RADIAL BASIS FUNCTION NEURAL NETWORK FOR PARTIAL DISCHARGE IDENTIFICATION IN HV GIS

Abstract

Gas Insulated switchgear comprise of many devices like circuit breaker, disconnector, Current transformer, voltage transformer, busbars and bus ducts. The insulation defects in these devices can be identified by Partial Discharge (PD) monitoring and analysis. The analysis of PD includes detection, recognition & classification of PD using various advanced mathematical tools & techniques. In the artificial intelligence, radial basis neural network methodology in MATLAB is one of the most popular and widely used for the analysis of PD. This work represents the generation of the partial discharge with known defects in GIS like cavity in epoxy bushing, particle on housing, and free particle etc. the signatures are used for training, testing and identification. The obtained PD pattern represents the characteristics of Partial discharge signal and the discrete spectrum interference signal with it. The PD signal that occur during testing and service conditions can be identified by expert RBF NN Tool. The expert algorithm will reduce the time in finding out the actual root cause that is creating PD.