基于CAS的企业竞争模型与仿真研究  

ON CAS BASED FIRM COMPETITION MODEL AND SIMULATION

作　　者：邵艳华[1] 沈昭华[1] 彭刚[1] 李怡[1] 

机构地区：[1]贵州民族学院经济管理学院,贵州贵阳550025

出　　处：《计算机应用与软件》2010年第4期149-151,共3页Computer Applications and Software

基　　金：贵州省科技基金项目(20072205)

摘　　要：企业系统本质上是一个复杂适应系统,是一个开放的、动态演化的、具有自组织能力的系统。人工神经网络具有非线性逼近能力和自组织、自适应能力。提出运用神经网络的学习能力模拟Agent适应性的方法,借鉴Holland的CAS理论的建模思想,建立基于CAS理论的企业竞争模型。通过开发基于Swarm平台的仿真程序,模拟不确定环境下企业的规模分布和动态演化过程。Firm system is a complex adaptive system substantively,it possesses characters of opening,dynamic evolution and self-organisation.Artificial neural network has the capability of approximating any nonlinear system and specialty of self-learning and self-adapting.In this paper,a method of simulating the adaptability of Agent with neural network＇s learning ability was proposed,an applied model of CAS-based firm competition was built by referring to the modelling scheme of Holland＇s CAS theory.Through developing the simulation program based on Swarm toolkit,the firm＇s size distribution and dynamic evolvement in uncertain environment was simulated.

关 键 词：神经网络 复杂适应系统 智能体 SWARM仿真平台 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]