Fault Diagnosis of Power Network Based on GIS Platform and Bayesian Networks

Abstract

In order to determine the location of the fault components of the power network quickly and give troubleshooting solutions, this paper obtains a simplify structure of relay protection and circuit-breaker as key equipment by analyzing the power network topology of GIS platform and uses the Bayesian networks fault diagnosis algorithm and finally designs the power network fault diagnosis module based on GIS platform. Fault diagnosis algorithm based on Bayesian networks is a new method for power network fault diagnosis which deals with the power network fault diagnosis with incomplete alarm signals caused by the protection device’s and the circuit breaker’s malfunction or refusal to move, device failure of communication and other reasons in the use of Bayesian networks method.  This method establishes the transmission line fault diagnosis model by using Noisy-Or, Noisy-And node model and similar BP neural network back propagation algorithm, and obtains the fault trust degree of each component by using the formula, and finally determines the fault according to the fault trust degree. The practical engineering application shows that the search speed and accuracy of fault diagnosis are improved by applying the fault diagnosis module based on GIS platform and Bayesian network.