基于复杂耦合时间线的深空探测器自主任务规划方法研究

Onboard autonomous mission planning is the key technology to enable long-term autonomous operation and control of deep space probes in complex environment. However, there are difficulties in planning and scheduling for physical systems due to complicated constraints, coupled subsystems, and dynamic changing parameters of deep space probes. In consideration of these characteristics of the probe, mechanism of hierarchical constraint propagation and dynamic technique is studied in order to establish a decoupling constraint propagation mechanism. Moreover, rapid mission planning technique based on state transition graph is put forward which employs lowest-cost-first heuristic search method. Finally, simulation and analysis are performed. Through theoretical analysis and numerical simulation, autonomous planning methods are deeply studied, which resolve the issues of propagating hierarchically during complicated constraints network and consist checking and issues of fast mission planning on board. In conclusion, the project provides a new way and a new technical basis for the design of the onboard autonomous planning system for deep space probe in the future.

星上自主任务规划技术是实现深空探测器复杂环境下长期自主运行的关键。针对深空探测器子系统互相耦合、约束关系复杂及参数动态变化等特点，研究约束分组映射传播与动态处理方法，建立复杂时间线状态转移图，设计最低代价优先的时间线快速任务规划方法，并进行仿真分析。通过理论分析和数值模拟，深入研究深空探测器系统自主规划方法，解决多系统耦合与约束复杂条件下的约束网络分组映射传播和一致性判断以及星上快速任务序列规划问题，为深空探测器星上自主任务规划系统的设计和实现提供新思路和技术途径。

星上自主任务规划技术是实现深空探测器复杂环境下长期自主运行的关键。针对深空探测器子系统互相耦合、约束关系复杂及动态变化等特点，对规划过程中时间约束推理方法和规划搜索技术进行了研究，提出了时间约束分组映射传播机制与优化处理方法，建立了复杂时间线状态转移图，设计了状态转移启发式，提出了最低代价优先快速任务规划方法，并进行仿真分析。通过理论分析和数值模拟，本项目方法提高了时间约束推理效率，降低了规划步数，从而减少了规划问题求解时间，为深空探测器星上自主任务规划系统的设计和实现提供新思路和技术途径。