On-Line Dynamic Security Assessment of Power Systems Utilizing Case-Based Reasoning Approach

Abstract

Online dynamic security of power system is one of the main issues of reliable operation of power system due to increasing stress on power system network and operating of systems near their stability limits. Dynamic security is the ability of power system to maintain its synchronism among system’s machines during and following contingencies and disturbances. This work shows effectiveness of Case-Based Reasoning (CBR) technique for dynamic security analysis. CBR is a type of machine learning method that falls within the broader domain of artificial intelligence. The principle behind CBR involves utilizing solutions from past problems and adapting them to estimate solutions for new problems by making necessary modifications. This technique is applied to dynamic security analysis of a standard IEEE 9 bus system, that is a highly nonlinear task. For creating initial data, the system was modeled in ETAP program and transient analysis is executed for different load-generation conditions of power system. Security level of power system is determined by calculating critical clearing time of a fault (CCT). To check robustness and efficiency of CBR technique, it is compared with the performance of Artificial Neural Network (ANN) for dynamic security assessment.