多重纳什均衡解的粒子群优化算法  

Particle Swarm Optimization Algorithm for Solving Multiple Nash Equilibrium Solutions

作　　者：瞿勇[1] 张建军[1] 宋业新[1] 

机构地区：[1]海军工程大学理学院,湖北武汉430033

出　　处：《运筹与管理》2010年第2期52-55,共4页Operations Research and Management Science

基　　金：国家自然科学基金项目(60774029;70471031);海军工程大学自然科学基金项目(HGDJJ07007)

摘　　要：提出了一种求解双矩阵对策多重纳什均衡解的粒子群优化算法。该算法通过随机初始点以及迭代粒子的归一化,保证粒子群始终保持在对策的可行策略空间内,避免了在随机搜索中产生无效的粒子,提高了粒子群优化算法求解纳什均衡解的计算性能。最后给出了几个数值例子,说明了粒子群优化算法的高效性。A Particle Swarm Optimization （PSO） algorithm for solving multiple Nash equilibrium solutions of bimatrix game is presented in this paper. By normalizing random initial particles and iterative particles in the proposed PSO algorithm, particles are always in the feasible strategy space that avoids producing ineffective particles in the random search procedure, so that the computation performance of the PSO algorithm for solving the Nash equilibrium solutions is improved. Finally, some numerical examples are given to illustrate the proposed algorithm.

关 键 词：粒子群优化算法 双矩阵对策 纳什均衡解 策略空间 

分 类 号：O225[理学—运筹学与控制论]