基于连续凸近似的多阶段弹道重规划求解算法  

An Algorithm for Multistage Trajectory Replanning Based on Successive Convex Approximation

作　　者：谭宇霄 彭博 邓智斌 王宇 李莉 TAN Yuxiao;PENG Bo;DENG Zhibin;WANG Yu;LI Li(University of Chinese Academy of Science,Beijing 100190,China;Beijing Institute of Astornautical Systems Engineering,Beijing 100076,China;MOE Social Science Laboratory of Digital Economic Forecasts and Policy Simulation at UCAS,Beijing 100190,China)

机构地区：[1]中国科学院大学,北京100190 [2]北京宇航系统工程研究所,北京100076 [3]数字经济监测预测预警与政策仿真教育部哲学社会科学实验室,北京100190

出　　处：《宇航总体技术》2024年第6期1-9,共9页Astronautical Systems Engineering Technology

基　　金：国家自然科学基金(12171151);中国科学院大学基金。

摘　　要：运载火箭弹道重规划问题是航天工程中至关重要的问题之一。目前运载火箭弹道重规划方法尚未对特定的多阶段弹道优化模型的特殊结构进行深入挖掘,计算效率有待进一步提升。通过建立多阶段弹道重规划问题模型,开展模型结构特性分析,最终在连续凸优化近似框架下构建了高效的定制化内点算法,以实现高精度、高鲁棒性的多阶段弹道重规划问题在线求解。仿真实验的结果表明,对典型火箭发射的SSO和GTO弹道,定制化内点算法的运行时间与传统算法相比分别减少了37.41%和47.86%,大幅提升了计算效率。The trajectory replanning problem of rocket launchers is one of the crucial issues in aerospace engineering.Currently,the existing trajectory replanning methods have not deeply explored the special structure of specific multi-stage trajectory optimization models,and the computational efficiency can be further improved.This paper establishes a multi-stage trajectory replanning problem model,analyzes the structural characteristics of the model,and finally develops an effective customized interior point algorithm by using successive convex approximation technology to achieve a high-precision and high-robustness online solution for trajectory replanning problems.The numerical simulation results show that,for typical SSO and GTO rocket orbits,the running time of the customized interior point algorithm is reduced by 37.41%and 47.86%respectively compared to traditional algorithms.The superior performance of the customized algorithm implies that it can meet the practical engineering demand of online trajectory replanning.

关 键 词：弹道重规划 连续凸近似 定制化内点算法 

分 类 号：V21[航空宇航科学与技术—航空宇航推进理论与工程]