Composite Power System Reliability Evaluation Based on Enhanced Sequential Cross-Entropy Monte Carlo Simulation

Abstract

Due to essential differences in modeling and computation, the traditional sequential cross-entropy method that is greatly effective for generation systems may lose its effectiveness for composite systems. An enhanced sequential cross-entropy method for composite systems is presented to achieve the effectiveness. In the pre-sample stage, a sequential cross-entropy optimization algorithm rather than using a traditional non-sequential optimization algorithm is proposed so that the samples in this stage could also contribute to calculations of reliability indices. In the second stage, an accelerated contingency analysis is presented for further reducing computational burdens, in which a bisection optimization algorithm is used to rapidly identify failure system states. The new likelihood ratios, i.e., frequency likelihood ratio and duration likelihood ratio are developed, resulting in accurate corrections of loss of load frequency (LOLF), loss of load probability (LOLP), and expected energy not supplied (EENS) indices. The performance of the proposed method is demonstrated by comparisons between the proposed and traditional methods using several composite power systems, including the IEEE RTS-79, a modified IEEE RTS-79, and the IEEE RTS-96.