Abstract: This paper defines the determined tree and designs a method to seek determined tree. This paper designs algorithm for strong planning based on the method. This algorithm finds the initial state corresponding to the determination of each tree. After finding the tree, strong planning solution has no need to search from the goal state to the initial state, and only needs to determine from the target state reverse to find any node of tree. Through the node reverse searching the initial state in the tree, a strong planning solution is got. The results show that the designed algorithm has higher efficiency than the reverse search method solution for strong planning algorithm.

Key words: determined tree, non-determinate plan, strong planning solution, reverse search

摘要： 定义了确定树，设计了求确定树的方法。基于该方法设计了求强规划解的算法，找到每个初始状态对应的确定树，在找到确定树之后，求强规划解不需要从目标状态一直搜索到初始状态，只需要从目标状态反向找到确定树的任意一个节点，再通过这个节点在确定树中反向搜索到初始状态从而得到一个强规划解。实验结果表明：所设计的算法比用反向搜索方法求强规划解的算法的效率高。

关键词: 确定树, 不确定规划, 强规划解, 反向搜索