Multivariable Feedback Control Design for Hypersonic Vehicles Based on Stochastic Robustness Analysis

For the multivariable coupling hypersonic vehicle (HV) dynamics, it is required to achieve a strong robustness and high-accuracy tracking under aerodynamic uncertainties. In this paper, we propose a stochastic robustness analysis-based multivariable feedback control method for HV to satisfy design requirements with high probability. First, the HV dynamic characteristics are analyzed. Then, we model the relation between controller parameters and the robustness evaluated by the probability of instability and violation of requirements. Next, by maximizing robustness, the optimal parameters are searched for using particle swarm optimization. To reach a prescribed probability level, Hoeffding inequality is applied to set the sampling size of Monte Carlo evaluation. Simulation is conducted on a six-degree-of-freedom nonlinear model of HV. By making a full use of the uncertainty distribution, the proposed method effectively improves the attitude tracking performance of HV while guarantees robust stability.