Power System Reliability Evaluation using Monte Carlo Simulation and Multi Label Classifier

Abstract

This paper presents a new method for evaluation of power systems reliability indices. In this study, a combination of Monte Carlo Simulation (MCS) and Multilabel Radial Basis Function (MLRBF) classifier is used for computing system reliability indices. Multilabel classification algorithms is different from single label approaches, in which each instance can be assigned into multiple classes. This study shows that MLRBF can be used to classify composite power system states (success or failure) without requiring optimal power flow (OPF) analysis, with exception of training phase. Therefore, this approach shows that the computational efficiency of the reliability evaluation analysis to evaluate reliability indices can be significantly increased. The proposed method is applied to the IEEE Reliability Test System (IEEE-RTS-79) for different load levels. The outcomes of case studies show that MLRBF algorithm provides good classification accuracy in reliability evaluation while reducing computation time substantially.