Fuzzy/terminal Sliding Mode Based Attitude Control For Hypersonic Vehicle

Due to the broad flight envelops spanning of the reentry process and the specific characteristics of the hypersonic vehicles, the design of the such control system presents the features of strong nonlinearity and strong coupling, and is challenging. In order to solve the attitude control problem of the hypersonic vehicles, the sliding mode control(SMC) strategies are studied extensively in this thesis.In order to improve the transient performance and to achieve the better robustness of the sliding mode control, an adaptive fast fuzzy integral sliding mode control(AFFSMC) is designed with the utilization of adaptive fuzzy logic system(FLS). Firstly, a manifold direct adaptive fuzzy control method is designed to guarantee the robustness and fast convergence property. Then, the direct adaptive fuzzy control is combined with integral sliding mode control to achieve the higher accuracy. The new sliding mode control law has the fast convergence property of the tracking errors, and better robustness than the direct adaptive fuzzy controller. Furthermore, to solve the problem of control input constraints, we design an auxiliary sliding mode manifold variable that can converge to zero in finite time. The stability of the closed-loop system can be proved via Lyapunov approach.Note that the exsiting SMC with PID-type sliding function suffers the problems such as the existence of reaching phase and poor performance on transient response. A robust adaptive fuzzy PID-type sliding mode control(AFPID-SMC) is designed with the utilization of radial basis function(RBF) neural network. Firstly, in order to improve the transient performance and ensure small steady state tracking error, the gain parameters of PID-type sliding mode manifold are adjusted by online learning on the basis of the attitude angle and angular velocity errors using adaptive fuzzy logic system(FLS). Additionally, the designed new adaptive law ensures that the closed-loop system is asymptotically stable. Meanwhile, the problem of the actuator saturation, caused by integral term of sliding mode manifold, is avoided even when large initial tracking errors. Then by considering the initial conditions of the system in the sliding function design, a global SMC with constant-gain PID-type sliding function is designed. Under the proposed attitude controller, the system states stay on the sliding surface from the very beginning of the control action, which ensures that the closed-loop system has the dynamic response determined by the sliding mode dynamics and improves the response performance of the closed-loop system greatly. Furthermore, to solve the problem that the upper bound information of lumped disturbances must be known before designing the sliding mode controller, RBF neural network observer is used to estimate the disturbances information.To make the attitude angular track the reference signals in finite time, a full-order nonlinear terminal sliding mode control scheme is proposed using terminal sliding mode and robust exact differentiator(RED). Firstly, a novel full-order terminal sliding mode is designed so that the derivatives of terms with fractional powers do not appear in the control law, the control singularities are avoiding. This control law makes the attitude angular track the reference signals in finite time. Then, to obtain the derivatives of the angular velocity signal in the sliding mode manifold, super-twisting algorithm(STA) is applied as a robust exact differentiator. Furthermore, the filter and boundary layer technique are combined to effectively alleviate chattering of the sliding mode controller.