Application of Approximate Entropy to Fault Signal Analysis in Electric Power System

Abstract

With the research of approximate entropy and its essence, it shows that approximate entropy has predominance in analysis of non-stationary random signals. A novel feature extraction method is proposed, and a new and effective feature parameter is provided to describe faulty signals efficiently. The analysis of ideal power signals with approximate entropy and the comparison between approximate entropy and information entropy testify the predominance of approximate entropy in power signal analysis. Simulations of its application to feature extraction of non-solid earthed network prove that approximate entropy can well realize the feature extraction of power faulty signals under the condition of short time-series, small magnitude and so on. Therefore, it is feasible to put the effective feature parameter into the use of power faulty signal analysis. Moreover, the prospect of approximate entropy’s application to power fault diagnosis has been forecasted.