再入轨迹多约束模型预测静态凸规划方法

A novel adaptive piecewise Chebychev pseudospectral method based-constrained model predictive static convex programming algorithm (P-CMPCP) is proposed for reentry glide trajectory planning of reusable vehicles with large lift-to-drag ratio. The high-precision trajectory solution can be solved iteratively under complete terminal state and multiprocess constraints. Firstly, considering the discontinuity of the control quantity caused by the overturning of the bank angle, it is difficult for the traditional planning method to strictly meet the terminal position and angle constraints simultaneously, and the process constraints are liable to exceed the limit, an adaptive piecewise pseudospectral discretization strategy is adopted. Subsequently, the sensitivity relationships among the flight state, process constraint and the control adjustment quantity at each collocation point are derived, so that the nonlinear optimal control problem is transformed into a static convex programming problem. Finally, the interior point method is utilized to optimize the control quantities of each segment. The smooth trajectory solution can be obtained efficiently without any simplification or approximation of the dynamic model. The simulation results indicate that the proposed approach has increased numerical accuracy compared to traditional methods.