基于知识推理的博弈树搜索算法  

Search Algorithm for Game of Checkers Based on Knowledge Inference

作　　者：蒋加伏[1] 陈蔼祥[1] 唐贤英[1] 

机构地区：[1]长沙交通学院计算机工程系,长沙410076

出　　处：《计算机工程与应用》2004年第1期74-76,156,共4页Computer Engineering and Applications

基　　金：湖南省自然科学基金(编号:00JJY2059)

摘　　要：针对搜索技术中存在的由于搜索空间过于巨大而引起的搜索效率下降的缺点,该文提出了一种把规则提取、知识推理、搜索相结合的新算法。并把该算法运用到中国象棋中去。实验结果表明,该算法不但能保证博弈水平,还能提高搜索效率。Search technology is an important part of Artificial Intelligence.The search efficiency will descend when the problem state is large.In this paper,a new algorithm which combines t he rule distill,Knowledge Inference and search is presented.we apply the algo rithm to china chess,the result shows that the algorithm can not only guarante e the computer's game level,but also improves its search effiency.

关 键 词：搏弈树 搜索算法 规则提取 知识推理 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]