Fault Diagnosis Method of Power System Based on Rough Set and Bayesian Networks

Abstract

In order to improve the efficiency and accuracy of large-scale power grid fault diagnosis, a grid fault diagnosis method was proposed based on Rough Set theory and Bayesian network. First of all, a decision-making table was built for grid fault diagnosis through the raw data used by description of the fault condition and definition and discretization of the attributes; Secondly, the redundant information contained in the decision-making table was reduced using the relevant knowledge of Rough Set, the minimal diagnostic rule was got; Then establish Bayesian network for minimum diagnostic decision rule, improve the efficiency and accuracy of power grid fault diagnosis using weights relationship among Bayesian network nodes. Simulation results show that the method is simple and feasible.