启发式深空探测器任务规划方法

Heuristic Mission Planning Approach for Deep Space Explorer

针对深空探测器的任务规划过程特点,考虑约束的数值特性,提出一种基于规划活动相关度的领域无关启发式规划方法。定义了活动的相关度,结合规划过程,给出自学习式的相关度动态更新过程。设计基于相关度的活动选择机制,形成启发式规划算法。数值计算结果表明,文中提出的方法有效地减少了规划过程中规划步数和回溯步数,提高了深空探测器规划的效率。

Considering the numeric constraints, a planning approach based on a domain independent heuristic approach is proposed, and the heuristic is abstracted from the planning activity relevancy. The relevancy between activities is defined, and a dynamic self-learning updating process is given. The heuristic is designed based on the relevancy, guiding the activity selecting in the planning process, and the heuristic planning algorithm is finally proposed. The simulation results show that the algorithm can efficiently reduce planning steps and backtracking steps and improve the planning efficiency for deep space explorer.