摘要： 引入复杂适应系统理论，将股市、房地产系统视为一类复杂适应系统。采用多智能体方法进行建模，设计随机决策Agent、模仿Agent、BP神经网络Agent等模块，引入订单簿技术实现Agent之间的信息交互。通过仿真模拟该类复杂系统的动态演化过程，验证了模型的有效性。

关键词: 复杂适应系统, 智能体, 神经网络, Swarm仿真平台

Abstract: In the paper, stock market and real estate system are regarded as a class of complex adaptive system. In order to researching the modeling of the class of system, the Complex Adaptive System(CAS) theory is applied into the paper. The modeling approach based on Multi-Agent is adopted in the paper, it designs the stochastic decision Agent, simulation Agent, BP neural network Agent. It introduces the order book technology to realize the communication between Agents. The dynamic evolution under uncertain environment of the class of system is simulated, and the validity of the model is confirmed.

Key words: Complex Adaptive System(CAS), Agent, neural network, Swarm simulation platform

中图分类号: 

• TP311