. which fits the computational requirements of online security assessment. This enables a very fast two level search for the near neighbors of the input pattern. For the successful operation of large-scale power systems the optimal estimation of the state is required. State estimation is the task of determining the actual values of the state variables . K.4 State Estimation The power system state estimation is another area were fuzzy logic applications are performed in recent times. The optimal estimator is controlled by the parameter W. then more weight is placed on the current state estimate in relation to the measured value and vice versa.One of the problems in automating a power system is the construction of reliable models of the system whose state variables can be identified sufficiently accurately using available noisy system data. which the weight is given to the current state estimate calculated using the WLS method. F. These methods security classification schemes are not well suited for large power systems because of convergence problems faced in designing the classifiers in a large dimensional pattern space. The weighted Squares (WLS) estimator is widely and extensively used due to their numerical stability and computational stability. tripping of generators.2.5 Security Assessment On line security assessment of a power system involves monitoring the current operating condition of the system and assessing the effects of probable contingencies (e. outages of transmission lines. An alternative state estimation approach. R. In this method variant of the Kalman State Estimation is taken as the basis. Prasad et al [SPS 96] formulated a method which uses the combination of weighted least squares and fuzzy logic based techniques to improve the state estimation of the power systems.

7 Load Forecasting Load forecasting is an important task for the efficient operation of a power system. Expert knowledge is used to model the system behavior and response. Hyun-Joon Cho and J. Fuzzy linguistic variables were used to characterize the load patterns of several types of days. The so-called Sagittal diagrams were build which represents the fuzzy relations for power systems and diagnosis were done using these diagrams.2.6 Fault Diagnosis and Restoration Fault diagnosis and restoration is perhaps the most popular area of the AI implementation where a large number of alarms have to be interpreted in real time to determine possible fault scenarios. based on which suitable restorative actions need to be taken. Many implementations for various fault diagnosis problems have been reported in the literature. Considerable improvement in the accuracy of the forecast hourly loads was reported. One of the major obstacles in implementing and using a SLTF (Short Term Load Forecast) has been the lack of user trust and confidence in the model. Ranaweera. The mathematical Special Study Report Page 41 of 56 . A fuzzy front-end processor was used in this work to enhance the forecasting accuracy by preprocessing the inputs. The proposed methodology uses fuzzy rules to incorporate historical weather and load data. The malfunctioning of relays and circuit breakers based on the alarm information and the estimated fault sections were estimated. Fuzzy expert systems are now being used for these applications to include vague constraints and express uncertainty. N. 5. The system provides the fault section candidates in terms of the degree of membership and the malfunction or wrong alarm.2. These fuzzy rules are obtained from the historical data using a learning-type algorithm. Following a fault an efficient restoration plan was generated using a heuristic search method. D. Torres and Mukhdekar [TM 89] developed a fuzzy knowledge based forecasting tool for distribution feeder load. Some recent works have reported successful application of fuzzy logic for expressing the vague relationship between forecast load and various parameters in which depends. both numerical as well as fuzzy. Park et al [HJ 97] proposes an expert system using fuzzy relations to deal with uncertainties imposed on fault section diagnosis of power systems. K. Application of fuzzy set theory in fault diagnosis was first reported by Xu et al [XZL 90]. Hsu and Ho [YK 92] first proposed a fuzzy expert system for short term load forecasting. Hubele et al [RHK 96] presented a fuzzy logic based short term load forecasting. coping with uncertainties. Here some of the advantages and important implementation issues based on practical experience were highlighted. A fuzzy method to deal with the uncertainty concerning fault location in distribution networks was also developed.Application of Artificial Intelligence techniques in Power Systems 5. The operator monitors these candidates and is able to diagnose the fault section. The load of each load points in the distribution system was estimated using a fuzzy expert system. K. F.

K.2.Application of Artificial Intelligence techniques in Power Systems complexity while designed to capture the nonlinear relationships between inputs (past load. fuzzy computations are performed. SVC and generator excitation meetings. 5. S.e. The linguistic variables of the system consists of 1. If these mathematical relationships could be reduced to logical table. medium (M). Thukaram et al [PuUTPa] presented a new technique using fuzzy set theory for reactive power control with the purpose of improving the voltage stability of the power system.For the output of the system the four terms are included as L. Sensitivity of the voltage stability index to control variables such as OLTC. past and predicted temperature) and outputs (predicted load) and does not offer the user an intuitive understanding. Different states are developed as low (L). G. The concept is as the same in reactive power planning and control which leads to better voltage profile. from the load flow algorithm incorporating the load characteristic and the generator control characteristics.8 Voltage Stability Enhancement Fuzzy Control Approach has been effectively presented in the Voltage Stability Enhancement too. This is found. N Udupa and D. The performance obtained from testing the above fuzzy controlled system was found to be encouraging. The load flow result is obtained for a given system operating characteristics or from the online state estimator. or assist in generating the system forecast. The Fuzzy conditional statements are then prepared Based on the values of the input variables fuzzy sets are formed. small ( S). First the L index is computed for the system. Voltage stability index. For the controllers three terms are used mainly i. Here the voltage stability index (L index) n and the controlling variables are translated into fuzzy set of notations to formulate the relation between voltage stability level and controlling ability of controlling devices. could be well developed solely from expert knowledge. which is in essence a set of logical statements. Then the L index sensitivity is computed. Base case load flow is performed ( or from state estimation) Special Study Report Page 42 of 56 . The fuzzy logic. M. L-index 2. Then a fuzzy ruled-based system is formed to select the controllers.THEN rule then there is the possibility that the user would gain confidence in the model and therefore use it to generate. Z. their movement direction and the step size. high (H) and very high (VH) for the L index value.medium(M) and large(L). The terms of the linguistic variables are used to describe the states of the system. such as a set of IF .Purushothama. Algorithmic steps in the proposed control methodology are 1. Using the terms of the linguistic variables and Rule base.

Application of Artificial Intelligence techniques in Power Systems 2. If the maximum L index value is not acceptable within tolerance and margin is available for the controllers to 4.indices with the suggested controller settings. 4. 6. Estimate new L. 3.index is acceptable within tolerance go to step 7. Perform the load flow with the suggested controller settings and output results. Observe the sorted list of nodes according to their L-index. Special Study Report Page 43 of 56 . If maximum L. 7. S' are found. Matrices S l. Using the available margin of the controller settings are evaluated so as to minimize the Lindex of those nodes where it is more than the acceptable level. 5. Corrections to the controller settings are evaluated so as to minimize the L-index of those nodes where it is more than the acceptable level. Sensitivity S is computed.

It is possible to analyze if the samples chosen are small in size. Here the error back scheme is widely used. which are the case of the practical system. 6. Generally at least twice of as many nodes in the hidden layer has been taken as Inputs. Also research results are available for dynamically designs hidden layers. Moreover its not easy to obtain good performance on training data followed by much worse performance on test data. and SOM) designed for specific tasks simplify the design process.Application of Artificial Intelligence techniques in Power Systems CHAPTER 6 ANALYSIS OF THE TECHNIQUES 6. But still some trial and error is needed to produce quick convergence and acceptable results. The introduction of the concept of structured ANNs (e. Some of the researchers gave an empirical formula as H = ni (ni-1) to calculate hidden layer where 'H' is the number of the hidden layer and 'n i' the input. But some experimentation is required to select the number of hidden layers and nodes. Perceptrons.1.1 Neural Network based Applications The most of the applications related to neural network is based on multilayer perceptron. there is no efficient way of generating a complete training set to cover all possible operating states. Cascaded correlation's begins with minimal network. Once the hidden layer is added it becomes a permanent feature detector in ANN.) the analysis will be extremely difficult.2 Training Set Generation In many applications. Fundamental aspects of Multilayer Perceptron networks are random initial start up state and convergence of connection weights to produce minimum error. Special Study Report Page 44 of 56 . For example. If the sample is large (500 buses.1 Design of Network As discussed in practical applications Multilayer Perceptron with at least one hidden layer is used. It has been reported that using greater number of hidden layer improve the overall performance. However there are no set rules for parameter selection associated with these algorithms. then automatically trains and adds new hidden units one by one. Hopfield Network.1. In the cases of power system security problem most of the literatures reports about offline simulation to obtaining the training sets. This architecture learns quickly.g. 6. So in using ANN models some trial and error is required. There can be improvement if some knowledge can be incorporated about the domain into the network architecture. This will be of greater concern in dealing with a problem of large on line data handling.

1.5 Knowledge Consistency and Interaction with the User Knowledge Consistency is an important concern in the training set of ANN research.1. 6. most of the implementations use a sequential algorithm on conventional computer to train the ANN. Another drawback is that the weights and thresholds are calculated based on the optimization process. When using the multilayer perceptron model. In many cases AI technique is required to interact to demonstrate the validity of the decision to the User. Similarly in preventive control an explanation might be necessary to validate and verify the control strategy. in node by node manner. The reason is that we should get the best rather than the feasible ranking of contingencies. which has to be repeated if any of the input parameters change.4 Training the Inputs Many of the ANN models (like perceptron. So feature selection is necessary to reduce this pattern space to a reasonable size.6 Practical Implementation In the hardware part most of the present day ANN schemes are single-processor simulations of the massively parallel ANN models. Ideally ANN schemes should be implemented in parallel processing machines to fully reap the benefits of their massively parallel structure. However. 6. SOM. This approach can potentially provide a very good performance. Special Study Report Page 45 of 56 . Also a mapping method is formulated from which the weights and thresholds for the particular optimization problem can be easily computed. Direct Implementation in which there is a physical-processing element for each neuron in the neural network. defined in terms of its weights and thresholds. ART Networks heavily rely on the information retained to the input features. For example in the diagnosis of faults in the system. this energy function has many local minima.3 Hopfield Network Hopfield Networks can be very useful in solving the optimization problems very quickly and efficiently by minimizing energy function. This is not acceptable especially in contingency screening.1. 6. There is mainly two way of implementation of ANN in the parallel computers. the operator might want to ascertain the validity of the reasoning employed. However it can support only a specific ANN model since it is fixed in the hardware. In any power system applications the input patterns space consists of a large number of features. The AI implementations are considered complete when they match with human competence and thus further research is needed in this area.Application of Artificial Intelligence techniques in Power Systems 6. The enhancement in the recent development of the architecture reduces these drawbacks.1. 1. These processes make loss of information.

The complexity of the problem makes the solution computationally intensive if solved by conventional technique. Special Study Report Page 46 of 56 . The problem has to be solved by human experts for daily operation and planning. If the problem involves uncertainty. Fuzzy systems are found to be very effective with problems dealing with most of these issues. 6. Common sense knowledge Representation It’s difficult to represent and manipulate common sense knowledge and there are no effective and sufficient methods to do so. It cannot be used if Input-Output mapping is difficult. Flexibility 4. 2. The various issues that needs to be addressed. Computationally less expensive and simpler tools. If the modeling of mathematical problem requires various many assumptions to be made. 5. imprecision and conflicting objectives. leading to an inaccurate models.2 Advantages of Fuzzy Logic Applications The main advantages of the fuzzy systems are 1. It's effective when the problem is non-linear in nature and if there is a convenient way to obtain Input-Output mapping.Application of Artificial Intelligence techniques in Power Systems 2. Virtual implementations (with general-purpose neuro computer) in which a processing element takes charge of multiple neurons and simulates them in a time-sharing fashion.2 Fuzzy Logic based Applications 6. 6. Thus functional knowledge in terms of heuristic rules are available. Ease of computation They are found to be very powerful in applications involving Uncertainties. vague constraints and/or multiple conflicting objectives.2. which lacks in knowledge engineering. 4.1Requirements of Fuzzy based Applications The main characteristics and requirement for a problem suitable for fuzzy logic applications are 1. 3. If the methodology cannot be expressed in terns of mathematical form. That means it depends on expert opinion and cannot decide the rule networks Genetic Algorithms and fuzzy clusters.2. Speed 2. 3. even though fuzzy logic has found in various applications are Creation of fuzzy logic Creation of fuzzy logic is mostly through experts.

Application of Artificial Intelligence techniques in Power Systems Fuzzy Logic Controller Stability Stability of the FLC cannot be assessed and there are no established methods to do that. Special Study Report Page 47 of 56 . This needs to be analyzed before they can be considered as alternative for conventional controller. There is a need to support applications that can be provided quality solutions. Tools and Practical Consideration The lack of tools for this generic development works handicaps the utilization of these systems. Moreover very few applications have been Implemented Practically though many applications are reported.

survey and the important analysis of typical applications of AI techniques (ANN and FUZZY LOGIC) in the field of Power systems. The tools for the simulation of these conditions also need to be enhanced for their limitations. As in the case of Fuzzy Logic applications it can be seen that these techniques can be blended with the conventional systems as well as with the other techniques like Neural Networks and Genetic Algorithms. Hybrid Systems combining the individual strengths of the ESs and ANNs along with the Fuzzy systems seems to be the most promising area in future and promising for the most of the Power system Applications. It helps to analyze and gives the result. The analysis of these techniques is indicated in a broader sense and the practical difficulties are narrated. The fundamentals of the Artificial Neural Network and the Fuzzy Systems are also described. Special Study Report Page 48 of 56 . operation and control. ANNs are mainly used for learning and pattern Recognition for depicting the reference knowledge database. The easiness in evaluating the vague or non-crisp concepts and the ability of these techniques to learn due to the technological improvement elevated the effect of these soft computing techniques. which can be substituted for any logical analysis. The application fields combining the conventional and these techniques can remarkably reduce the difficulties faced in the Power Systems design. It gives the understanding of the strengths of the models. which are used in Power Systems. and the future hybrid models that are useful. The study presents concepts. The hybrid systems thus formed can be the most powerful systems for design. Also the future concentration on the modification of the techniques is analyzed to obtain better result and making these techniques competitive to the human brains. The concepts of the AI techniques are reviewed to understand those categories of models. Moreover there are sufficient scope in the improvement of the various soft-computing techniques to increase their strengths and capability.Application of Artificial Intelligence techniques in Power Systems CHAPTER 7 CONCLUSION The importance of the use of the AI tools has been felt in all the areas of the Power Systems and the need is emphasized. planning and control & Operation of practical problems.

Study on fault Location System for Transmission Lines using neural Networks. S. Monticelli. Wall. IEEE Trans.P.T. Jpn . pp 338-353 (1965). R.H. Wirgan.O. Happ and K. Granville. Vol 1.Cao . Portland OR. Fernades.IEEE Int. Y. Hope.M. 1988 [GPM 88] S. V. pp 413-416 (1989).C.F. A. USA pp 486-489 (1989). Transient Stability Evaluation by PR approach using Multivalent logic. Kanemaru. 1986 [CCM 86] S.Power Apparatus System (PAS). Of Power System .Pereira and A. pp 209-217 (1989). Inst. IEEE Inst. Malik and G.S. Proc. Power System . kaneta and K. Prorano and E. H.Kanoh. Khan. Chow and S.Zadeh.C. R. Large Scale Reactive Power Planning. IEEE Trans. System Control Staudy Group. A. Symp. Lange.C. [BPW 89] [FKCR 89] Special Study Report Page 49 of 56 . J.P. Cheng. Screening Power System Contingencies using a back propogation trained multipercptron. SC 88-22. M. Circuits and Systems (ISCAS). An adaptive synchronous Machine Stabilizer. Fischl. H. J.A. M. Doc. J. Engg. pp 101-109 (1986). [HMK 88] 1989 [AV 89] A. Ricciardi. Burdett. J. 1983 [FLBHW 83] R. An Integrated Methodology for VAR Sources Planning. 3 (1988). Symp. Fuzzy Sets Information Control. Electr. Implementation of Intelligent Controller using neural network state estimator.V. pp 1083-1088 -102 (1983). pp 53 ( 1988). IEEE Trans. Machias. Intelligent Control. F. EPSR Vol 17.Application of Artificial Intelligence techniques in Power Systems BIBLIOGRAPHY 1965 [LAZ 65] L. Bialasiewiez. Proc. pp 549-557.

H. R. IEEE Trans. USA. IEEE Trans on Power system. M. UA.Sun. Vol 5(2) . pp 639-643. Z.R. pp 285-311 (1989). Nonlinear Stabilising Control of MultiMachine System. Proc. IEEE Trans. on Power Systems. Fischl. Ricciardi and H. Seattle. Yee. A Neural Network Processing Strategy for the Detection of Incipient Fault on Distributin Feeders. Expert System Applications to Power Systems. sensor validation for Power Plants using adaptive back-propogation neural networks IEEE Trans. Power Topological Observability analysis using a neural network Model. Real Time Application of ANN for Incipient Fault detection of Induction Machines. pp 1040-1047 (1990). Power Delivery. Chow .H.Pao.H.Machias and D. pp 2925-2928 (1990). Brighton UK. Proc. WA. 1. MukhedKar. Symp. [LS 89] [MT 89] [PS 89] [SMP 89] [TM 89] 1990 [CY90] M.V. Ebron.Application of Artificial Intelligence techniques in Power Systems [JHA 89] J. IEEE Trans. [EL90] [EU 90] [FKCRY 90] Special Study Report Page 50 of 56 . (1989). L. Power System . S. D. The control of synchronous machine using Optimal Control Theory. Distribution feeder load modeling An approach using fuzzy sets. pp 1030-1036 (1990). J. Q. . E. (1989). Vol 4. Vol 59 pp 25-35 ( 1989).C. 37(1) ( Part 2). New Orlean LA USA. IEEE. Circuits systems (ICSAS). H. Souflis.308. J. Papadias. London.J. Sci. Eryurek and B. Vol 4(3).Anderson. Yan. White. Industrial and Engg Application of AI and Expert Systems (IEA/ AIE). Khan. A Neural Network based dynamic Security Assessment for electrical Power System. Sobajic. An application of fuzzy concepts to transient Stability evaluation. G. Mori and S.. Lamber Torree and D.O. Y. Vol 4.USA.L. Nucl.Conf Publ. Upadhyaya. Proc.Chow and S. L. no.Lubkeman and M.pp 905-914 (1990). IEEE Int. and Simp. Lu. D.C. An Improved Hopfield Model for Power System contigency classification.Y. Y. pp-220-228 (1989). Tsuzuki. Charleston. Proc.No. pp 236-241 (1989). Eur Conf : Circuit Theory and Design. IEE. S. A.

138 (1991). Proc.Park. R. Design of Fuzzy Power System Stabiliser for Multi Machine Power Systems. Procee.Y. Xu. IEEE Trans. State Estimation using ANN. SC. [HC 91] [HY91] [MOG 91] [PEM 91] Special Study Report Page 51 of 56 . pp-271-277.Organised Feature maps for day type identifaction. pp 536-545 Vol 6 ( 1991). ANN Methodology for short term load forecasting. Power System state evaluation by structural neural Networks.P. pp 407-413.Y.C. USA pp 227-233 (1990). Nishimura and M.Hsu. S.IEEE Proc.Application of Artificial Intelligence techniques in Power Systems [HC 90] Y. Design of Fuzzy Power System Stabilizer for multimachine Power Systems .T. Y.A. Y. IEEE Trans.Hsu and C. pp 552-556 (1990). Joint Conf.Damborg. IEEE Trans on energy Conv 6(3). H.Yang.E. S. C. 6 pp-442-449 (1991). O. Hsu.R.6 pp 612-619 (1991). Yee . A. Engg. 137 : pp 233-238 . IEEE Trans. Energy Convers . Hsu and C. (1991).Park. Lin. K.Lee.3 pp 233-238 (1990).Kanekar and A. Energy Covers. Int. Feliachi. 22nd South Eastern Symp. Atlas and M. CA. Y. (1990).Chen. IEE Proc Vol 137.H. Part 1: Self. Indst.El-Sharkawi. Hassan. Electric Load Forecasting using an ANN. K.Y. Fuzzy diagnosis method in machinery condition monitoring and failure diagnosis .J. Malik and G. (1990). System Theory CS Press. Zhang and Y. Design of ANN for Short Term Load Forecasting.Cha and J.Y. PartC. M. WSF Workshop ANN in Power system Engineering methodology Clemson. Y. Tuning of Power System Stabilizers using an Artificial Neural Network. USA. C.Y. L. Cheng. pp 407-413. Marks II. A. M. USA (1990). A.Y. Proc. Cheng. CA. J. D. C. Los Alamtos C. No. Methodology for online Incipient Fault Detection in Single Phase Squirrel Cage Induction Motor using ANN. Hope. A fuzzy logic based stabilizer for synchronous Machine. [KF 90] [LCP 90] [NA 90] [XZL 90] [YC 90] 1991 [CY 91] M. Rai.conf : Los Angels.C. Electr. Power System. M. O.J.Chow and S. Neural networks (IJCNN). Sandiego.

Part C. Hassan and O.H. Shahidehpour.Vankayala .pp 399-408 (1994). Y. IEE Proc.Lecture Notes in Artificial Intelligence (LNAI 638) Edition 1992. A Neural Network ShortTerm Load Forecaster. IEEE Trans. Djukanovic. D. mammone. (1991) 1992 [AFR 92] A.Application of Artificial Intelligence techniques in Power Systems [RY 91] Richard. EPSR. Vidya Sagar . Northern Ireland. EPSR Vol 16.Hsu and H. No.Hong and C.C. Ho.L. Yehoshyua Zeevi. 139 (6): pp 471-477 (1992).Cheng. S. IEEE Trans.Liu.Y. Edited 1991.Liew and C. Neural Computing Research and Applications-. pp227-234 (1994).Based Optimal Reactive Power Control.Vol 28.Abdul_rahman and S. M. 28 pp 67-79 (1993).2.C. J. M. No. [HM 93] [VVR 93] 1994 [DLC 94] Dipti Srinivasan.Y.6 .M.Rocha. Neural Nets -.Y. A.C. D.A theory of Brains and Machines. N. Hsu and K. Vol 8.M. Neural Net Based Tangent Hypersurfaces for Transient Security Assessment of Electric Power Systems. Pao.. Y. pp 662-670 (1993). [GO 92] [HL 92] [YK 92] 1993 [AS 93] K.A.F. Inc.Research and Applications Proc.J. Heuristic based fuzzy relationing approach for [DSP 94] [YH 94] Special Study Report Page 52 of 56 . Y. Implementation and laboratory test results of a fuzzy logic based self-tuned PSS.S. Fuzzy expert Systems: An Application to short term load forecasting. Power System. EPSR. 25-26 June 1992. 2nd Irish Neural networks Conference Belfast. IEEE Trans. Power system: pp 505-512 (1992). H.Sobajic and Y. Neural Networks Theory and Application. A heuristic and algorithmic approach to VAR planning. Kuo. Academic Press. Rao. Artificial Neural Networks and their applications to Power Systems---Bibliographical Survey. P. on energy conversion 8 (2) : pp 221-228 ( 1993). Malik. A Fuzzy.

Neural Networks Applied to Preventive Control Measures for the Dynamic Security of [CTH 96] [FPV 96] Special Study Report Page 53 of 56 . A. Irisarri. The Application of Artificial Neural Networks to Substation Load Forecasting. Mohammed and N. Y. EPSR 38 pp 153-160 (1996). Hwang.Optimal Shunt Capacitor allocation by Fuzzy Dynamic Programming. Miller. Kirschen.Rao. pp 391-397 (1995).Sinha. Zadeh. EPSR. Power System. IEEE Trans.Integration of Artificial Intelligence Applications in the EMS : Issues and Solution.Artificial Neural Network based Fault Diagnosis System for Electric Power Dustribution Feeders. A.1.Ruzic. B.Tzeng.A. A Neural Net based Short Term Load forecasting using moving Window Procedure. J. Y. M. pp 475-482 (1996). Power 1995 [AJA 95] Feed Forward neural Networks Vector Decomposition Analysis. Miranda.P.Fidalgo. Witold Pedrycz. Hong-Chan Chin. Modelling and Anolog Implementation.S. [WP 95] 1996 [BIKMM 96] Bann. Vol 17. [AKS 95] [AQ 95] [DRSP 95] [HC 95] [MR 95] E.J. CRC Press (1995). Alves DaSilva. Power System Security Assessment using pattern Recognition and fuzzy estimation.Chen. pp51-60 (1995). pp 1-10 (1995). pp 133-139 (1995).A. (1994). Quintana. B. pp11-19( 1995). J. C.H. Pao. Vol 11. EPSR Vol 17.H.No. No. D. EPSR Vol 35. Fuzzy Sets Engineering with Foreward by Lotif. No 6.M.K. On Delivery 9 : 2. Pattern Analysis in Power System State Estimation. S. EPSR 35.C. No1. IEEE Trans.V.D. D.Babic.1. Sobajic. S. Vol17.Application of Artificial Intelligence techniques in Power Systems distribution system service restoration. EPSR. Mokkhtari . Kluwer International Series in Engineering and Computer Science (1995). A.J. Djukanovic. J. Pecas Lopes and V.

EPSR Vol 37 pp 39-44 (1996).ANN and their Applications in Power Systems-Special Study Report (1996). M.Ghosh and B .I. M. H.F.4. N. S. Vol 11.A. pp 1749-1755 (1996). A.No. EPSR.Term Generation Scheduling with Take-Or_pay Fuel Contract. Security-Constrained Optimal Rescheduling of Real Power Using Hopfield Neural Network. A.H.2. Vol 11. Power Systems. 4 pp 1811-1816 (1996). S. G.Speed Deviation Driven Adaptive Neural Network based Power System Stabilizer.Ranaweera. Kit Po Wong. No. H.Lafuente.El-Sherbiny. Prudenzi.Orsolini Cencellli. C.1 pp 128-136 (1996). Vol 18. EPSR Vol 18. Combined Genetic Algorithm/Simulated Annealing/Fuzzy Set Approach to Short. Agarwal. [JJ 96] [KSY 96] [KWA 96] [LPSCO 96] [PHL 96] [RHK 96] [RRMLMP 96] [SC 96] [SPS 96] [ShSaI 96] Special Study Report Page 54 of 56 . N. Rodriguez. EPSR . pp55-60 (1996).K.H. No.Vol. EPSR. 18. Chowdhury. Lamedica. [GC 96] S. Warke and S.R. R. No. S.Perez . 5 pp 217-277 (1996). IEEE Trans. Design of an ANN for fast line flow contingency ranking. Vol 39. IEEE Trans.Khaparde.A. Y.Caciotta. A fuzzy logic supported weighted least squares state estimation. Vol 11.H.4 pp 1743-1748 (1996). D.Muguerza and J. No.H. F. IEEE Trans.Hubele and G.Fault Analysis with Modular Neural Networks. Sforna. N. A Neural Network Based Technique for short-term Forecasting of Anomolous Load Periods.Vol 18. M. pp 215-222 (1996).Park.Smolleck. Fuzzy logic for short term load forecasting. Jovitha Jerome . EPSR . Chowdhury. 4 . Hyun and J.pp 169-175 (1996). El-Saady and E. No.V.G. Lee .M. S. EPSR . Karady. A. IEEE Trans.Rementeria.pp 297-305 (1996). Power Systems. Power Systems Vol 11 No. Power System Stabilizer based on inverse dynamics using an Artificial Neural Network. An adaptive approach in distance protection using an ANN. Power System.K.Prasad. J.Application of Artificial Intelligence techniques in Power Systems Isolated Power Systems with Renewables. Shabani.Vol 38 .Ghosh and B .Martin. J. No. Suzannah and Yin Wa Wong.5 . pp99-110 (1996). Ibrahim.

M.4 pp. Srinivas Pillutla.Djukanovic.Morgan. Identification of Power System Load Dynamics Using ANN. Vol 12.C. C. 4 . A.D.Dobrijevic. C. A new Constructive ANN and its Application to Electric Load Representation. Y.pp1569-1577 (1997). [XSJMW 96] 1997 [AFZL 97] A. IEEE Trans. No.3 pp 654-659 (1996). A new fuzzy Control Approach to Voltage Profile Enhancement for Power Systems.M. H. IEEE Trans. Vol 12. No 4. Printed by prentice-Hall of India Pvt.J.Ahalt. Gwo-Jen Chiou .B. An expert System for Fault Diagnosis of Power Systems using Fuzzy Relations.M.Y. Ferreira. Power System.Xuan. A.A.Application of Artificial Intelligence techniques in Power Systems [SL 96] Ching-Tzong Su and Chein-Tung Lin. Power Systems.Williams. pp 489-499 (1997). 4.IEEE Trans.S.Choueiki. pp 1759-1763 (1997).No6. No. lambert-Torres. IEEE Trans on Power Systems Vol 12.1. Power System Stabilization based on Modular Neural Network Architecture. Ltd (1997). Building A Quasi Optimal Neural Network To Solve The Short-Term Load Forecasting Problem. Power Systems.Koplowitz.M. IEEE Trans Power System. Ali Keyhani.M. EPSR Vol 19.C. J. No.Song. No.Sobajic. C. Vol 11.Novicevic and D. M.A fast Computation Hopfield Method to Economic Dispatch of Power Systems. pp1468-1472 (1997). pp 57-66 (1996).8 . Ching Tsong Su. R. No.D. pp 411418 (1997). Taylor. M.Zambroni de Souza and G. Power System.12. EPSR . Bart Kosko. Vol.T.Vol 36.Alves da Silva. [BK 97] [BKT 97] [CG 97] [DCNS 97] [HJ 97] [MCS 97] [SA 97] Special Study Report Page 55 of 56 . Q. Performance of an Adaptive Protection Scheme for series compensated EHV transmisssion systems Using Neural Network.Vol 19. No. S. W.Campbell. pp 342-348 (1997).Vol 12. Calovic.1432-1439 (1997).P.H.Coordinated strabilising control for the exciter and governor loops using fuzzy set theory and neural nets. Hyun-Joon Cho and Jong-Keun Park.Bostanci. IEEE Trans. EPSR .Johns. Neural networs and Fuzzy Systems—A dynamical Systems Approach to machine Intelligence.

Online Topology Determination and Bad Data Suppression in Power System Operation using ANN. Vol. Malik.13.Khorasani. No. IEEE Trans. IEEE Trans. P. Vol.C. IEEE Trans. Application of a Radial Based Function ( RBF) Neural Network for Fault Diagnosis in a HVDC system. R. No. IEEE Trans. K .Nerandra.1. Vermaak.Application of Artificial Intelligence techniques in Power Systems 1998 [CH 98] G. Power Systems.Botha. Feuillet. Vol.1. No. No3. Vol 13. Sood.Design and Experience with a Fuzzy Logic Controller for Automatic generation Control (AGC). Power Systems.A. Patel.M. M. IEEE Trans.13. Power Systems. Improvement of Transient Stability Limit in Power System Transmission Lines Using Fuzzy Control of FACTS Devices. pp917-922 (1998) J. Vol 13.Sadeghzadeh. No.3.G.chown and R. IEEE Trans. No. Narendra Uduppa. Hartman. Vol 13. Jinyu Wen. Phd Thesis (1998). On line development of Intelligent Tools for Applications in Energy Control Center. [JSM 98] [NSKP 98] [NU 98] [SLA 98] [SEHFH 98] [VB 98] Special Study Report Page 56 of 56 .13.N. pp-796-803 (1998). E. Shijie Cheng and O. S. V.Hadj Said. R. K.3. Power Systems. A Synchronous Generator Fuzzy Excitation Controller Optimally Designed with a Genetic Algorithm. pp 177-183 (1998).3 pp 965-970 (1998). Power Systems. Ehsan. pp884-889 (1998). Power Systems. Recurrent Neural Networks for Short-Term Load Forecasting.C. K. pp 126131 (1996).