Active Disturbance Rejection Control of Hypersonic Vehicle Based on Q-Learning Algorithm

The control of hypersonic vehicle is characterized by strong coupling, large parameter fluctuation, nonlinearity and uncertainty. To solve the above technical difficulties, active disturbance rejection control (ADRC) is presented to track the expected pitch angle of hypersonic vehicles. Due to the problem that extended state observer (ESO) and nonlinear state error feedback (NLSEF) parameters in ADRC need to be debugged many times, this paper develops a Q-learning algorithm to adjust the optimal parameters of ADRC within a certain range. Simulation results indicate that the proposed control strategy has a better tracking performance.