Abstract:The robustness improvement strategy for guidance of rocket powered vertical landing is investigated in this paper based on the receding horizon online convex optimization. The upper bound of thrust in receding horizon planning is level-wise released to cope with insufficient control ability during the ending phase of landing. Meanwhile, the online planned thrust command is compensated using apparent acceleration within each guidance cycle to mitigate the effect of disturbance and accumulative error, improving the recursive feasibility of receding horizon planning. Finally, the effectiveness of proposed strategies is validated with numerical experiments and comparative analysis.