基于MAS理论的无人化行车调度仿真模型

A Simulation Model of Unmanned Vehicle Scheduling Based on MAS Theory

为解决制造业车间多机多车多任务的复杂行车调度NP难问题,根据车间作业特点建立了基于MAS(Multi-Agent System)理论的无人化行车调度仿真模型。将行车、工位、作业任务抽象为独立的Agent,每个Agent均具有独立性、自主性、灵活性、智能性;行车Agent能够相互协商从而完成各自的任务。以宝钢湛江实际作业历史数据为实例验证基于MAS理论的行车调度系统的可行性,实验证明该系统能够有效地自主进行调度作业。

To solve NP difficult problem,which is a multi-vehicle and multi-task scheduling problem,the system based on MAS（Multi-Agent System） theory is applied to the simulation model of unmanned vehicle operation.The crane,the station and the task is abstracted as an independent Agent individually,which has the properties of independence,autonomy,flexibility,intelligence.These Agents can negotiate with each other to complete their tasks.Based on the practical operation data of Zhan Jiang,the feasibility of crane control system based on MAS theory is verified.The experiment results show that the system can work autonomously.