A fault line selection method of small current grounding system based on wavelet packet and Bayes theory

Abstract

In view of problems of complex fault situation and unobvious characteristics of fault signal of fault line selection method of existing small current grounding system, a fault line selection method of small current grounding system based on wavelet packet and Bayes theory was proposed. Firstly, the method divides faults of small current grounding system into strong fault mode, medium fault mode and weak fault mode according to transition resistance, and constructs the Bayes classifier. Then, in extraction of fault features, it selects characteristic frequency band according to the maximum energy by db wavelet packet, and inputs the characteristic frequency band of every line into the classifier to judge fault type after training the Bayes classifier. Finally, it votes the output results of the three-classifier by majority principle, and acquires the fault line selection result. The simulation results show that the method has high accurate rate of line selection.

Keywords

• small current grounding system
• fault line selection
• fault mode
• bayes classifier
• characteristic frequency band