基于Agent的芯片制造生产线动态调度方法研究

基于智能体技术，提出了芯片制造生产线动态调度新方法，实现了投料调度、工件调度与设备维护调度的集成。首先，给出了基于智能体的动态调度模型，该模型中包括管理智能体、投料智能体、工件智能体、设备智能体、运输智能体与人员智能体。投料智能体用于实现投料控制，工件调度通过工件智能体与设备智能体之间的协商实现，设备维护调度由设备智能体实现。通过智能体间的合作，能够实现投料控制、工件调度与设备维护调度的协同进行。然后给出了工件智能体与设备智能体问协商使用的协商协议（即带有时间约束的单步协商协议），以及智能体决策中使用的调度算法（包括投料调度算法，工件调度算法与设备维护调度算法）。最后，通过实例，给出了方法的使用过程，并进行了模拟仿真。仿真结果表明，集成的调度方法能更好地优化模型的生产率、加工周期与设备利用率，最终提高准时交货率。

Research on agent-based dynamic scheduling approach for semiconductor wafer fabrication

A new agent-based dynamic scheduling approach for semiconductor wafer fabrication was proposed,which integrated release control,dispatching and machine preventive maintenance scheduling.Firstly,an agent-based dynamic scheduling model was presented,which consisted of six kinds of agents,i.e.manager agent,input agent,workpiece agent,machine agent,transportation agent and operator agent.Release control was fulfilled by input agent,dispatching was performed by the negotiations between machine agents and workpiece agents and machine maintenance scheduling was implemented by the machine agents.The release control,dispatching and machine preventive maintenance scheduling were realized by agent collaboration.Secondly,the negotiation protocol among machine agents and workpiece agents,single-step negotiation protocol with time bound,was given.Thirdly,decision-making algorithms of agents were offered,including release control algorithm,dispatching algorithm and machine maintenance scheduling algorithm.Finally,the simulation results indicated that the proposed integration approach was advanced than original approaches without integration on throughput rate,cycle time, machine utility and on-time-delivery rate with better performance.