多种搜索算法的五子棋博弈算法研究

摘要

Gobang is selected as the target for studying,which combines with Alpha-Beta pruning algorithm and replacement Table to achieve game results.Iterative deepening and local search methods are adopted to improve chess level.Monte Carlo method and depth of learning on the basis of a combination of ways are applied to improve chess skills.Experiments show that the proposed algorithm has been significantly improved in comparison with the above-mentioned methods.%主要选择五子棋为研究对象,应用Alpha-Beta剪枝算法、置换表技术搜索算法,研究人工智能模拟人类思考的推算过程,实现博弈效果.在Alpha-Beta剪枝算法中引入迭代加深以及局部搜索方法,提高程序棋技.在此基础上使用Monte Carlo方法和深度学习方法结合的方式来提高下棋技巧.实验结果表明,该算法相比于上述几种方法有明显的改进.