行星着陆不确定因素智能映射表征与轨迹优化设计方法研究

The uncertainty of the complex dynamic system near the planet brings great difficulties and challenges to the design of the planetary landing trajectory. Under the traditional theoretical framework, it often relies on improving the redundancy to make up for the lack of research on the system uncertainty. However, this also seriously restricts the improvement of trajectory design performance. This project innovatively introduces the intelligent statistical learning idea into the analysis of the uncertainty propagation mechanism of the planetary landing dynamics. From the perspective of statistical analysis, it reveals the coupling relationships and influence laws between uncertainties and landing trajectory parameters, and establishes a direct mapping model of uncertainty. Additionally, the statistical characteristics are applied into the trajectory maneuver control design to achieve the optimal design and efficient evaluation of the planetary landing trajectory in complex dynamic environment. The project will focus on the complex landing environment of irregular weakly gravitational asteroids and atmospheric planets, to study the intelligent mapping model with uncertainty and trajectory optimization design method of planetary landing. The influence law of uncertainties on the trajectory parameters in complicated dynamics environment is striven to be given clearly and completely, and a systematic optimal trajectory design method and efficient evaluation method of uncertain nonlinear planetary landing are established, to provide innovative ideas for the basic theory of planetary precise landing, and to provide an effective tool of rapid analysis for the performance improvement of deep space exploration missions in the future.

行星附近复杂动力学系统不确定性的存在给行星着陆轨迹设计带来了极大的困难和挑战。传统理论方法往往依赖于提高设计冗余度来弥补对系统不确定性认识的不足，严重制约了轨迹设计性能的提升。本项目创新性地将智能统计学习思想引入行星着陆动力学不确定传播机理分析过程，从统计分析的角度揭示不确定因素与着陆轨迹参量间的耦合关系与影响规律，建立不确定性映射表征模型，并将其纳入轨迹机动优化设计中，实现在复杂动力学环境下对行星着陆轨迹的优化设计与高效评估。项目将重点针对不规则弱引力小行星和有大气行星这两类天体的复杂着陆环境，开展行星着陆不确定性智能映射表征与轨迹优化设计方法的研究，力求能够清晰完整地给出复杂动力学环境下不确定因素对着陆轨迹参量的影响规律，建立较为系统的不确定非线性行星着陆轨迹优化设计与高效评估理论方法，为解决行星精确着陆的基础理论问题提供创新性的思路，为未来深空探测任务设计提供有效的快速分析工具。

行星着陆轨迹设计的难点在于行星附近复杂动力学系统不确定性的存在。传统的理论方法往往依赖于提高设计冗余度来弥补对系统不确定性认识的不足，严重制约了轨迹设计性能的提升。本项目提出从统计映射角度运动智能机器学习思想解决行星着陆轨迹的优化设计与高效评估问题，取得的主要研究成果为：1）提出了小天体引力场检测点样本数据获取方法，解决了大规模高质量样本数据生成问题；2）从数据统计规律挖掘的全新角度，将高斯回归过程原理引入行星迊动力学模型，直接建立了小天体引力场与检验点的映射关系；3）提出了基于统计映射的全引力场高效建模计算方法，有效提升了建模效率；4）提出了行星着陆轨迹数据样本生成方法，构建了大规模的行星着陆轨迹数据样本集合；5）从数据统计角度深入挖掘着陆轨迹关键特征，提出了行星着陆不确定性与轨迹特征的直接映射学习方法；6）提出了不确定因素下鲁棒轨迹优化设计方法，有效提升了着陆轨迹的精度。.本项目研究成果学术论文9篇，其中SCI检索论文3篇；申请国家发明专利10项，其中已授权专利5项。出版学术专著1部—《地外天体着陆自主导航与制导》（共213000字）。获国防科技进步二等奖1项。本项目相关研究成果应用于火星探测器制导导航与控制系统的方案论证与设计优选，研究成果得到了应用单位的充分肯定。此外，项目相关成果还可为我国未来小天体探测任务中小天体附近动力学分析与操作任务设计提供支撑。