Fuzzy Neural Network PID Control Based on RBF Neural Network for Variable Configuration Spacecraft

The mass distribution of spacecrafts with variable configuration and flexible appendages will change significantly during the configuration variation, and this will perform big disturbance to the attitude of the spacecraft. The traditional PID controller with manual optimization cannot perform well in this situation, to solve this problem, a novel adaptive FNN (Fuzzy Neural Network) PID controller based on RBFNN (Radial Basis Function Neural Network) is proposed. The parameters of PID controller were adaptively adjusted by FNN, and RBFNN with PSO (Particle Swarm Optimization) method is used to estimate the dynamic model and adjust parameters of the FNN with online Gradient Descent algorithm. The simulation results verify the effectiveness and practicability of the FNN PID controller based on RBFNN. It has better control quality in converge time, overshoot and accuracy compared with traditional PID controller.