Robustness optimization approach for hypersonic vehicle controller based on parallel computing

Air-breathing hypersonic vehicle maneuvering in large flight envelope has imposed urgent requirements for performance and robustness enhancement of control system. Considering the characteristics of multi-variables, strong coupling and complex nonlinearity, a nonlinear state feedback structure is developed based on feedback linearization technique, which is combined with tracking differentiator extracting differential signal and transient profile eliminating the contradiction between rapidity and overshoot. In presence of multiple parametric uncertainties, a novel cost function is derivated intuitively to evaluate the system robustness. Taking full advantage of hybrid particle swarm optimization, an optimization approach is developed to improve the robustness precisely and is significantly accelerated by parallel computing. Simulation results indicate that the robustness optimization approach based on parallel computing performs well both in robustness enhancement and computation speedup, with the controller tracking the commands rapidly and precisely.