To avoid the waste of time and hardware resources caused by repeated operations, this paper takes Chinese chess as the research object, and proposes a method for constructing computer chess game knowledge base (including red chess and black chess).The knowledge base automatically records the best movement and current Chessboard situation that are searched by Alpha-Beta pruning algorithm and historical heuristic algorithm during the computer "thinking".When we meet the same chessboard situation, directly follow the moves in the knowledge base.At the same time, in order to reduce the time consumption caused by the knowledge base retrieval, we use a hash value in the Zobrist hash technique to uniquely identify a chessboard situation.In view of the fact that Alpha-Beta search is not meaningful for the Opening, this paper introduces a variety of experts movements.Finally, the comparison of time efficiency in Visual Studio C++ environment proves the validity of the knowledge base of this paper.%为了解决重复局面导致时间资源和硬件资源的浪费问题,以中国象棋为研究对象,提出了一种构建计算机象棋(执红棋和执黑棋)博弈知识库的方法,知识库自动记录每次计算机"思考"时经过Alpha-Beta算法和历史启发算法搜索到的最佳走法和当前棋盘局面,下一次遇到相同局面时,直接检索知识库获取最佳对弈走法;使用Zobrist哈希技术中的一个哈希值来唯一标识一个棋盘局面,以减少知识库检索时造成的时间消耗;针对开局就使用Alpha-Beta算法搜索意义不大的问题,引入了多种专家开局走法.在Visual Studio C++环境下对时间消耗进行对比实验,结果证明了所提出知识库的有效性.
展开▼
