Reliable Controller Design Based on Self-organization Fuzzy Neural Network for Satellite Attitude Control Systems

摘要

A fault detection and diagnosis (FDD) technique of the satellite attitude control systems (ACS) with actuators faults is investigated. A robust controller based on neural network (NNC) is presented which is insensitive to the disturbances. The system faults can be identified by the self-organizing fuzzy neural network (SFNN), and then the obtained estimation is used to reconfigure the control law. The neural network estimation system (NNES) begins to work after the fault is detected, which can reduce the computational complexity. Simulation results demonstrate the efficiency of the proposed FDD for satellite ACS.