基于多agent的虚拟股市仿真研究  

Study on Simulation of Virtual Stock Market Based on Multi-Agent

作　　者：刘大海[1] 王治宝[1] 孙洪军[1] 王秀峰[1] 

机构地区：[1]南开大学信息技术科学学院,天津300071

出　　处：《计算机工程与应用》2003年第25期26-28,共3页Computer Engineering and Applications

基　　金：天津市自然科学基金(编号:023601411)

摘　　要：复杂适应系统的基本思想认为系统的复杂性起源于其中个体的自适应性。该文借助Multi-agent的建模技术,对真实股市和投资者的行为进行合理的约简,设计出具有自适应能力的股民(agent),通过自适应股民的相互行为,从而构建一个虚拟的股票市场。最后对仿真的结果进行分析并展望了未来的研究方向。Holland John put forward the theory of Complex adaptive system in1994;its basic idea is that complexity of system was originated from adaptability of individual among system.First,we predigest actual stock market reasonable.Then we construct adaptive agents possessing ability of self-study base on the modeling technology of Multi-Agent.Thus,model of stock market is constructed based on simple mutual behavior of adaptive agents successfully.Finally,we analyze the result of experiment and the prospect future research work.

关 键 词：复杂适应系统 MULTI—AGENT 虚拟股市 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]