A Deep Learning Based Data-Driven Thruster Fault Diagnosis Approach for Satellite Attitude Control System
Bing Xiao, Shen Yin
Abstract
The thruster fault diagnosis problem of the satellite attitude control system is investigated in this article. This challenging problem is first changed into the binary image classification issue. A deep learning based data-driven fault diagnosis approach is then presented. It benefits from this approach that the stuck-open and the stuck-close faults of thruster are detected, diagnosed, and located online with high accuracy. The proposed method is purely data-driven and directly implemented by using raw measurement data only. It is independent of the dynamics of the thruster and the mathematical model of the satellite attitude control system. The effectiveness of the proposed approach is finally demonstrated on a satellite example.