Composite Power System Reliability Evaluation Using Importance Sampling and Convolutional Neural Networks

Abstract

This paper proposes a new approach for evaluation of power systems reliability using Monte Carlo Simulation. Using standard Monte Carlo Simulation, a composite of Convolutional Neural Networks (CNN) and Importance Sampling (IS) is proposed for computing power system reliability indices. It is shown that the computational efficiency can be dramatically increased if the machine learning techniques are used in conjunction with well-known variance reduction technique of importance sampling. The IEEE Reliability Test System (IEEE-RTS) is used for studying the proposed method. The results of case studies show that CNNs together with importance sampling provide a good classification accuracy while reducing computation time substantially.