Fractional Order Sliding Mode Based Attitude Control Design For Hypersonic Vehicle

With the development of the hypersonic flight, the requirements of the robustness and the control accuracy of the controller are higher and higher. The hypersonic vehicle’s high mach, the wide area of the flight, the coupling between the channels, the complexity of the parameter perturbation and the external uncertain distubance puts forwards challenge for the control system design. The appreance of the fractional order calculus opens a new gate to the sliding mode control. In this paper, the fractional order sliding mode strategy is proposed for the hypersonic vehicle.Based on the fractional order characteristic ratio assignment, a new kind of linear fractional order sliding mode strategy with no overshooting is designed, which makes the error of the attitude angle converge to zero. The speed of the response of the close-loop system can be adjusted by the corresponding time constant without changing the overshooting of the response.Aiming at the shortcomings of the linear fractional order sliding mode based on the fractional characteristic ratio assignment, a new kind of nonlinear fractional order sliding mode controller is proposed, to solve the contradiction between the speed of the response and the saturation of the control input. More than that, a time-constant-adaptive fractional order sliding mode surface is proposed, where the time constant is adaptive. The speed of the convergence of the sliding mode is improved, so that the reaching phase is shorter and the system becomes robuster.In order to solve the problem that the integral term of the PID sliding mode is too large, the fractional PID sliding mode controller is proposed, which add a freedom on the order of the integral term, then the phenomenon is weakened. It can make the error of the system asymptotically converge to zero. Finally, a fixed-time convergent fractional order time varying slidng mode is proposed based on the fractional PID controller. Then the error converges to zero within a fixed time. The error converges to zero since the intial moment, so the robustness of the system is enhanced.