行星表面变尺度特征识别与导航观测构型抗差优化研究

项目名称： 行星表面变尺度特征识别与导航观测构型抗差优化研究

项目编号： No.61304248

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 朱圣英

作者单位： 北京理工大学

项目金额： 24万元

中文摘要： 行星表面存在大量的陨石坑、岩石等形貌特征，基于行星表面特征进行自主导航，能够取得优异的导航性能，是未来深空探测器的重要导航方式。由于行星表面形貌的复杂性以及深空环境的特殊性，给自主导航带来了挑战，导航信息获取、观测误差抑制是亟需解决的技术难点。本项目针对行星表面特征导航中难点问题，通过对仿射几何不变性和系统可观测度的深入分析，重点开展行星表面变尺度特征提取与匹配、观测构型抗差优化等研究工作，提出基于几何描述/仿射不变性的导航特征匹配识别方法、基于数据深度加权机制的导航特征误匹配校正方法、基于可观测度分析的观测构型抗差优化设计方法等，突破行星表面特征导航的相关关键技术，期望对我国将来进行的深空探测任务有所帮助。

中文关键词： 深空探测；自主导航；导航信息；观测构型优化；

英文摘要： The terrain features, such as craters and rocks, are very rich on the surface of planets. Therefore, autonomous navigation based on the terrain features is expected to achieve high level of navigation performance, and is regarded as the main way of autonomous navigation for future deep space explorer. However, the complexity of the surface topography and the particularity of the deep space environment have brought great challenges to the autonomous navigation system. In order to cope with the challenges, it's desiderated to accurately obtain navigation information under variable scale, to solve the miss matching of the characteristics and to overcome the influence of the position error. Aiming at the mentioned problems, this project will research deeply on affine geometric invariance and system observability, and the issues of feature extraction and matching and the optimization design of desensitized observation geometry are investigated based on variable scale and system observability theory. The navigation features identification and matching based on geometric description and affine invariance, the landmark matching error correction based on data depth weight theory and the optimization design of desensitized observation geometry based on system observability theory are proposed, and the proposed key technol

英文关键词： deep space exploration；autonomous navigation；navigation information；Observation geometry optimization；