计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (23): 6-8.DOI: 10.3778/j.issn.1002-8331.2009.23.002

• 博士论坛 • 上一篇    下一篇

复杂系统多智能体建模的博弈进化仿真

杨 波,徐升华   

  1. 江西财经大学 信息管理学院,南昌 330013
  • 收稿日期:2009-05-08 修回日期:2009-06-09 出版日期:2009-08-11 发布日期:2009-08-11
  • 通讯作者: 杨 波

Evolutionary simulation game based on module of complex multi-agent system

YANG Bo,XU Sheng-hua   

  1. School of Information Management,Jiangxi University of Finance & Economics,Nanchang 330013,China
  • Received:2009-05-08 Revised:2009-06-09 Online:2009-08-11 Published:2009-08-11
  • Contact: YANG Bo

摘要: 采用基于复杂系统多智能(Multi-Agent)体的整体建模仿真方法,利用Swarm仿真软件平台建立进化博弈仿真模型(EGM),对该模型进行了描述和定义,运用EGM对各类对称博弈模型进行仿真运行,并对结果进行分析。仿真结果表明在不同类型的对称博弈中初始策略选择概率和收益矩阵对动态稳定策略的获得有不同的影响,在给定不同的收益矩阵和初始策略选择概率下可以获得不同的进化稳定策略。

关键词: 复杂系统, 多智能体, 进化博弈, Swarm, 整体建模仿真

Abstract: Adopted by the method of evolutionary simulation game based on ensemble module of Complex Multi-Agent Aystem,the article builds a module of evolutionary simulation game—EGM.By making use of Swarm simulation software,the module is decribed and defined.EGM is applied to different symmetric game models and the result is obtained.It reveals that the initial probability of selection strategies and gains matrix have different effects on the gaining of dynamic stability strategy in different symmetric game models,and it can obtain different evolutionary stable strategies.

Key words: complex system, multi-agent, evolutionary game, Swarm, ensemble modeling and simulation method

中图分类号: