Research Review and Prospect of Fault Diagnosis Method of Satellite Power System Based on Machine Learning

Abstract

Aiming at the complex characteristics of the current satellite power supply system and the mass data produced in its operation process, an integrated fault diagnosis method based on deep learning and expert diagnosis was proposed. The deep learning has the capability to deeply investigate the incidence relation between abnormal data and faults, and the has the superiority of making full use of domain experts’ experience and simulating their reasoning process expert system es for solutions. The combination of them both can improve the reliability of satellite power failure diagnosis and provide a new way of thinking for the development of satellite power failure diagnosis method.