【作者】 闵飞炎； 杨明；

【Author】 Feiyan Min Ming Yang (Control and Simulation Center. Harbin Institute of Technology, Harbin,150080)

【机构】 哈尔滨工业大学控制与仿真中心；

【摘要】 基于智能体的方法是对智能个体的学习和适应行为建模有效手段。这些智能个体的学习和适应行为,是感知、认识和行动等过程相互作用的结果。在DEVS中分别以外部和内部转移函数描述这些过程,并对相应的状态变量更新,从而实现对学习和适应过程的建模。这一方法得以实现,是建立在对对智能体的状态变量分类,即目标变量,相关变量和行为变量的基础上的。此外,这类智能体模型的运行环境和相应的仿真策略应进行了讨论。最后,在行人路经选择实例的基础上讨论了这类模型实现的细节。更多还原

【Abstract】 Agent based modeling is considered to be a sound and effective way to the simulation of adaptive and intelligent units and their interactions. The intelligent behavior, comprehended as combination of perception, cognition and motor, is integrated in DEVS, a widely accepted modular modeling formalism. This DEVS based agent model, become possible, lies on the fact that agent state variables are classified into three subclasses. The intelligent behavior is specified as three functions defined on some of these subclasses to some other variables. The detail modeling theories and methodology is provided. In addition, the environment and corresponding algorithms with which the agent model executed is also discussed.更多还原