火星表面大范围漫游探测协同导航与控制方法研究

After landing on the Mars eight years, the NASA Mars Exploration Rover Opportunity had successfully explored 34.4 kilometers till Jan. 24, 2012. Although this distance has exceeded the sum of all the distances explored on other planetary surfaces, this kind of rover, which is with the average distance of 12 meters per day, cannot satisfy the high-level planetary exploration missions in the near future. Therefore, in this project, we propose a novel scenario in which the multi-agent formation is employed to the planetary surface exploration. The long-range rover exploration is achieved by using the cooperative navigation and control for multi-agents. The novelties of this project can be summarized as follows: (1) the integration of the space (air) information with the planetary surface information and the strategy of the cooperative navigation for multi-agents; (2) the long-range rover global localization and the pose refinement by matching the planetary surface data with the orbital maps; (3) the cooperative control of the multi-agent formation and the consensus investigation in a distributed 3D environment; (4) experimental studies based on a simulated planetary surface. The research work of project is expected to fulfill much larger average driving speed which can be up to kilometers per day. The current planetary exploration by one rover is considered as the narrow and flat pattern. However, the proposed planetary exploration by heterogeneous multi-agents using the cooperative navigation and control would be a wide and tridimensional pattern, which will great benefit the future high-level scientific exploration on planets such as Mars and moon.

2012年1月24日，NASA"机遇"号火星漫游探测器在着陆8周年后成功地进行了34.4千米的漫游探测，这一探测距离超过了目前所有地外星表漫游探测距离的总和，但是这种平均每日约12米探测范围的单漫游器探测方式还远无法满足未来深层次行星探测任务的需要。本项目提出由星表多智能体系统进行编队探测的新思路，利用多智能体系统的协同导航与控制实现大范围行星表面探测的新方案，并对其中几个比较关键的科学问题进行理论探索和模拟实验。本项目中有特色的研究内容包括：(1) 太空（空中）和地面信息一体化技术与非同质多智能体协同导航策略；(2) 基于地表数据与轨道地形数据匹配的大范围漫游探测姿态修正与全局定位；(3) 3D空间多智能体编队的协同控制和一致性算法；(4) 基于模拟星表环境测试场地的导航与协同探测实验。本项目提出的研究方法预期实现每日千米级的星表漫游探测，能够满足未来对火星等行星深入科学探测任务的需要。

随着航天技术的日趋成熟与航天工业的蓬勃发展，火星探测任务已经逐步从方案论证走向了工程实践。无论在火星上开展何种科学探测，精准高效的火星车导航与控制算法是必不可少的，这也是科学任务得以安全实施的保障。2012年1月24日，NASA“机遇”号火星漫游探测器在着陆8周年后成功地进行了34.4千米的漫游探测，这一探测距离超过了目前所有地外星表漫游探测距离的总和，但是这种平均每日约12米探测范围的单漫游器探测方式还远无法满足未来深层次行星探测任务的需要。. 本课题针对目前单火星漫游器探测方式存在探测手段单一、范围有限、效率偏低、探测通道狭窄、导航精度低等缺点，结合未来火星探测发展方向和任务需求，提出了由星表多智能体系统进行编队探测的新思路，利用多智能体系统的协同导航与控制实现大范围行星表面探测的新方案，并对其中几个比较关键的科学问题进行理论探索和模拟实验。本课题中有特色的研究内容包括：(1) 太空（空中）和地面信息一体化技术与非同质多智能体协同导航策略；(2) 基于地表数据与轨道地形数据匹配的大范围漫游探测姿态修正与全局定位；(3) 3D空间多智能体编队的协同控制和一致性算法；(4) 基于模拟星表环境测试场地的导航与协同探测实验。. 本课题的建立的探测模式，形成的研究成果，开展的相关地面试验相信能够为后续火星等实际深空探测任务提供有价值的参考和理论依据，并且期望能够推动多智能体协同导航与控制的理论应用在国内快速展开。