Online-SVR-based GPC control for airframe/engine integrated near-space hypersonic vehicle

Due to its multi-mission profiles, complicated flight conditions, and large Mach number, the flight control system of NHV (near-space hypersonic vehicle) might have strong nonlinearity, coupling and fast-variability and be impacted by strong external disturbance and internal uncertainty. It is a great challenge for control law design. Based on the configuration of NHV, the aerodynamic/engine integrated control model for longitudinal motion of NHV is studied first. After that, for a class of uncertain nonlinear systems, an ideal GPC (generalized predictive control) strategy based on sliding-mode surface is derived and an adaptive law is proposed to eliminate the systems´ uncertainty using the online SVR (support vector regression), then the control character of this method is analyzed via Lyapunov method. At last, the proposed strategy is used as flight/propulsion integrated control law for the longitudinal motion of NHV and the simulation results demonstrate the good performance of this method.