利用局势变化攫取优势博弈  

Achieve Preponderant Game Through Changing Situation

作　　者：周玮[1] 王友政[1] 马庆增[1] 

机构地区：[1]东北大学流程工业综合自动化教育部重点实验室,沈阳110004

出　　处：《系统仿真学报》2008年第17期4505-4508,共4页Journal of System Simulation

基　　金：国家自然科学基金资助项目(60574050)

摘　　要：针对目前在中国象棋计算机博弈中,广泛使用的估值和搜索方法不能客观体现博弈双方各自实力的变化趋势,实现博弈者战略意图的实际情况,提出利用局势变化攫取优势博弈的方法,该方法以数学、人工免疫聚类等方法为工具,在中国象棋计算机博弈中建立"稳定的"和"变化的"局势模型,为博弈者将棋局局势了然于胸,驾驭或控制局势的变化和发展,提供理论依据。实践表明,基于博弈局势的估值和搜索方法,摆脱了机器博弈的传统思维定式,为走出好的"关键步",更好的全局寻优能力,奠定了基础。在中国象棋计算机博弈以及其他领域的博弈研究中有实际应用价值。The static evaluation and search methods used in Chinese chess computer game currently could not reflect the strength trend of game players objectively and accomplish the strategic purpose of game players, so game situation varying based on mathematics and artificial immunity cluster are presented to realize predominance game, which can create model of stable and varying situations in Chinese chess computer game, and also provide game players with theoretical rules to get a better understanding of game situation and control situation＇s trend. Practices show that the estimation and search methods based on game situation could get rid of the traditional thinking formularies and lay the basis for making good key moves and gaining a better globally optimizing ability, so it is of great value in Chinese chess computer game and other fields in game research.

关 键 词：中国象棋计算机博弈 优势博弈 局势变化 人工免疫聚类 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]