Lifetime Prediction of Composite Insulator Based on BP Neural Network

Abstract

Composite insulators in long-term operation are subjected to ultraviolet, electric field, moisture, temperature difference and other factors, which inevitably lead to aging. It is important to carry out prediction research on the lifetime of composite insulator so as to replace the aging and deteriorated insulators to ensure the normal operation of power grid. In this paper, the performance indicators with strong correlation with insulator life were selected by Poisson correlation analysis, and the operation lifetime was predicted based on BP neural network. The research results show that when nine performance indexes such as hardness, hydrophobicity, water diffusion leakage current and tensile strength are used to predict the lifetime of the insulator, the difference between the predicted and actual values is within 5%.