基于Agent的企业演化模型仿真研究  

Research about model and simulation of firm evolution based on Agent

作　　者：邵艳华[1,2] 李坚石[1] 田丰[1] 陈天健[1] 王锦荣[1] 

机构地区：[1]贵州大学计算机科学与技术学院,贵州贵阳550025 [2]贵州民族学院经济管理学院,贵州贵阳550025

出　　处：《计算机工程与设计》2009年第10期2489-2492,共4页Computer Engineering and Design

基　　金：贵州省科技基金项目(20072205)

摘　　要：针对企业演化过程的动态性和不确定性,借鉴Holland教授的复杂适应系统理论的建模思想,提出运用神经网络的学习能力模拟智能体适应性的方法,建立基于复杂适应系统理论的企业演化模型。主要从企业行为的视角对处于动态竞争中的企业如何应对竞争实现企业演化的问题进行探讨,期望充实关于竞争和演化的理论。通过开发基于swarm平台的Ob-ject-c仿真程序,模拟了不确定环境下企业的规模分布和动态演化过程,并对仿真结果进行分析。In accordance with the dynamics and tmcertainty during the enterprise evolvement. Referring to the modeling scheme of Holland＇s Echo model, a method of simulating the adaptability of agent by neural network is proposed, the issues related to how to respond and evolve in the dynamic competition from the perspective of evolutionary theory is discussed mainly. Secondly, an applied model of CAS-based firm evolvement was built. Finally, the size distribution and evolvement ofthe firm tmder uncertain environment is simulated through the program of Object-c based on swarm, and the results of the simulation are analyzed.

关 键 词：神经网络 复杂适应系统 智能体 SWARM仿真平台 动态演化 

分 类 号：TP319.9[自动化与计算机技术—计算机软件与理论]