Power System Security Assessment using Binary SVM Based Pattern Recognition S Kalyani, Member, IEEE, and K Shanti Swarup, Senior Member, IEEE Abstract—Power System Security is a major concern in real time operation. Conventional method of security evaluation consists of performing continuous load flow and transient stability studies by simulation program. This is highly time consuming and infeasible for on-line application. Pattern Recognition (PR) is a promising tool for on-line security evaluation. This paper proposes a Support Vector Machine (SVM) based binary classification for static and transient security evaluation. The proposed SVM based PR approach is implemented on New England 39 Bus and IEEE 57 Bus systems. The simulation results of SVM classifier is compared with the other classifier algorithms like Method of Least Squares (MLS), Multi- Layer Perceptron (MLP) and Linear Discriminant Analysis (LDA) classifiers. Keywords—Static Security, Transient Security, Pattern Recogni- tion, Classifier, Support Vector Machine. I. I NTRODUCTION S ECURITY evaluation is an important issue in planning and operation stages of an electric power system. The present trend toward deregulation has forced modern elec- tric utilities to operate the systems under stressed operating conditions closer to their security limits. Under such fragile conditions, any disturbance could endanger system security and may lead to system collapse. Therefore, there is a pressing need to develop fast on-line security monitoring method, which could analyze the level of security and forewarn system operators to take necessary preventive actions in case need arises [1]. Power System Security is defined as the ability of the system to withstand unexpected failures and continue to operate without interruption of supply to consumers [2]. One of the challenging problems in the real-time operation of power system is security assessment. Security analysis may be broadly classified as static security assessment (SSA) and transient security assessment (TSA). Static Security Analysis evaluates the post contingency steady state condition of the system neglecting the transient behavior and other time de- pendent variations. Transient Security Analysis evaluates the performance of the system as it progresses after a disturbance. Analysis of rotor angle stability is an essential component in TSA [3]. Any on-line TSA tool must provide a fast stability evaluation and system security analysis under perturbations. This paper presents a Support Vector Machine (SVM) based approach for on-line security evaluation. One of the S. Kalyani is with the Department of Electrical Engineering, Indian Institute of Technology, Madras, Chennai-600036. She is on deputation from K.L.N College of Engineering, Pottapalayam, Sivagangai, Tamilnadu to pursue Ph.D programme at IIT Madras (e-mail: kal yani 79@yahoo.co.in). K. Shanti Swarup is working as Professor in the Department of Electrical Engineering, Indian Institute of Technology, Madras, Chennai - 600036. (e- mail: swarup@ee.iitm.ac.in). important consideration in applying SVM to power system security evaluation is the proper selection of training feature set, characterizing the behavior of the power system. Many feature selection algorithms are available in the literature such as fisher discrimination analysis, entropy maximization, fisher discrimination [4]. The main problem with the existing feature algorithms is that it works well with linearly separable classes, but not well established on non-linearly separable classes [5]. In this paper, feature selection is performed by a simple approach called Sequential Forward Selection (SFS) method. Power system security evaluation is a complex non-linear problem, which has non-linear separability between secure and insecure classes. Literatures have reported the use of conventional algorithms like linear programming, least squares [6], decision trees [7] and different artificial neural network architectures [8] for design of classifier. To handle the problem of non-linear separability, SVM technique is adopted in the classification phase of the Pattern Recognition system. Further- more, in this paper, the logic of binary security assessment is considered, i.e., a given operating condition is deemed as either secure (1) or insecure (0). An operator likes to know exactly the disturbances that could cause insecurity and abnormality resulting from each disturbance for a given system operating condition, rather than its degree of security. The proposed SVM based classification approach is implemented on New England 39 bus and IEEE 57 bus systems. The simulation results prove that SVM classifier gives a better classification, enhancing its suitability for on-line security evaluation. II. POWER SYSTEM SECURITY The term ‘Security’ as defined by NERC (1997) is the ability of the electric systems to withstand sudden disturbances such as electric short-circuits or unanticipated loss of system element [9]. The main goal in security analysis is to increase the power system’s ability to run safely and operate within acceptable economic bounds. A set of most probable contin- gencies is first specified for security evaluation. This set may include outage of a line/generator, sudden increase in load, three phase fault in the system, etc [10]. A. Static Security Assessment Static security (also referred to as steady state security) is the ability of a power system to reach a steady state operating point without violating system operating constraints [11]. The violations of thermal limits of transmission lines and bus voltage limits are main concern for static security analysis. Under normal operating conditions, the following constraints World Academy of Science, Engineering and Technology International Journal of Electrical and Computer Engineering Vol:3, No:4, 2009 746 International Scholarly and Scientific Research & Innovation 3(4) 2009 scholar.waset.org/1307-6892/10828 International Science Index, Electrical and Computer Engineering Vol:3, No:4, 2009 waset.org/Publication/10828