Trajectory Tracking Control For The Air-breathing Hypersonic Vehicle Via Adaptive Dynamic Programming

Due to its great military strategic significance and civil value,Air-breathing hypersonic vehicle(AHV)has become a strategic commanding point attached to great importance by the world powers.As a key technology,flight control technology has attracted the extensive attention of scientists all over the world.Aiming at the trajectory tracking control of Air-breathing hypersonic vehicle in the cruise phase,this paper carries out relevant research based on adaptive dynamic programming(ADP)theory.The specific work is as follows:(1)Aiming at the problem of uncertainty in affine nonlinear systems,a robust controller design method based on ADP is proposed.First,an uncertainty term is introduced into the performance function.The robust controller solution problem is transformed into a nominal system optimal control problem.Then,an evaluation neural network is constructed,and a policy iterative algorithm is used to train the neural network.An approximate solution to the nominal system Hamilton Jacobi Bellman equation(HJB)is obtained.Furthermore,the optimal control law is derived directly.Finally,strict mathematical proof is given to verify the stability of the system.(2)An ADP based algorithm is proposed to solve the on-line tuning problem of PID gain parameters of the AHV attitude controller.By building the evaluation network and the action execution network respectively,the approximation of the optimal objective function of attitude subsystem and the fitting of PID control parameters are realized.And the gradient descent method is used to give a neural network weight adjustment formula that only depends on the observation data.The online closed-loop fast adaptive tuning of PID attitude control parameters is realized.The simulation results verify the excellent characteristics of the proposed algorithm.(3)In order to realize the altitude and speed tracking control of the AHV cruise section,a sliding mode controller design based on the ADP algorithm is proposed.The input and output feedback linearization method is adopted.Establish the dependencies between the inputs and outputs of the velocity subsystem and the altitude subsystems.Design sliding mode controller separately.The ADP control term is introduced for the obtained sliding mode dynamic system to Improve transient response and adaptive capability of sliding mode controller.Numerical simulation experiments verify the effectiveness of the controller.