The Recognition of Fault Type of Transmission Line Based on Wavelet Transmission and FNN

The article firstly utilizes wavelet transform to do denoising, filtering and other preprocessing to the information collected and realizing the purification of three-phase component and zero sequence current component which used for determining the ground fault, then using the effective combination between fuzzy set theory and neural network to build five-layer fuzzy neural network, taking T-S fuzzy reasoning model as an interference layer in network, then regarding the purified and optimized fault characteristic quantity preprocessed by wavelet transform as the input of fuzzy neural network, regarding fault type as output to train and learn the fuzzy neural network to further realize the way of recognition of fault type. Finally, simulation testing shows this way can effectively realize the recognition of fault type, strongly enhancing the recognition rate of fault types.