Fuzzy logic based on-line fault detection and classificationmethod of substation equipment based on convolutional probabilistic neural network with discrete wavelet transform and fuzzy interference

In power systems engineering, early faults detection, as well asa location in networks is a major challenge that results in damage and revenue to facilities and equipment, energy loss.The reason for the delay in detection is that most network operators rely on data/complaints supplied by customers without having a system in place to check and verify whether the information is misleading or veridical.This paper proposed the novel fault detection in power system operators for sub-station equipment by a convolutional neural network (CNN)for detecting and relegating blemish of the network by training voltage and power values by analyzing the circuit breakers of the network. Then, at that point, discrete wavelet change (DWT) has been coordinated with the preparation part to get qualities of voltage and current signs in numerous recurrence groups of the organization at the sub-stations and this will bean master arrangement of issue location. The fault locations have been done using a fuzzy interference-based fault location system (FI-FLS) of the network. By this fault detection and locator, the aberrant and mundane conditions of the networks has been relegated and located. The simulation results show feasibility, effectiveness, average precision. The fault located by this proposed method is line to ground, line to line and, phase fault predicted on the fault parameter values.The proposed approach attains 96 % average accuracy for the detection that shows the efficiency of the performance.