Fuzzy PID Controller Based on Improved Neural Network for Satellite Attitude

In the field of satellite technology and its application, most of the satellite attitude control system uses PID control. However, the conventional PID control has certain limitations in large inertial system, nonlinear system and complex system. In order to overcome these problems, this paper puts forward a novel fuzzy PID controller by using the radial basis function neural network, genetic algorithm and Takagi-Sugeno fuzzy logic. Based on the dynamic model of the satellite attitude, the approach uses the rule from the improved fuzzy PID controller to ensure system stability. And the proposed control system is found to own many excellent performance of faster response speed, higher efficiency and smaller overshoot errors compared with the conventional PID control system. Finally, simulation results and comparative analysis are carried out to demonstrate that the proposed controller applied in the satellite attitude control system can effectively achieve attitude control and improve the response sensitivity. Therefore, it is feasible and effective in practical application to meet the demands of the Euler angles for satellite attitude control.