On-board Rapid Planning for Booster Trajectory of Multistage Rocket Based on SQP and Surrogate Model

Based on the Sequential Quadratic Programming (SQP), this paper proposes an on-board rapid planning guidance for the boost phase of multistage rockets. Considering the time-consuming of the SQP algorithm, a surrogate model is built in order to reduce the times of iteration. Firstly, according to the multistage rocket model and traditional SQP algorithm, an initial trajectory is achieved by flight trajectory optimization. Secondly, considering the varieties of constraints and aerodynamics deviation, a polynomial surrogate model is built for calculating the approximate guidance command in every planning period. Finally, improved SQP algorithm is used for on-board trajectory planning with the approximate guidance command. Numerical simulation results show that surrogate model can effectively reduce the time consuming of SQP algorithm and on-board trajectory planning problems with multiple constraints and uncertain aerodynamic coefficient can be solved by the improved SQP effectively.