基于Multi-Agent的露天矿山无人运输系统的研究

Research on Driverless Transportation System in Open-pit Mines Based on Multi-Agent

针对露天矿山无人化作业的业务需求,提出一种基于多智能体的露天矿山无人运输系统。分析了该系统构架下各agent应具备的功能,并概要描述了矿山无人运输的作业流程。其中:感知agent实现矿区环境下多目标识别;路径规划agent实现全局与局部路径寻优;控制agent实现无人矿车精准的横向与纵向控制;调度agent实现运输总运量优化与品位配比;交管agent通过交管指令的决策优化,提高整个车队通行效率。最后,通过一个应用实例验证了系统的可行性与有效性。

In order to meet requirements of business requirements of driverless transportation in open pit mines,Driverless transportation system in open-pit mines based on Multi-Agent is proposed.The functions of each agent are analyzed,the Operation flow of autonomous driving truck is briefly described.Firstly,perception agent can realize multi-target recognition in mining environment;Planning agent can realize global and local path optimization;Control agent can realize the accurate lateral and longitudinal control of the driverless mine-used truck;Dispatch agent can realize the optimization of total transportation volume and grade proportions;through traffic control instructions of traffic agent,the traffic efficiency of the whole fleet is improved.An example of proto-type system is implemented to validate the effectiveness of autonomous driving system.