Fault Detection Method of Power Insulator Based on Deep Convolution Neural Network

Abstract

Aiming at the problem of low detection accuracy of traditional power insu-lator fault detection methods, a power insulator fault detection method basedon deep convolution neural network is designed. For the training of deepconvolution neural network, the fault detection of power insulator based ondeep convolution neural network is realized by anchor design, loss functiondesign, candidate region selection mechanism establishment and sharingconvolution features. The experimental results show that the fault detectionmethod of power insulator based on deep convolution neural network is moreaccurate than the traditional method, and the detection time is less.