Research on satellites fault diagnosis method based on artificial intelligence

Abstract

In order to enable satellites to use fault diagnosis, fault prediction, failure prediction and other data models to determine the health status of satellites autonomously, According to the statistics of satellite orbit faults, this paper gives an artificial intelligence-based satellite fault diagnosis method and compares the advantages and disadvantages of three different satellite fault diagnosis methods to ensure that when a satellite fault occurs, it can implement autonomous reconfiguration, redundancy switching, autonomous mode switching or downgrade use when its own conditions allow The study ensures that when a satellite failure occurs, it can implement autonomous reconfiguration, redundancy switching, autonomous mode switching or downgrade use when its own conditions allow, ensuring the safety of satellite operation in orbit. Through the analysis of the application of typical fault diagnosis methods for a satellite control subsystem, it can be seen that artificial intelligence-based satellite fault diagnosis methods have gradually replaced traditional methods and become the core key technology for the future application of autonomous satellite health management system in orbit, which is also one of the core functions that future software-defined satellites should have.