IJEST Special Issue Complete

MultiCraft

ISSN 2141-2839 (Online); ISSN 2141-2820 (Print)

Available online at www.ijest-ng.com

International Journal of Engineering, Science and TechnologySpecial Issue:Application of Computational Intelligence in Emerging Power Systems

Guest Editors:S.N. Singh Department of Electrical Engineering, Indian Institute of Technology Kanpur Kanpur- 208016, India

K.S. Verma Department of Electrical Engineering, Kamla Nehru Institute of Technology Sultanpur- 228118, India

Vol. 2, No. 3, 2010

International Journal of Engineering, Science and Technology (IJEST)

Aims and scope

IJEST is an international refereed journal published by MultiCraft. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of engineering, science and technology. Original theoretical work and application-based studies, which contributes to a better understanding of engineering, science and technological challenges, are encouraged.

Journal policy

International Journal of Engineering, Science and Technology (IJEST) publishes articles that emphasize research, development and application within the fields of engineering, science and technology. All manuscripts are pre-reviewed by the editor, and ifappropriate, sent for blind peer review. Contributions must be original, not previously or simultaneously published elsewhere, and are critically reviewed before they are published. Papers, which must be written in English, should have sound grammar and proper terminologies.

Call for papers

We invite you to submit high quality papers for review and possible publication in all areas of engineering, science and technology. All authors must agree on the content of the manuscript and its submission for publication in this journal before it is submitted to us. Manuscripts should be submitted by e-mail to the Editor at: [email protected]

Call for Reviewers

Scholars interested in serving as volunteer reviewers should indicate interest by sending their full curriculum vitae to us. Reviewers determine submissions that are of quality. Since they are expected to be experts in their areas, they should comment on the significance of the reviewed manuscript and whether the research contributes to knowledge and advances both theory and practice in the area.

International Journal of Engineering, Science and Technology (IJEST)Editor: S.A. Oke, PhD, Department of Mechanical Engineering, University of Lagos, Nigeria E-mail : [email protected] Associate Editor (Electrical Engineering): S.N. Singh, PhD, Department of Electrical Engineering, Indian Institute of TechnologyKanpur, Kanpur-208016, India

EDITORIAL BOARD MEMBERS

Kyoji Kamemoto (Japan) M. Abdus Sobhan (Bangladesh) Sri Niwas Singh (India) Shashank Thakre (India) Jun Wu (USA) Jian Lu (USA) Raphael Jingura (Zimbabwe) V. Sivasubramanian (India) Shaw Voon Wong (Malaysia) S. Karthikeyan (Sultanate of Oman)

Amir Nassirharand (Malaysia) N. W. Ingole (India) Fatih Camci (Turkey) Milorad Bojic (Serbia) Asim Kumar Pal (India) Prasanta Sahoo (India) Vidosav D. Majstorovich (Serbia) K. Somasundaram (India) Shashi Anand (India) Ian Blenkharn (UK)

Petr Konas (Czech Republic) Angelo Basile (Italy) Syed Asif Raza (Qatar) Atif Iqbal (India) Kampan Mukherjee (India) P.K. Kapur (India) Haitao Huang (Hong Kong) P. Thangavelu (India) Yechun Wang (USA) Abdul Ravoof Shaik (Australia)

Alistair Thompson McIlhagger (UK) K.I. Ramachandran (India) Jian Ma (USA) P.K. Tripathy (India) Alan Rennie (UK) J. Paulo Davim (Portugal) MKS Sastry (West Indies) Rajneesh Talwar (India) A. Moreno-Muoz (Spain) Prabin K Panigrahi (India)

Vinay Gupta (India) Mohammed Al-Nawafleh (Jordan) Bhu Dev Sharm (India) Tien-Fu Liang (Taiwan) Ranjit Kumar Biswas (Bangladesh) Siba Sankar Mahapatra (India) Sangeeta Sahney (India) Kuang-Yuan Kung (Taiwan) Eleonora Bottani (Italy) Evangelos J. Sapountzakis (Greece)

A.M. Rawani (India) Saurabh Mukherjee (India) P. Dhavachelvan (India) A. Bandyopadhyay (India) Velusamy. Sivasubramanian (India) S. Vinodh (India) Vctor Hugo Hinojosa Mateus (Chile)

International Journal of Engineering, Science and Technology (IJEST)REVIEWERS

The following reviewers have greatly helped us in reviewing our manuscripts and have brought such submissions to high quality levels. We are indebted to them.

Marcus Bengtsson (Sweden) Kit Fai Pun (West Indies) Peter Koh (Australia) Erhan Kutanoglu (USA) Jayant Kumar Singh (India) Maneesh Singh (Norway) RRK Sharma (India) Jamil Abdo (Oman) Agnes S. Budu (Ghana) Yuan-Ching Lin (Taiwan)

Withaya Puangsombut (Thailand) Abd Rahim Abu Bakar (Malaysia) Fakher Chaari (Tunisia) Ghosh Surojit (India) Umut Topal (Turkey) Maloy Singha (India) Parviz Malekzadeh (Iran) G. Possart (Germany) Masoud Rashidinejad (India) Vera Meshko (Republic of Macedonia)

Jun Luo (China) Uday Kumar (Sweden) Tamer Samir Mahmoud (Egypt) Arijit Bhattacharya (Ireland) M.R. Sharma (India) Hyung Hee Cho (Korea) Souwalak Phongpaichit (Thailand) Elsa Rueda (Argentina) Ming-Kuang Wang (Taiwan) Ruey-Shin Juang (Taiwan)

Marisa Viera (Argentina) Shiguo Jia (China) S. Devasenapati Babu (India) Rajeeb Dey (India) Subrata Kumar Ghosh (India) Timothy Payne (Australia) Diwakar Tiwari (India) Mustafa Soylak (Turkey) Jerzy Merkisz (Poland) Md Fahim Ansari (India)

Jiun-Hung Lin (Taiwan) Tzong-Ru Lee (Taiwan) Subir Kumar Sarkar (India) Kee-hung Lai (Hong Kong) Jochen Smuda (Switzerland) Roland Hischier (China) Ahmed Abu-Siada (Australia) Hamzah Abdul Rahman (Jordan) Chih-Huang Weng (Taiwan) Yenming Chen (Taiwan)

Dinesh Verma (USA) Devanandham Henry (USA) M. Habibnejad Korayem (Iran) Radu Radescu (Romania) Hsin-Hung Wu (Taiwan) Amy Trappey (Taiwan) A.B. Stevels (Netherlands) Liang-Hsuan Chen (Taiwan) Richard Hischier (Switzerland) Shyi-Chyi Cheng (Taiwan)

Andrea Gerson (Australia) Ingrid Bouwer Utne (Norway) Maruf Hossain (Bangladesh) Enso Ikonen (Finland) Kwai-Sang Chin (Hong Kong) Jiunn I Shieh (Taiwan) Hung-Yan Gu (Taiwan) Pengwei (David) Du (US) Min-Shiang Hwang (Taiwan) Ekata Mehul (India)

Shashidhar Kudari (India) Khim Ling Sim (USA) Rong-Jyue Fang (Taiwan)Chandan Guria (India) Rafael Prikladnicki (Brazil) Juraj Kralik (Slovak) Indika Perera (Sri Lanka) R K Srivastava (India) Ramakrishnan Ramanathan (UK)Suresh Premil Kumar (India)

Fernando Casanova Garca (Colombia)J. Ashayeri (The Netherlands) Siddhartha Kumar Khaitan (USA)Jim Austin (UK) Rafael Prikladnicki (Brazil)V. Balakrishnan (India)P. Dhasarathan (India) R. Venckatesh (India) Sarmila Sahoo (India) Avi Rasooly (USA)

Barbara Bigliardi (Italy) Huiling Wu (China) Ahmet N. Eraslan (Turkey)

Ryoichi Chiba (Japan)P.K.Dutta (India)Kimon Antonopoulos (Greece)

Josefa Mula (Spain) Amiya Ku.Rath (India) Fabio Leao (Brazil)

Francisco Jesus Fernandez Morales (Spain)

MultiCraft

International Journal of Engineering, Science and Technology

Vol. 2, No. 3, 2010, pp. i-ii

INTERNATIONAL JOURNAL OF

ENGINEERING, SCIENCE AND TECHNOLOGY

www.ijest-ng.com 2010 MultiCraft Limited. All rights reserved

Editorial

Due to increased interconnection and loading along with deregulation and environmental concerns, electric power systems, around the world, are changing in terms of structure, operation, management and ownership. This makes the electric power system complex, heavily stressed and thereby vulnerable to cascade outages. Electric power utilities are trying to provide intelligent and smart solutions with economical, technical (secure, stable and good power quality) and environmental goals. There are several challenging issues in the smart grid solutions such as, but not limited to, forecasting of load, price, ancillary services; penetration of new and renewable energy sources; bidding strategies of participants; power system planning & control; operating decisions under missing information; increased distributed generations and demand response in the electric market; tuning of controller parameters in varying operating conditions, etc. The conventional methods in solving the power system design, planning, operation and control problems have been very extensively used for different applications but these methods suffer from several difficulties due to necessities of derivative existence, providing suboptimal solutions, etc. Computational intelligence (CI) is a new and modern tool for solving complex problems which are difficult to be solved by the conventional techniques. Computation intelligent methods can give better solution in several conditions and being widely applied in the electrical engineering applications. In this special issue, sixteen papers, which are focusing on various areas of emerging power systems such as improving power transfer capability, transformer protection, multi-area economic dispatch, power system security, distribution system planning and reliability, power quality, shipboard power systems, automatic generation control, flexible ac transmission systems, collaborative control, etc., are considered. These papers utilize and address the various computational intelligence techniques such as artificial neural networks, genetic algorithms, particle swarm optimization, evolutionary techniques, ant colony and harmony searches, fuzzy systems, etc. First paper summarizes the various power system problems and potential CI techniques to solve those problems. Key issues and challenges in power quality problems are discussed in detail in one of the papers. These papers will be very useful to the researchers and academicians, but not limited to, in carrying out the research in power systems using computational intelligence. Being guest editors of the special issue on Application of Computational Intelligence in Emerging Power Systems for the International Journal of Engineering, Science and Technology (IJEST), we would like to thank all the authors for publishing their papers, all the reviewers for reviewing the papers of this special issue and the Editor, IJEST, for encouragement.

Prof. S.N. Singh and Prof. K.S. Verma Guest Editors,

Application of Computational Intelligence in Emerging Power Systems

ii

Sl. No.

Title of paper Page No.

1 Editorial S.N. Singh, K.S. Verma

i

2 Application of computational intelligence in emerging power systems D. Saxena, S.N. Singh, K.S. Verma

1-7

3 Application of artificial neural networks to improve power transfer capability through OLTC A. Abu-Siada, S. Islamand, E.A. Mohamed

8-18

4 Application of radial basis neural network for state estimation of power system networks J.P. Pandey, D. Singh

19-28

5 Improved transformer protection using probabilistic neural network and power differential method Manoj Tripathy, R. P. Maheshwari, H. K. Verma

29-44

6 Intelligent agents under collaborative control in emerging power systems H.F. Wedde, S. Lehnhoff, C. Rehtanz, O. Krause

45-59

7 Requirement analysis for autonomous systems and intelligent agents in future Danish electric power systems

Arshad Saleem, Morten Lind

60-68

8 Influence of TCSC on social welfare and spot price- A comparative study of PSO with classical method

S. K. Joshi, K. S. Pandya

69-81

9 Optimal trading strategy for GenCo in LMP-based and bilateral markets using self-organising hierarchical PSO

C. Boonchuay, W. Ongsakul, J. Zhong, F. F. Wu

82-93

10 Allocation of optimal distributed generation using GA for minimum system losses in radial distribution networks

T.N. Shulka, S.P. Singh, K.B. Naik

94-106

11 Distribution system reliability evaluation considering fuzzy factors Xufeng Xu, Joydeep Mitra

107-118

12 Application of ant colony optimization for reconfiguration of shipboard power system Sri Hari Krishna Vuppalapati, Anurag K. Srivastava

119-131

13 Multi-area economic dispatch with tie-line constraints employing evolutionary approach Manisha Sharma , Manjaree Pandit, Laxmi Srivastava

132-149

14 Overall performance evaluation of evolutionary designed conventional AGC controllers for interconnected electric power system studies in a deregulated market environment

Y.L. Karnavas, K.S. Dedousis

150-166

15 A modified harmony search based method for rural radial line planning Yan Yang, Wenxin Guo, Fushuan Wen, Danyue Wu, Yan Lin

167-177

16 Supervised fuzzy C-means clustering technique for security assessment and classification in power systems

S. Kalyani, K.S. Swarup

175-185

17 Power quality event classification: Overview and key issues D. Saxena, K.S. Verma, S.N. Singh

186-199

MultiCraft


Vol. 2, No. 3, 2010, pp. 1-7




Application of computational intelligence in emerging power systems

D. Saxena*, S.N. Singh+1 and K.S. Verma#

* Department of Electrical and Electronics Engineering, Invertis Inst. of Engg.& Tech., Bareilly (UP), INDIA.

+ Department of Electrical Engineering, CET, Denmark Technical University, Kgs. Lyngby, DENMARK. # Department of Electrical Engineering, K.N.I.T Sultanpur (UP), INDIA.

E-mails: [email protected] (D. Saxena), [email protected] (S.N. Singh 1, 1Corresponding author), [email protected] (K.S. Verma)

Abstract Electric power systems, around the world, are changing in terms of structure, operation, management and ownership due to technical, financial and ideological reasons. Power system keeps on expanding in terms of geographical areas, assets additions, and penetration of new technologies in generation, transmission and distribution. This makes the electric power system complex, heavily stressed and thereby vulnerable to cascade outages. The conventional methods in solving the power system design, planning, operation and control problems have been very extensively used for different applications but these methods suffer from several difficulties due to necessities of derivative existence, providing suboptimal solutions, etc. Computation intelligent (CI) methods can give better solution in several conditions and are being widely applied in the electrical engineering applications. This paper highlights the application of computational intelligence methods in power system problems. Various types of CI methods, which are widely used in power system, are also discussed in the brief. Keywords: Power systems, computational intelligence, artificial intelligence.

1. Introduction

Increased interconnection and loading of the power system along with deregulation and environmental concerns has brought new challenges for electric power system operation, control and automation. In liberalized electricity market, the operation and control of power system become complex due to complexity in modeling and uncertainties. Power system models used for intelligent operation and control are highly dependent on the task purpose. In competitive electricity market along with automation, computational intelligent techniques are very useful. As electric utilities are trying to provide smart solutions with economical, technical (secure, stable and good power quality) and environmental goals, there are several challenging issues in the smart grid solutions such as, but not limited to, forecasting of load, price, ancillary services; penetration of new and renewable energy sources; bidding strategies of participants; power system planning & control; operating decisions under missing information; increased distributed generations and demand response in the electric market; tuning of controller parameters in varying operating conditions, etc. Risk management and financial management in electric sector are concerned with finding an ideal trade-off between maximizing the expected returns and minimizing the risks associated with these investments. Computational intelligence (CI) is a new and modern tool for solving complex problems which are difficult to be solved by the conventional techniques. Heuristic optimization techniques are general purpose methods that are very flexible and can be applied to many types of objective functions and constraints. Recently, these new heuristic tools have been combined among themselves and new methods have emerged that combine elements of nature-based methods or which have their foundation in stochastic and simulation methods. Developing solutions with these tools offers two major advantages: development time is much shorter than when using more traditional approaches, and the systems are very robust, being relatively insensitive to noisy and/or missing data/information known as uncertainty. Due to environmental, right-of-way and cost problems, there is an increased interest in better utilization of available power system capacities in both bundled and unbundled power systems. Patterns of generation that results in heavy flows, tend to incur

1 Corresponding author. Tel: +91-512-2597009

Saxena et al. / International Journal of Engineering, Science and Technology, Vol. 2 No. 3, 2010, pp. 1-7

2

greater losses, and to threaten stability and security, ultimately make certain generation patterns economically undesirable. Hence, new devices and resrources such as flexible ac transmission systems (FACTS), distributed generations, smart grid technologies, etc. are being utilized. In the emerging area of power systems, computation intelligence plays a vital role in providing better solutions of the existing and new problems. This paper lists various potential areas of power systems and provides the roles of computational intelligence in the emerging power systems. A brief review of computational techniques is also presented. 2. Potential Area of Research in Power System Using Computational Intelligence

There are several problems in the power systems which cannot be solved using the conventional approaches as these methods are based on several requirements which may not be true all the time. In those situations, computational intelligence techniques are only choice however these techniques are not limited to these applications. The following areas of power system utilize the application of computational intelligence.

Power system operation (including unit commitment, economic dispatch, hydro-thermal coordination, maintenance scheduling, congestion management, load/power flow, state estimation, etc.)

Power system planning (including generation expansion planning, transmission expansion planning, reactive power planning, power system reliability, etc.)

Power system control (such as voltage control, load frequency control, stability control, power flow control, dynamic security assessment, etc.)

Power plant control (including thermal power plant control, fuel cell power plant control, etc.) Network control (location and sizing of facts devices, control of facts devices, etc.) Electricity markets (including bidding strategies, market analysis and clearing, etc.) Power system automation (such as restoration and management, fault diagnosis and reliability, network security, etc.) Distribution system application (such as operation and planning of distribution system, demand side management &

demand response, network reconfiguration, operation and control of smart grid, etc.) Distributed generation application (such as distributed generation planning, operation with distributed generation, wind

turbine plant control, solar photovoltaic power plant control, renewable energy sources, etc.) Forecasting application (such as short term load forecasting, electricity market forecasting, long term load forecasting,

wind power forecasting, solar power forecasting, etc.) Several research papers are published in various journals and conferences. Some conferences in the power system areas are completely dedicated to intelligent system applications and organized regularly such as intelligent system application to power systems (ISAP) held alternate years in different locations of the world. Power system computing conference is another very important conference held once in three years. Similarly, several reputed journals are dedicated to CI applications in the field of engineering and science. There are several books which address of CI application in power systems (Fogel et al, 1966; Sobajic, 1993; Song, 1996; Warvick, 1997; El-Hawary, 1998; Lai, 1998; Wehnekel, 1998; Momoh, 2000). 3. Various Computational Intelligence Techniques Computational intelligence (CI) methods, which promise a global optimum or nearly so, such as expert system (ES), artificial neural network (ANN), genetic algorithm (GA), evolutionary computation (EC), fuzzy logic, etc. have been emerged in recent years in power systems applications as effective tools. These methods are also known as artificial intelligence (AI) in several works. In a practical power system, it is very important to have the human knowledge and experiences over a period of time due to various uncertainties, load variations, topology changes, etc. This section presents the overview of CI/AI methods (ANN, GA, fuzzy systems, EC, ES, ant colony search, Tabu search, etc.) used in power system applications. 3.1 Artificial Neural Networks An artificial neural network (ANN) is an information-processing paradigm that is inspired by the biological nervous systems, such as the brain, process information (Bishop, 1995). The key element of this paradigm is the novel structure of the information processing system composed of a large number of highly interconnected processing elements (neurons) working in unison to solve the specific problems. ANNs, like people, learn by example. The starting point of ANN application was the training algorithm proposed and demonstrated, by Hebb in 1949, how a network of neurons could exhibit learning behaviour. During the training phase, the neurons are subjected to a set of finite examples called training sets, and the neurons then adjust their weights according to certain learning rule. ANNs are not programmed in the conventional sense, rather they learn to solve the problem through interconnections with environment. Very little computation is carried out at the site of individual node (neuron). There is no explicit memory or processing locations in neural network but are implicit in the connections between nodes. Not all sources of


3

input feeding a node are of equal importance. It all depends on weight which can be negative or positive. Inputs arriving at a node are transformed according to the activation function of node. The main advantages of ANNs are as follows: ANNs with suitable representation can model any degree of non-linearity and thus, are very useful in solving the non-

linear problems. These do not require any apriori knowledge of system model. ANNs are capable of handling situations of incomplete information, corrupt data and they are fault tolerant. Massive parallelism provided by the large number of neurons and the connections amongst them provide good search and

distributed representations. ANN is fast and robust. It possesses learning ability and adapts to the data.

Though the neural network (NN) training is generally computationally expensive, it takes negligible time to evaluate correct outcome once the network has been trained. Despite the advantages, some disadvantages of the ANN are: (i) large dimensionality, (ii) selection of the optimum configuration, (iii) choice of training methodology, (iv) the black-box representation of ANN they lack explanation capabilities and (v) the fact that results are always generated even if the input data are unreasonable. Another drawback of neural network systems is that they are not scalable i.e. once an ANN is trained to do certain task, it is difficult to extend for other tasks without retraining the NN. Artificial neural networks are most promising method for many power system problems and have been used for several applications. ANNs are mainly categorized by their architecture (number of layers), topology (connectivity pattern, feed forward or recurrent, etc.) and learning regime. Based on the architecture ANN model may be single-layer ANN which includes perceptron model (suggested by Rosenblot, in 1959) and ADALINE (suggested by Widrow & Hoff in 1960). ANN model can be further categorized as Feed forward NN and Feed Backward NN based on neuron interactions. Learning of ANN may be Supervised Learning, Unsupervised Learning, and Reinforcement Learning. Based on neuron structure ANN model may be classified as multilayer perceptron model, Boltzman machine, cauchy machine, Kohonen self-organizing maps, bidirectional associative memories, adaptive resonance theory-I (ART-1), adaptive resonance theory-2 (ART-2), counter propagation ANN. Some other special ANN models are parallel self-hierarchical NN, recurrent NN, radial basis function NN, knowledge based NN, hybrid NN, wavelet NN, cellular NN, quantum NN, dynamic NN, etc. 3.2 Genetic Algorithms Genetic algorithm (GA) is an optimization method based on the mechanics of natural selection and natural genetics. Its fundamental principle is the fittest member of population has the highest probability for survival. The most familiar conventional optimization techniques fall under two categories viz. calculus based method and enumerative schemes. Though well developed, these techniques possess significant drawbacks. Calculus based optimization generally relies on continuity assumptions and existence of derivatives. Enumerative techniques rely on special convergence properties and auxiliary function evaluation. The genetic algorithm, on the other hand, works only with objective function information in a search for an optimal parameter set. The GA can be distinguished from other optimization methods by following four characteristics.

(i) The GA works on coding of the parameters set rather than the actual parameters. (ii) The GA searches for optimal points using a population of possible solution points, not a single point. This is an important

characteristic which makes GA more powerful and also results into implicit parallelism. (iii) The GA uses only objective function information. No other auxiliary information (e.g. derivatives, etc.) is required. (iv) The GA uses probability transition rules, and not the deterministic rules.

Genetic algorithm is essentially derived from a simple model of population genetics. It has five following components:

Chromosomal representation of the variable characterizing an individual. An initial population of individuals. An evaluation function that plays the role of the environment, rating the individuals in terms of their fitness that is

their aptitude to survive. Genetic operators that determine the composition of a new population generated from the previous one by a mechanism

similar to sexual reproduction. Values for the parameters that the GA uses.

The advantages of GA over traditional techniques are as follows:

It needs only rough information of the objective function and puts no restriction such as differentiability and convexity on the objective function.

The method works with a set of solutions from one generation to the next, and not a single solution, thus making it less likely to converge on local minima.


4

The solutions developed are randomly based on the probability rate of the genetic operators such as mutation and crossover as the initial solutions would not dictate the search direction of GA.

Major disadvantage of GA method is that it requires tremendously high computational time in case of the large variables and

constraints. The treatment of equality constraints is also not well established in G.A. Alander (1996) has presented a bibliography of genetic algorithm in the power systems. GA has been widely used in the power system. There are several versions of GA available for different applications. 3.3 Fuzzy Logic Fuzzy logic (FL) was developed by Zadeh (Zadeh, 1965) in 1964 to address uncertainty and imprecision, which widely exist in the engineering problems. FL was first introduced in 1979 for solving power system problems. Fuzzy set theory can be considered as a generalization of the classical set theory. In classical set theory, an element of the universe either belongs to or does not belong to the set. Thus, the degree of association of an element is crisp. In a fuzzy set theory, the association of an element can be continuously varying. Mathematically, a fuzzy set is a mapping (known as membership function) from the universe of discourse to the closed interval [0, 1]. Membership function is the measure of degree of similarity of any element in the universe of discourse to a fuzzy subset. Triangular, trapezoidal, piecewise-linear and Gaussian functions are most commonly used membership functions. The membership function is usually designed by taking into consideration the requirement and constraints of the problem. Fuzzy logic implements human experiences and preferences via membership functions and fuzzy rules. Due to the use of fuzzy variables, the system can be made understandable to a non-expert operator. In this way, fuzzy logic can be used as a general methodology to incorporate knowledge, heuristics or theory into controllers and decision makers. The advantages of fuzzy theory are as follows:

(i) It more accurately represents the operational constraints of power systems and (ii) Fuzzified constraints are softer than traditional constraints.

Momoh et al. (2000) have presented the overview and literature survey of fuzzy set theory application in power systems. A recent survey shows that fuzzy set theory has been applied mainly in voltage and reactive power control, load and price forecasting, fault diagnosis, power system protection/relaying, stability and power system control, etc.

3.4 Evolutionary Computation: Evolutionary Strategies and Evolutionary Programming Natural evolution is a hypothetical population-based optimization process. Simulating this process on a computer results in stochastic optimization techniques that can often perform better than classical methods of optimization for real-world problems. Evolutionary computation (EC) is based on the Darwins principle of survival of the fittest strategy. An evolutionary algorithm begins by initializing a population of solutions to a problem. New solutions are then created by randomly varying those of the initial population. All solutions are measured with respect to how well they address the task. Finally, a selection criterion is applied to weed out those solutions, which are below standard. The process is iterated using the selected set of solutions until a specific criterion is met. The advantages of EC are adaptability to change and ability to generate good enough solutions but it needs to be understood in relation to computing requirements and convergence properties. EC can be subdivided into GA, evolution strategies, evolutionary programming (EP), genetic programming, classified systems, simulated annealing (SA), etc. The first work in the field of evolutionary computation was reported by Fraser in 1957 (Fraser, 1957) to study the aspects of genetic system using a computer. After some time, a number of evolutionary inspired optimization techniques were developed. Evolution strategies (ES) employ real-coded variables and, in its original form, it relied on mutation as the search operator and a population size of one. Since then, it has evolved to share many features with GA. The major similarity between these two types of algorithms is that they both maintain populations of potential solutions and use a selection mechanism for choosing the best individuals from the population. The main differences are:

ESs operates directly on floating point vectors while classical GAs operate on binary strings, GAs rely mainly on recombination to explore the search space while ES uses mutation as the dominant operator and ES is an abstraction of evolution at individual behavior level, stressing the behavioral link between an individual and its

offspring, while GA maintains the genetic link. Evolutionary programming (EP), which is a stochastic optimization strategy similar to GA, places emphasis on the behavioral linkage between parents and their offspring, rather than seeking to emulate specific genetic operators as observed in nature. EP is similar to evolutionary strategies, although the two approaches were developed independently. Like ES and GA, EP is a useful method of optimization when other techniques such as gradient descent or direct analytical discovery are not possible. Combinatorial and real-valued function optimizations, in which the optimization surface or fitness landscape is rugged and possessing many locally optimal solutions, are well suited for evolutionary programming.


5

3.5 Simulated Annealing This method was independently described by Scott Kirkpatrick, C. Daniel Gelatt and Mario P. Vecchi in 1983 (Kirkpatrick et al., 1983), and by Vlado Cerny in 1985 (Cerny, 1985). Based on the annealing process in the statistical mechanics, the simulated annealing (SA) was introduced for solving complicated combinatorial optimization problems. In a large combinatorial optimization problem, an appropriate perturbation mechanism, cost function, solution space, and cooling schedule are required in order to find an optimal solution with simulated annealing. SA is effective in network reconfiguration problems for large-scale distribution systems and its search capability becomes more significant as the system size increases. Moreover, the cost function with a smoothing strategy enables the SA to escape more easily from local minima and to reach rapidly to the vicinity of an optimal solution. The advantages of SA are its general applicability to deal with arbitrary systems and cost functions; its ability to refine optimal solution; and its simplicity of implementation even for complex problems. The major drawback of SA is repeated annealing. This method cannot tell whether it has found optimal solution or not. Some other methods (e.g. branch and bound) are required to do this. SA has been used in various power system applications like transmission expansion planning, unit commitment, maintenance scheduling, etc. 3.6 Expert Systems AI programs that achieve expert-level competence in solving the problems by bringing knowledge about specific tasks are called knowledge-based or expert systems (ES) which was first proposed by Feigenbaum et al. in the early 1970s (Feigenbaum et al, 1971). ES is a knowledge-based or rule based system, which uses the knowledge and interface procedure to solve problems that are difficult enough to be solved by human expertise. Main advantages of ES are:

(a) It is permanent and consistent (b) It can be easily transferred or reproduced (c) It can be easily documented.

Main disadvantage of ES is that it suffers from a knowledge bottleneck by having inability to learn or adapt to new situations. The knowledge engineering techniques started with simple rule based technique and extended to more advanced techniques such as object-oriented design, qualitative reasoning, verification and validation methods, natural languages, and multi-agent systems. For the past several years, a great deal of ES applications has been developed to prepare plan, analyze, manage, control and operate various aspects of power generation, transmission and distributions systems. Expert system has also been applied in recent years for load, bid and price forecasting. 3.7 Ant Colony and Tabu Search Dorigo introduced the ant colony search (ACS) system, first time, in 1992 (Dorigo, 1992). ACS techniques take inspiration from the behavior of real ant colonies and are used to solve functional or combinational problems. ACS algorithms to some extent mimic the behavior of real ants. The main characteristics of ACS are positive feedback for recovery of good solutions, distributed computation, which avoids premature convergence, and the use of a constructive heuristic to find acceptable solutions in the early stages of the search process. The main drawback of the ACS technique is poor computational features. ACS technique has been mainly used in finding the shortest route for transmission network. Tabu search (TS) is basically a gradient-descent search with memory. The memory preserves a number of previously visited states along with a number of states that might be considered unwanted. This information is stored in a Tabu list. The definition of a state, the area around it and the length of the Tabu list are critical design parameters. In addition to these Tabu parameters, two extra parameters are often used such as aspiration and diversification. Aspiration is used when all the neighboring states of the current state are also included in the Tabu list. In that case, the Tabu obstacle is overridden by selecting a new state. Diversification adds randomness to this otherwise deterministic search. If the Tabu search is not converging, the search is reset randomly. TS is an iterative improvement procedure that starts from some initial solution and attempts to determine a better solution in the manner of a greatest descent neighborhood search algorithm. Basic components of TS are the moves, Tabu list and aspiration level. TS is a metahuristic search to solve global optimization problem, based on multi-level memory management and response exploration. TS has been used in various power system application like transmission planning, optimal capacitor placement, unit commitment, hydrothermal scheduling , fault diagnosis/alarm processing, reactive power planning, etc. 3.8 Particle Swarm Optimization The particle swarm optimization (PSO) method introduced by Kennedy and Eberhart (Kennedy et al., 1995) is a self-educating optimisation algorithm that can be applied to any nonlinear optimisation problem. In PSO, the potential solutions, called particles,


6

fly through the problem space by following the best fitness of the particles. It is easily implemented in most programming languages and has proven to be both very fast and effective when applied to a diverse set of optimization problems. In PSO, the particles are flown through the problem space by following the current optimum particles. Each particle keeps the track of its coordinate in the problem space, which is associated with the best solution (fitness) that it has achieved so far. This implies that each particle has memory, which allows it to remember the best position on the feasible search space that has ever visited. This value is commonly called as pbest . Another best value that is tracked by the particle swarm optimizer is the best value obtained so far by any particle in the neighborhood of the particle. This location is commonly called as .gbest The position and velocity vectors of the ith particle of a d-dimensional search space can be represented as ),.....,( 21 idiii xxxX = and ),.....,( 21 idiii vvvV = respectively. On the basis of the value of the evaluation function, the best previous position of a particle is recorded and represented as ),......,( 21 idiii PPPpbest = . If the thg particle is the best among all particles in the group so far, it is represented as 1 2( , ,.. )g g g gdgbest pbest P P P= . The particle tries to modify its position using the current velocity and the distance from pbest and gbest . The modified velocity and position of each particle for fitness evaluation in the next iteration are calculated using the following equations

)()( 2211``1 k

idgdkidid

kid

kid xgbestrandcxpbestrandcvwv ++=+ (1)

`11 ++ += kidkidkid vxx (2) where, w is the inertia weight parameter, which controls the global and local exploration capabilities of the particle. 21,cc are cognitive and social coefficients and 1rand and 2rand are random numbers between 0 and 1. A large inertia weight factor is used during initial exploration and its value is gradually reduced as the search proceeds. The concept of time-varying inertial weight (TVIM) is given by

minmax

maxminmax )( witer

iteriterwww += (3) where maxiter is the maximum number of iterations. The velocity update expression (1) can be explained as follows. Without the second and third terms, the first term (representing inertia) will keep a particle flying in the same direction until it hits the boundary. Therefore, the first term tries to explore new areas and corresponds to the diversification in the search procedure. In contrast, without the first term, the velocity of the flying particle is only determined by using its current position and its best positions in history. Therefore, the second representing memory) and third terms (representing cooperation) try to converge the particles to their Pbest and/or Gbest and correspond to the intensification in the search procedure. 3.9 Support Vector Machines Support vector machine (SVM) is one of the relatively new and promising methods for learning, separating functions in pattern recognition (classification) tasks as well as performing function estimation in regression problems. It is originated from supervised learning systems derived from the statistical learning theory introduced by Vapnik for distribution free learning from data (Vapnik, 1998). In this method, the data are mapped into a high dimensional space via a nonlinear map, and using this high dimensional space, an optimal separating hyper-plane or linear regression function is constructed. This process involves a quadratic programming problem and produces a global optimal solution. The great advantage of SVM approach is that it greatly reduces the number of operations in the learning mode and minimizes the generalization error on the test set under the structural risk minimization (SRM) principle. Main objective of the SRM principle is to choose the model complexity optimally for a given training sample. The input space in a SVM is nonlinearly mapped onto a high dimensional feature space. The idea is to map the feature space into a much bigger space so that the boundary is linear in the new space. SVMs are able to find non-linear boundaries if classes are linearly non-separable. Another important feature of the SVM is the use of kernels. Kernel representations offer an alternative solution by nonlinearly projecting the input space onto a high dimensional feature space. The advantage of using SVMs for classification is the generalization performance. SVM performs better than neural networks in term of generalization. There is problem of over fitting or under fitting if so many training data or too few training data are used. The computational complexity is other factor for the choice of SVMs as classifier. The other advantage of SVM based system is that it is straight forward to extend the system when new types of cases are added to the classifier. 4. Conclusion The current interest in using the computation intelligence (CI) for power system applications is increasing amongst the researchers and academicians. It is noticed that huge amount of research papers and articles are available in all the area of


7

engineering and science. There is no specific guideline for the application of CI techniques in the power systems. In this paper, various computational techniques widely used in power system applications are briefly described. The potential areas of CI application are also highlighted. New intelligent system technologies using digital signal processing techniques, expert systems, artificial intelligent and machine learning provide several unique advantages in solving the power system problems. Acknowledgement Authors acknowledge the suggestions and corrections made by Dr Guang Ya Yang, Centre of Electric Technology, Electrical Engineering Department, Denmark Technical University, Denmark. Dipti Saxena is grateful to the CM and Director, Invertis Institute of Engineering and Technology, Bareilly for permitting her to carry our PhD work from UP Technical University, Lucknow, India. References Alander J. T., 1996, An indexed bibliography of genetic algorithm in power engineering, Power Report Series 94-1. Bishop C.M., 1995, Neural networks for pattern recognition, Oxford University Press, Oxford. Cerny V., 1985, A thermodynamical approach to the travelling salesman problem: an efficient simulation algorithm. Journal of

Optimization Theory and Applications, vol.45, pp.41-51. Dorigo M., 1992, Optimization, learning and natural algorithms, PhD thesis, Politecnico di Milano, Italy. El-Hawary, Mohamed E., 1998, Electric power applications of fuzzy systems, John Wiley USA. Feigenbaum, E.A., Buchanan, B.G., Lederberg J., 1971, On generality and problem solving: A case study using the DENDRAL

program, in Machine Intelligence 6, Edinburgh University Press. Fogel L.J., Owens A.J., Walsh M.J., 1966, "Artificial intelligence through simulated evolution", Wiley, New York Fraser A. S. 1957, Simulation of genetic systems by automatic digital computers, I. Introduction.Australian. J. Biol. Sci., vol.10,

pp.484491. Kennedy J., and Eberhart R., 1995, Particle swarm optimization, Proc. of the IEEE Int. Conf. on Neural Networks, Piscataway,

NJ, pp. 19421948. Kirkpatrick S., Gelatt C. D., Vecchi M. P., 1983, "Optimization by simulated annealing". Science. New Series 220, pp.671680. Lai, Loi Lei, 1998, Intelligent system applications in power engineering: evolutionary programming and neural networks, John

Willey & Sons, UK. Momoh James A., EL-Hawary Mohamed E., 2000, Electric systems, dynamics, and stability with artificial intelligence, Marcel

Dekker, Inc. USA. Sobajic Dejan J., 1993, Neural network computing for the electric power industry, Routledge Publisher, USA. Song Yong-Hua Song, Johns Allan, Aggrawal Raj,1996. Computation intelligence applications to power systems, Kluwer

Academic Publishers, USA. Vapnik N., 1998, Statistical learning theory, John Wiley and Sons, New York. Warwick K., Ekwue Arthur, Aggarwal Raj, 1997, Artificial intelligence techniques in power systems, IEE UK. Wehenkel Louis A., 1998, Automatic learning techniques in power systems, Kluwer academic publisher, USA. Zadeh L.A.,1965, Information and Control, vol. 8, no. 3, pp. 338-353. Biographical notes: D. Saxena obtained B. Tech. (Hons) in Electrical Engineering from KNIT Sultanpur (UP), India and M. Tech. in Process Control from Netaji Subhas Institute of Technology, New Delhi, India in 1999 and 2003, respectively. Presently, she is working as reader in Electrical and Electronics Engineering Department at Invertis Institute of Engineering And Technology, Bareilly. She is registered as PhD candidate in UP Technical University Lucknow, India. Her research interests are power quality, power electronics, control systems and DSP application in power. S. N. Singh received M. Tech. and Ph.D. from Indian Institute of Technology Kanpur, India in 1989 and 1995, respectively. He is a Professor in the Department of Electrical Engineering, Indian Institute of Technology Kanpur, India. Presently he is on leave from IIT Kanpur and working with the Centre for Electric Technology (CET), Technical University of Denmark, Denmark. His research interests include power system restructuring, power system optimization & control, voltage security and stability analysis, power system planning, and ANN application to power system problems. He is a Fellow of IE (India), Fellow of IETE (India) and senior member of IEEE. K.S. Verma received his B. Tech. (Hons) in Electrical Engineering and M. Tech. in Electrical Engineering (Power Systems) both from KNIT Sultanpur (India) in 1987 and 1998, respectively. He obtained his Ph.D. degree in Electrical Engineering (Power Systems) from Indian Institute of Technology, Roorkee, Roorkee (India) in 2003. Presently, Prof Verma is Director, KNIT Sultanpur (India). His research interests include FACTS Technology, distributed generation, power system optimization & control, power quality and AI application in power system. Received February 2010 Accepted March 2010

MultiCraft


Vol. 2, No. 3, 2010, pp. 8-18




Application of artificial neural networks to improve power transfer

capability through OLTC

A. Abu-Siada1*, S. Islam1 and E.A. Mohamed2

1*Department of Electrical and Computer Engineering, Curtin University of Technology, Perth, AUSTRALIA 2 Department of Electrical Power Engineering, Ain Shams University, Cairo, EGYPT

*Corresponding Author: e-mail: [email protected], Tel +61-8-92667287, Fax.+61-8-92662584

Abstract On load tap changing (OLTC) transformer has become a vital link in modern power systems. It acts to maintain the load bus voltage within its permissible limits despite any load changes. This paper discusses the effect of different static loads namely; constant power (CP), constant current (CI) and constant impedance (CZ) on the maximum power transfer limit from the generation to the load centre through the OLTC branch and in turn on the static stability margin of power systems. Then the paper introduces a novel approach for the on-line determination of the OLTC settings using artificial neural network (ANN) technique in order to improve the power transfer capability of transmission systems. The proposed approach is tested on a six-bus IEEE system. Numerical results show that the setting of OLTC transformer in terms of the load model has a major effect on the maximum power transfer in power systems and the proposed ANN technique is very accurate and reliable. The adaptive settings of OLTC improve the power transfer capability according to the system operating condition. Keywords: OLTC, Static Load, Voltage Stability, ANN. 1. Introduction Modern power systems are being stressed with the continuous growth of load requirements. Power system operator should ensure the quality and reliability of supply to the loads by maintaining the load bus voltages within their permissible limits. The power transfer from generators to load centers affects load bus voltages. In addition to restoring the load, OLTC also extend the power transfer ability to the load center (Zhu et al., 2000). It is well known that the operation of OLTC has a significant influence on voltage stability. Most literature has concentrated on the contribution of OLTC to voltage stability (Tylor 1994, Afzalian et al. 2008). The power transfer capability is measured by the margin from a base case operating point to the operation-limiting boundary, measured along the load growth trajectory. The static load models have an impact on the power transfer capability from generating station to the load center and thus on the static stability margin of power systems (Abu-Siada et al. 2008, Feng et al. 2004). The paper considers first the effect of different static load models namely; constant power (CP), constant current (CI) and constant impedance (CZ) load models on the power transfer limits to the load centre. Then the paper presents a novel approach for evaluating the on-line optimal settings of OLTC transformer that corresponds to the maximum power transfer to the load and hence to improve the static stability limits using artificial neural network (ANN). 2. Effect of OLTC Setting on Power Transfer Assume the simple power system shown in Figure 1. The load power is supplied from an infinite bus via OLTC. The load flow equation at load bus is given by:

c

L

LT

L

jXV

VjQP

XjnVnE

+= *2 (1)

Abu-Siada et al. / International Journal of Engineering, Science and Technology, Vol. 2, No. 3, 2010, pp. 8-18

9

where 0jEE += is the emf of the equivalent voltage source, irL jVVV += is the load bus voltage, treqT XXX += is the equivalent reactance of power system & transformer, P and Q are the load active and reactive power and Xc is the reactance of the shunt capacitor compensated load.

Figure. 1 Simplified power system

2.1 Constant Power (CP) Load: For constant Power load;

Kcc

PQLet

cQandcP

q

p

qp

====

Equation (1) can be re-written as below: jKPP

XXnXnX

VVjnXEV

nXEVj

cT

Tcir

T

i

T

r =++ 22

22 )( (2)

Equating real parts in both sides, P

EnXV Ti = (3)

Equating imaginary parts in both sides and substituting for Vi from (3), 0)( 2

22

22

2

22

=++ rTcc

rTC

CTT VXnX

EnXVP

XnXXKXn

PEXn (4)

Solving for P,

)()())(2

()(2 2

2222

2

2

2

rTc

cr

TTcT

c

TcT

c VXnX

EnXVnXE

XnXXEKX

XnXXEKXP +=

(5)

This solution is subject to:

T

c

rTc

cr

TTcT

C

XXn

and

VXnX

EnXVnX

EXnXX

EKX

0)()())(2( 2222

2

2

f (6)

From equation (6), one can conclude that there is a limit to the power transfer to the load depends on the degree of compensation, OLTC setting and load bus voltage. Figure 2 shows the effect of OLTC tap ratio on the maximum power transfer for compensated and uncompensated load (Xc=10 pu. and Xc= respectively). For both cases the maximum power transfer limit is increased by increasing the tap ratio. Once the optimal power is reached (0.5 pu. in case of uncompensated load and 0.56 pu. in case of compensated load) the power slightly decreases with the increase of tap ratio. It should be noted that for the compensated load, the power is dramatically increased when the resonant case is reached ( 472.4==

T

c

XX

n in this case).

Figure 2. Effect of OLTC on power transfer to the CP load (K=0.75, XT=0.5 pu., Vr=0.8 pu.)

Rearranging equation (4) and solving for Vr,


10

)())(2

()(2 22

2222 PE

XXnX

KXPXn

XnXEnX

XnXEnX

V TTc

cT

Tc

c

Tc

cr +=

(7)

Figure 3 shows the PV curve for the uncompensated load for different tap ratios. The figure shows that the maximum power limit is constant and equal 0.5 pu. in all cases.

Figure 3. P-V curve at different tap ratios (Xc=)

The PV curves in Figures 4 and 5 show that the maximum power transfer limit to a compensated load can be increased by either increasing the tap ratio of OLTC or by increasing the degree of compensation. The power increase is attributed to the fact that the OLTC tap settings allow the match between the network impedance and the reflected compensated load impedance.

Figure 4. P-V curve at different tap ratios (Xc=1 pu)

Figure 5. P-V curve at different load compensation (n=1)

2.2 Constant Current (CI) Load: For constant current load;

KPQVcQVcP LqLp === ,

Then equation (1) can be re-written as: )1()( 2

222 jKVc

XXnXnXVVj

nXEV

nXEVj Lp

cT

Tcir

T

i

T

r =++ (8)

Equating real parts in both sides,

r

pT

pTi

irpTi

LpT

i

VcnXE

cnXV

VVcnXVE

VcnXEV

22

22222

)(

)()(

=

+==

(9)


11

Equating imaginary parts and substituting for Vi

2)()()(

)( 22

2

2

22

pTTc

pcTc

Tc

pTr cnXEXnXE

KcXXnnX

XnXEcnXE

V = (10)

22

22

22222

)(

))(

)(1(

pT

rL

pT

pTrirL

cnXEEVV

cnXEcnX

VVVV

=

+=+= (11)

Tc

pTcpT

Tc

cpLp XnX

KcXXncnXE

XnXXnc

VcP 222

222 )( ==

(12)

This solution is subject to:

T

c

pT

pT

XX

n

and

cXEn

cnXE

p

f 0)( 22 (13)

Figure 6. Effect of OLTC on power transfer to the CI load

Figure 6 shows the effect of OLTC tap ratio on the power transfer limit to constant current load. The maximum power transfer limit is increasing with the increase of the tap ratio. When the maximum limit is reached (0.5 pu. in case of uncompensated load and o.62 pu. in case of compensated load), the power is decreasing with the increase of tap ratio. The shunt capacitor increases the power transfer limit and shifts the optimal setting of OLTC to a higher value. 2.3 Constant Impedance (CZ) Load: For constant impedance load;

KPQVcQVcP LqLp === 22 (14)

Then equation (1) can be re-written as: )1()( 2

222 jKP

XXnXnXVVj

nXEV

nXEVj

cT

Tcir

T

i

T

r =++ (15)

Equating the real parts in both sides,

E

PnXV Ti = (16) Equating imaginary parts,

PXXnXnXKc

EcnXV

cT

Tcp

p

Tr )( 2

2+= (17)

p

T

cT

Tcp

p

TirL c

PPE

nXPXXnXnXKc

EcnXVVV =++=+= 22222

22222 )()()(

222

)()(

/1

EnX

EKnX

EXncXnX

cP

TT

cp

Tc

p

++= (18)

Figure 7 shows the effect of OLTC tap ratio on the power transfer limit to constant impedance load. Again, the power is increasing with the increase of OLTC tap ratio till the optimal power is reached then the power is decreasing with the increase of tap ratio. The optimal power in case of uncompensated load (Xc= ) is 0.5 pu. corresponding to 1.72 tap ratio. However, the maximum power for a compensated load (Xc= 10 pu.) is 0.62 pu. corresponding to 2.1 tap ratio. It should be noted that in all load models (CP, CI and CZ) the power transfer limit is increased by increasing the degree of compensation i.e. decreasing the value of Xc. When the load power cannot be met with the increasing compensation degree, the power can be increased by increasing the OLTC tap ratio.


12

Figure 7. Effect of OLTC on power transfer to the CZ load

3. Digital Simulation Figure 8 shows a one-line diagram of the IEEE 6-bus test system. The system consists of 6-nodes, 7 lines, two OLTC transformers, two generators and four loads (Tamura et al. 1993). The two OLTCs are varied one at a time, during the calculation of the maximum power transfer to the load centers.

Figure 8. One line diagram of the system under study

The results obtained using the conventional methodology cover a wide range of operating conditions. A sample of the simulation results is shown in Figures 9-14. These results are presented to explore the effect of both OLTC settings and the load model on the maximum power transfer. Figures 9-11 show the relation between the power transfer to the three load buses (1, 2& 4) and the transformer tap ratio n under loading condition PL = QL= 0.35 pu.(pf= 0.707 lag), for the three different load models (CP, CI & CZ). From these figures, it can be seen that the optimum value of n and the corresponding maximum power transfer to the loads are different for the three load models. Figures 12-14 present similar results at another loading condition PL = 0.8 pu. and QL = 0.13 pu. (pf = 0.987 lag). It can be seen from these figures that the maximum power transfer limit is higher at this operating condition than the previous case and the OLTC setting n1 has a strong effect on the load at bus #1 while the OLTC setting n2 has a stronger impact on the load at bus #2. On the other hand, the OLTC setting n2 has a unique correlation and effect on the load at bus #4. These figures prove that the power transfer limit is affected by the OLTC setting for all load models. Also, Results show that before the reverse points are reached, the upward of OLTC operation increases the maximum power transfer to the load centre. Also these results give an indication that the maximum point of power transfer and thus the levels of these powers are different for different load models. It must be noted here that, in the constant impedance (CZ) load model, the nonlinear load behavior is very sensitive to any system variation near to the static stability limit. That can be depicted from the sharp peaks indicated on Figures 10 and 14. In case of constant power (CP) load model, the load power is constant however, the power transfer to the load bus through the OLTC branch is maximized and the other shares through other branches are varied.


13

Figure 9. Effect of OLTC (n1) on Power transfer to load 1





14



3. Determination of OLTC Setting using ANN Owing to the fast development in computing systems, application of intelligent systems such as artificial neural network (ANN) and fuzzy logic have paid a great attention in power system applications (Abu-Siada et al. 2009, Nor et al. 2003). Fuzzy logic needs a prior experience about the system to enable designing fuzzy logic rules based on the input output behaviors. ANN is selected in this paper because of its capability of learning a tremendous variety pattern mapping relationships without having prior knowledge of a mathematical function. ANN is able to handle complex non-linear problems and it overcomes the complex tedious calculation problems. The enhancement of power transfer and voltage stability can be achieved in real-time using the proposed method without any additional cost involved for installing capacitor banks and the related switches (El-Keib et al. 1999). The developed ANN model can be implemented using the currently available neural chips. The main drawback of ANN is the extensive training data required to ensure the reliability of the network.

Figure 15. Typical ANN architecture model


15

3.1 Topology of the ANN: The proposed ANN scheme uses a multi-layer feed-forward ANN which consists of an input layer, a hidden layer and an output layer as shown in Figure 15 (El-Sharkawi 1991, Kashem et al. 2001).

3.1.1 Input layer: Appropriate selection of input variables is the key to the success of the ANN application. Usually heuristic knowledge is required in choosing the appropriate variables. The power transfer from generation to the load centre is mainly concerned with the transformer OLTC settings. The optimal setting of the OLTC is affected by the load level. Therefore, these loading conditions are chosen as inputs to the ANN. These inputs are the load active and reactive power in addition to the generation active and reactive power at each bus in the system.

3.1.2 Hidden layer: The computational power of the ANN increases with the addition of hidden layers. There are no general guidelines to determine the number of hidden layers and number of neurons per layer. Many applications have proved that ANN with single hidden layer has sufficient capability of capturing complicated relations between input and output variables. In this paper, one hidden layer with five neurons was found satisfactory for estimating the optimal transformer OLTC settings.

3.1.3 Output layer: Determination of the output layer is quite forward. Two output neurons were used for setting the OLTC of the two transformers. The sigmoidal transfer functions have been used for all the neurons as they are suitable for any nonlinear mapping of input-output combinations because they have noise immunity for low inputs, normal outputs for middle range inputs and saturation for large inputs. Appropriate scaling of the input and output variables were carried out in the (0, 1) range. 3.2 Training Scenario: The scenario of the training process was started by initiating the ANN topology; i.e. the number of input nodes, the number of hidden layers, the number of neurons in different hidden layers, the number of neurons in the output layer, the type of hidden and output activation functions, and the number of presentation cycles (epochs). It must be noted, that, by adjusting the number of hidden layers and the number of simulated neurons within each layer, the performance of an ANN can be either enhanced or degraded. There are no general guidelines for a priori knowledge of which ANN architecture would perform the best for a given application. The researchers have experimented with different architectures to find out the most suitable configuration. The results obtained from the digital simulation of the system in Figure 8 were used in training and testing the proposed ANN. Back-propagation algorithm was used to train the proposed ANN using 100 training patterns with a learning rate of 0.96. The Neural Network toolbox in MATLAB is used to implement the proposed ANN. The network weights and biases initialized using random values. During training, the weights and biases are adjusted to minimize the error between the network outputs and the targets. It has been found from the simulation results that the OLTC settings and the corresponding maximum power transfer to load centers are varying with the load type. Therefore, different ANNNs have been trained for different load models. 3.3 Testing Scenario: For validating the ANN performance the scenario of the testing process is started once the training process is completed and a set of new input variables which were not used for training the ANN model are applied to the designed ANN. The proposed method has been applied to the system shown in Figure 8 and the ANN results are compared with the simulation results. The obtained results for this scenario are recorded in Tables 1, 2 and 3. 4. Results and Discussions The three load models under consideration are tested on different operating conditions. The transformers turns ratios are varied one at a time. The results in Table 1 have been obtained for the constant power (CP) load model. The table shows that the maximum testing absolute errors are: 5.06% (n1) & 5.623% (n2). Similarly, Tables 2 & 3 show these errors for the constant current (CI) and the constant impedance (CZ) load models, respectively. The maximum errors obtained in this case are: 6.0% (n1), 3.94% (n2) and 3.66% (n1), 3.49% (n2), respectively. Also tables present the average and the standard deviations for the three different load models. These deviations are: 1.092/1.731 (n1), 1.762/2.215 (n2) for CP load; 1.499/2.375 (n1), 1.275/1.636 (n2) for CI model; and 1.66/1.474 (n1), 0.825/1.19 (n2) for CZ model. From these results, it can be seen that the CP load and the CI load models have nearly the same error level. While, the CZ load model has the less error level. The ANN is efficient for the prediction of the optimal value of transformer turns ratio settings corresponding to the maximum power transfer to the load centre. 5. Conclusions An efficient method for the on line prediction of OLTC transformer settings giving maximum power transfer to the load centre is analysed. The proposed technique has been tested on the IEEE Six-bus power system. Numerical results show that the operation of OLTC transformer has a major effect on the maximum power transfer and thus on the static stability margin. Static load models such as constant power, constant current and constant impedance have a great effect not only on the maximum power transfer but also on the optimal settings of transformer OLTCs. The Artificial neural network (ANN) proved to be an efficient tool for the prediction of the optimum value of OLTC transformer settings corresponding to the maximum power transfer. The developed ANN model can be implemented using the currently available neural chips.


16

Table 1. Results of ANN in case of constant power (CP) load model

Training Scenario Testing Scenario Predicted Values

Desired Values

%Error Predicted Values

Desired Values

% Error

n1 n2 n1 n2 n1 n2 n1 n2 n1 n2 n1 n2 .9983 .9985 .997

.9985

.9981

.9975

.9985

.9983

.9975

.9983

.9972

.9972

.9982 .998

.9962

.9943

.9952

.8988

.9933

.9852

.9202 .994

.9909

.9339

.9898

.9324

.8942

.9914

.9839

.8411

1 1 1 1 1 1 1 1 1

.98 1 1

.98 1 1

1 1.02 .9 1 1

.93 1 1

.93 1

.93 .9 1

.95

.85

-.17 -.15 -.30 -.15 -.19 -.25 -.15 -.17 -.25

1.867 -.28 -.28

1.857 -.2 -.38

-.57 -2.48 -.133 -.670 -1.48 -1.05 -.6 -.91 .419 -1.02 .258 -.644 -.86 3.58 1.07

.9983

.9982

.9965

.9984

.9976

.9974

.9984

.9981

.9971

.9981 .997 .9968 .9981 .9978 .9963

.993

.992

.852

.991

.966

.917

.993

.989

.897

.986

.919

.843

.982

.983

.873

.98 1 1

.98 1 1

.99 1 1

.95 1 1

.96 1 1

1 .98 .85 1

.95 .9 1

.98

.88 1 .9

.85 1

.93

.83

1.86 -.18 -.35 1.88 -.24 -.26 .899 -.19 -.29 5.06 -.3 -.32 3.99 -.22 -.37

-.77 1.143 .897 -.9

1.684 1.888 -.670 -.908 2.125 -1.54 1.544 -.318 -1.08 5.623 5.337

Average deviation Standard deviation

.443 .7368

1.047 1.361

1.092 1.71

1.762 2.215

Table 2. Results of ANN in case of constant Impedance (CI) load model


Desired Values

% Error Predicted Values

Desired Values

% Error

n1 n2 n1 n2 n1 n2 n1 n2 n1 n2 n1 n2 .9703 .9843 .9927 .9489 .9910 .9931 .9483 .9878 .9931 ,9581 .9952 ,9947 .9579 .9849 .9929

.9915

.9406

.8067

.9920

.9440

.8606

.9922

.9458

.8328

.9876

.7850

.8032

.9876

.9171

.7577

.98 1 1

.95 1 1

.95 1 1

.93

.78 1

.93 1 1

1 .95 .80 1

.93

.85 1

.95

.83 1 1

.78 1

.90

.75

-.989 -1.57 -.73

-.116 -.90 -.69

-.179 -1.22 -.69 3.02 -.48 -.53 3.0

-1.51 -.71

-.85 -.989 .838 -.80 1.51 1.24 -.78

-.442 .337 -1.24 .641 .40

-1.24 1.9

1.027

.9456

.9862

.9935

.9453

.9924

.9939

.9466

.9893

.9943

.9531

.9958

.9953 ,9540 .9869 .9947

.9929

.9275

.7767

.9929

.9252

.8383

.9926

.9342

.7923

.9897 ,7319 ,7753 .9899 .8990 .6724

.95 1 1

.93 1 1

.94 1 1 .9 1 1 .9 1 1

1 .93 ,78 1 .9 .83 1

.93 .8 1

.75

.78 1

.88

.70

-.463 -1.38 -.65

1.645 -.76 -.61

-.702 -1.07 -.57 5.9 -.42 -.47 6.0

-1.31 -.53

-.71 -.269 -.33 -.71 2.8 1

-.74 .45

-.963 -1.03 -2.41 -.602 -1.01 2.159 -3.94


1.225 1.563

.9489 1.067

1.499 2.375

1.275 1.636


17

Table 3. Results of ANN in case of constant current (CZ) load model


Desired Values

% Error Predicted Values

Desired Values

% Error

n1 n2 n1 n2 n1 n2 n1 n2 n1 n2 n1 n2 .9444 .9955 .9965 .9872 .9980 .9982 .9750 .9975 .9978 .7996 .9979 .9972 .8527 .9963 .9951

.9956

.9494

.8510

.9856

.7494 .785

.9912

.8291

.7706

.9984

.7996

.9343

.9977

.8299

.8023

.95 1 1

.98 1 1

.97 1 1 .8 1 1

.85 1 1

1 .95 .85 1

.75

.78 1

..83 .78 1 .8

.93 1

.83 .8

-.59 -.45 -.35

-.735 -.2 -.18 .515 -.25 -.22 -.05 -.21 -.28 .318 -.37 -.49

-.44 -.06 .118 -1.44 -.08 .641 -.88

-.108 -.94 -.16 -.05 .46 -.23 .012 .288

.9437

.9959

.9965 .964 .9978 .9981 .9695 .9975 .9977 .9844 .9978 .9974 .8226 .9965 .9954

.9948

.9251

.8280

.9919

.7437

.7498

.9910

.8018

.7407

.9651

.7883

.9052

.9978

.7965

.7821

.93 1 1

.93 1 1

.95 1 1

.79 1 1 .8 1 1

1 .93 .83 1

.73

.75 1 .8

.75 1

.78

.90 1

.80

.78

1.47 -.41 -.35 3.66 -.22 -.19

2.053 -.25

-1.56 1.823 -.22 -.26 2.83 -.35 -.35

-.52 -.527 -.24

-.810 1.877 -.27 -.90 .18

-1.24 -3-49 1.064 .578 -.22

-.438 .269


.349 .3836

.394

.547 1.66

1.474 .825 1.19

References Abu-Siada A., S. Islam and E. A. Mohamed, 2008. Adaptive Setting of OLTC to Improve Power Transfer Capability of Power

Systems, CMD08, Beijing, China, March. Abu-Siada A., S. Islam and S. Lai, 2008. A Novel Application of Gene Expression Programming in Transformer Diagnostics,

AUPEC08, Sydney, December. Abu-Siada A., S. Islam and S. Lai, 2009. Remnant Life Estimation of Power Transformer using Oil UV-Vis Spectral Response,

IEEE PES Power Systems Conference & Exhibition (PSCE), USA, March. Afzalian A., A. Saadatpoor and W. Wonham, 2008. Systematic supervisory solutions for under-load tap-changing transformers,

Control Engineering Practice, Vol. 16, pp1035-1054. Dilon T. S., 1991. Artificial Neural Network Applications to Symbolic Methods, Journal of Electrical Power and Energy System,

Vol. 13, No. 2, pp. 66-72. El-Keib A. and X. Ma, 1999. Application of Artificial Neural Networks in Voltage Stability Assessment, IEEE Transaction on

Power Systems, Vol. 10, Nov. EL-Sharkawi M. A., R, J. Marks and Siri Weerasooriya, 1991. Neural Networks and Their Application to Power Engineering,

control and Dynamics serial Academic Press. Feng Dong et al, 2004. Impact of Load Tab Changing Transformers on Power Transfer Capability, Electric Power Components

and Systems, Vol. 32, pp. 1331-1346. Gwang Won Kim and Kwang Y. Lee, 2005. Coordination Control of ULTC Transformer and STATCOM Based on an Artificial

Neural Network, IEEE Transactions on Power Systems, Vol. 20, No. 2, pp 580-586. Kashem M. A., V. Ganapathy and G. B. Jasmon, 2001. On-line Network Reconfiguration for Enhancement of Voltage Stability in

Distribution Systems using Artificial Neural Networks, Electrical Power Components and Systems, Vol. 29, pp. 361-373. Lai S., A. Abu-Siada, S. Islam. 2008. Correlation between UV-Vis Spectral Response and Furan Measurement of Transformer

Oil, CMD08, Beijing, China, March. Les M.Hajagos, and Behnam Danai, 1998. Laboratory Measurements and Models of Modern Loads and Their Effects on voltage

Stability Studies, IEEE Transaction on Power Systems, vol. 6, No.2 May, pp.584-591. Nor Haidar et al., 2003. Application of ANN to Determine the OLTC in Minimizing the Real Power Losses in Power System,

National Power and Energy Conference Proceddings, Malaysia. Qin Zhou et.al., 1994. Application Of Artificial Neural Networks In Power System Security And Vulnerability Assessment, IEEE

Transaction on Power Systems, Vol.9, No.1, February, pp. 525-532. Qiu J. and S.M.Shahidelhpour; 1987. A New Approach For Minimizing Power Losses and Improving Voltage Profile. IEEE

Transaction on Power Systems, Vol. PWRS-2, No.2, May, pp. 287-295. Tamura, Y.Mori, H.Iwamoto,S., 1993. Relationship between Voltage Instability and Multiple Load Flow Solutions in Electric

Power Systems, IEEE Transaction on Power Apparatus and Systems, PAS-102, No.5, pp.1115-1125. Taylor C. W., 1994, Power System Voltage Stability, New York; McGraw-Hill, Inc. Zhu T. X., S.K. Tso. and K.L.Lo., 2000. An Investigation into the OLTC Effects on Voltage Collapse, IEEE Transaction on

power Systems, vol.15. No.2, May, pp.515-521.


18

Biographical notes: Ahmed Abu-Siada received the B.Sc. and M.Sc. degree from Ain Shams University, Egypt and the PhD degree from Curtin University of Technology, Perth, Australia, All in Electrical Engineering. Currently, he is a lecturer in the Department of Electrical and Computer Engineering at Curtin University of Technology. His research interests include power System Stability and Control, Power Electronics, Power Quality, Condition Monitoring, Energy Technology and System Simulation. He is a member of IEEE. Syed Islam (S81, M, 83, SM93) received the B.Sc., MSc, and PhD degree all in electrical power engineering in 1979, 1983, and 1988 respectively. He is currently the Chair Professor in Electrical Power Engineering and Head of Department of Electrical and Computer Engineering at Curtin University of Technology, Perth, Australia. He received the IEEE T Burke Hayes Faculty Recognition award in 2000. He has published over 140 technical papers in his area of expertise. His research interests are in Condition Monitoring of Transformers, Wind Energy Conversion, and Power Systems. He has been a keynote speaker and invited speaker at many international workshops and conferences. He is the current Vice-Chair of the Australasian Committee for Power Engineering (ACPE) and a member of the steering committee of the Australian Power Institute. He is a Fellow of the Engineers Australia, a senior member of the IEEE IAS, PES and DEIS, a Fellow of the IET and a chartered engineer in the United Kingdom. He is regular reviewer for the IEEE Trans. on Energy Conversion, Power Systems and Power Delivery. Prof. Islam is an editor of the IEEE Transaction on Sustainable Energy. E.A. Mohamed received the B.Sc. and M.Sc. degree from Ain Shams University, Egypt and the PhD degree from the University of Manitoba, Canada, All in Electrical Engineering. Currently, he is a Professor in the Department of Electrical Engineering, Ain Shams University, Egypt His research interests include power System Stability and Control, Power Quality, Application of AI to power systems and System Simulation. Received August 2009 Accepted March 2010 Final acceptance in revised form March 2010

MultiCraft


Vol. 2, No. 3, 2010, pp. 19-28




Application of radial basis neural network for state estimation of power

system networks

J.P. Pandey1*, D. Singh2

1*Department of Electrical Engineering, Kamla Nehru Institute of Technology Sultanpur INDIA 2 Department of Electrical Engineering, I.T,B.H.U Varanasi. INDIA

*Corresponding Author: e-mail: [email protected],

Abstract An original application of radial basis function (RBF) neural network for power system state estimation is proposed in this paper. The property of massive parallelism of neural networks is employed for this. The application of RBF neural network for state estimation is investigated by testing its applicability on a IEEE 14 bus system. The proposed estimator is compared with conventional Weighted Least Squares (WLS) State Estimator on basis of time, accuracy and robustness. It is observed that the time taken by the proposed estimator is quite low. The proposed estimator is more accurate and robust in case of gross errors and topological errors present in the measurement data. Keywords: Radial Basis Function Neural Networks, State Estimation.

1. Introduction

Electric Power System deregulation has transformed state estimation from an important application to a critical one. The system operator has to make equitable, security related, congestion management decisions to curtail or deny power transfer rights in real time. It has to be founded and justified on a precise model of the power system derived from the state estimation process. Moreover fast and accurate state estimation is foundation of locational marginal pricing methodologies for transmission management costing.

The state estimation provides the real time representation of the conditions in a power network. A state estimator is a data processing algorithm, which transforms meter readings and the switch status information into an estimate of the systems state (voltage magnitudes and phase angles at all the nodes). Real and reactive bus power injections, and real and reactive line flows and bus voltage magnitudes are the measurements, which are transmitted to computer control system via telemetry system. These measurements contain random noise due to instrument and phenomenon errors. The state estimation program obtains a best fit for the power system state variables by minimizing these errors. Ideally state estimation should run at the scanning rate of the telemetry system (say at every two seconds). Due to computational limitations, most practical state estimators run every few minutes or when major changes occur.

Most of the state estimation problems are formulated as over determined system of non-linear equations and solved as a weighted least squares problem. The Weighted Least Squares Estimation (WLSE) is by far the most popular approach in industry. The least squares technique is slow and computational requirements are prohibitively large since there is large number of redundant measurement data normally available in form of nodal injection and line flows. To overcome this difficulty, a number of alternatives algorithms including modification and refinements of the basic WLSE have been presented (Horisberger et al, 1976; Garcia et al, 1979). The state of the art in state estimation algorithms is presented in (Wu 1990; Monticelli, 2000). Most of the practical implementation of state estimation in electric power systems is based on the Gauss-Newton methods. The state estimates, i.e., the voltage magnitudes and the bus voltage angles are solved through an iterative procedure in least squares sense.

Pandey et al. / International Journal of Engineering, Science and Technology, Vol. 2, No. 3, 2010, pp. 19-28

20

The slowness of the state estimators due to computational requirements has been a major drawback of the present methods of state estimation employed. The conventional state estimation is based on algorithmic method of solving a large number of non-linear equations based on network line flows and /or bus injections and network constraints similar to power flow problem.

The present work proposes entirely different paradigm of state estimation problem. In this paper state estimation is addressed as a pattern recognition problem and solved using learning approach. The learning based methods have found wide applications in some EMS applications such as load forecasting, Topology Processing, Optimal unit commitment, Load Flows etc (Singh et al, 2001). In this paper an original application of radial basis function neural network for state estimation is proposed. The property of massive parallelism of neural network is employed for faster state estimation. The proposed estimator is studied for various cases to show its utility for state estimation in terms of accuracy and time requirements. 2. WLS State Estimator

Most state estimation programs in practical use are formulated as over determined systems of nonlinear equations and solved as WLSE problem. Consider the nonlinear measurement model

jj exhjz += )( (1) where jz is the

thj measurement, x is the true state vector, )(xhj is a nonlinear scalar function relating the thj measurements to

states, and je is the measurement error, which is assumed t

Documents

IJEST Special Issue Complete