摘要
博弈论是专门研究有利害冲突关系的数学理论,零和博弈是一类在社会竞争中适用面广泛的数学模型。解有限零和博弈问题的经典方法是线性规划方法。但它的运算复杂度随博弈参加者数量和选择战略数量的增加而急剧增加,并且编程复杂。该文所阐述的布朗方法是一种基于仿真的求解零和博弈的解法,它通过仿真具体的博弈过程来获得博弈的解。该方法的运算复杂度不会随着博弈加者数量的增加而急剧增加,克服了使用线性规划单纯型算法的缺陷。并且该方法计算步骤简单,易于编程实现,因此适用于计算机求解大规模的零和博弈问题。该文对这个算法做了具体介绍,并使用该方法求解了一个具体的博弈问题,最后根据得出的结果对布朗方法的特点进行了讨论。
Game Theory is the theory that studies the conflict relation. Zero sum game is an important mathematical game model. Linear program method is the classical method for solving finite zero game problem. There is large computation in it and it is hard to be realized on computer because its difficulty in programming. Brown method showed in this paper is the method that solves zero sum game problems based on simulation. Brown method is adapted to solve large scaled game problem because its computation is slow - added with the number of participator and is easy to be realized on computer. This paper introduces Brown method detailedly and solutes an exact example. At the end of this paper, there shows some characteristics of this method based on the solution.
出处
《计算机仿真》
CSCD
2005年第8期87-90,143,共5页
Computer Simulation
关键词
布朗方法
博弈论
有限零和博弈
Brown method
Game theory
Finite zero sum game
