具有完备信息的非合作计算机博弈系统的研究与实现

发布时间：2018-11-19 20:27

【摘要】：计算机博弈是计算机科学与人工智能领域最具挑战的研究方向,是一个新兴的研究领域。它以人工智能和各种计算机博弈技术为基础,研究如何让计算机像人类一样从事高度智能的博弈活动。同时,它的研究也为计算机技术与人工智能带来了很多重要的方法和理论,近年来被广大的学者所接受,其研究成果产生了广泛的社会与学术影响。如果能够掌握下棋的本质,也就掌握了人类智能行为的核心通过大量阅读相关参考文献,本文对于完备信息非合作计算机博弈的相关技术进行了探讨。按照计算机博弈的一般原理和基本规则,设计了一个机器博弈系统软件。本文中提出的机器博弈系统可以分为以下两个主要部分：对战平台和博弈引擎。本文中所做的工作主要包括如下几个方面：(1)对计算机博弈搜索技术进行了研究,包括极大极小算法、负极大值算法、α-β剪枝算法等,设计了一个博弈对战平台和博弈搜索引擎独立的计算机博弈系统。(2)研究了对战平台的关键技术,包括棋局数据结构表示、匿名管道技术、对战平台与博弈引擎间的数据通信等,并实现了该对战平台。(3)研究了博弈搜索引擎的关键技术,给出了几个搜索算法的具体实现,解释了棋局的评估方法。以苏拉卡尔塔棋作为搜索引’擎的例子,实现了基于“兵力”模型的动态非线性搜索技术的算法。尽管还有一些不足之处有待改进,但程序运行的结果证明该对战平台在计算机博弈中是有效的。并且,还对博弈搜索引擎进行了一些实验,结果表明计算机表现得更加“智能”。这说明本系统达到了预期要求。
[Abstract]:Computer game is the most challenging research direction in the field of computer science and artificial intelligence, and it is a new research field. Based on artificial intelligence and various computer game techniques, it studies how to make computers engage in highly intelligent game activities like human beings. At the same time, its research has brought a lot of important methods and theories for computer technology and artificial intelligence, which has been accepted by many scholars in recent years, and its research results have produced extensive social and academic influence. If we can grasp the essence of chess, we can grasp the core of human intelligent behavior. Through reading a lot of relevant references, this paper discusses the relevant technology of complete information non-cooperative computer game. According to the general principle and basic rules of computer game, a machine game system software is designed. The machine game system proposed in this paper can be divided into two main parts: the battle platform and the game engine. The work in this paper mainly includes the following aspects: (1) the computer game search technology is studied, including minimax algorithm, negative maximum algorithm, 伪-尾 pruning algorithm and so on. A computer game system with independent game platform and game search engine is designed. (2) the key technologies of the platform are studied, including data structure representation of chess game, anonymous pipeline technology, etc. The data communication between the battle platform and the game engine is realized. (3) the key technology of the game search engine is studied, and the implementation of several search algorithms is given, and the evaluation method of the chess game is explained. Taking Sulacarta as an example, a dynamic nonlinear search algorithm based on "force" model is implemented. Although there are still some shortcomings to be improved, the results of the program show that the platform is effective in computer game. Some experiments on game search engine show that the computer is more intelligent. This shows that the system meets the expected requirements.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP18

【相似文献】