Reliability Prediction of UHF Partial Discharge Sensor Based on Inverse Gaussian Process

The degradation process of Ultra high frequency PD sensors is complicated and random due to the harsh distribution station scenario. The reliability research of Ultra high frequency PD sensors has become an important topic. This article predicts the reliability and lifespan of sensors using performance degradation data by establishing an inverse Gaussian model. Firstly, the working performance and degradation mechanism of Ultra high frequency PD sensor are analyzed, and the output port voltage is determined as the performance degradation reference; Then, by selecting multiple groups of test data, the constraint relationship between drift parameters and diffusion parameters of the Inverse Gaussian Process is calculated, and the failure criteria are finally determined. The reliability function and reliability life of Ultra high frequency partial discharge sensors under normal operating temperature conditions are obtained, which verifies the feasibility of the reliability prediction method based on the inverse Gaussian process.