智能仓储多机器人动态任务分配方法  

Research on Dynamic Task Allocation Method for Multi-robot Intelligent Warehouse

作　　者：陈明智 Chen Mingzhi(School of Computer and Electronic Information Engineering,Wuhan City Polytechnic,Wuhan 430070,Hubei)

机构地区：[1]武汉城市职业学院计算机与电子信息工程学院,湖北武汉430070

出　　处：《武汉工程职业技术学院学报》2024年第1期31-35,共5页Journal of Wuhan Engineering Institute

摘　　要：针对智能仓储系统中多机器人任务分配方法进行研究,根据智能仓储中物流任务数量庞大且实时新增的动态特点,创新地提出了一种总体分配并根据当前工况实时插入的动态任务分配方法,将一定时间阈值内的物流任务集合利用综合时间代价、路径代价、空载代价和任务均衡度代价的自适应多层编码遗传算法进行任务整体规划,为智能仓储系统中每个机器人分配需执行的物流任务序列及执行顺序。任务执行过程中,各机器人通过对当前自身剩余任务量成本及插入实时任务后路径成本的判断,在不影响当前任务完成总时间的情况下,对下一时间阈值内产生的实时物流任务是否插入当前执行任务序列展开竞争,实现任务的实时插入,未被实时插入的物流任务则等待下一次总体分配。实验结果表明,该方法能在缩短系统运行总时间和总路程的基础上,极大的降低机器人无任务空载等待的时间,提高系统运行效率。This study focuses on the task allocation methods for multi-robot systems in intelligent warehousing.Considering the large quantity of logistics tasks and the dynamic nature of real-time task additions in intelligent warehousing,a novel approach is proposed.It introduces an overall allocation method that dynamically inserts tasks based on current conditions.Within a certain time threshold,a logistics task set is planned using an adaptive multi-layer encoding genetic algorithm that considers comprehensive time cost,path cost,idle cost,and task balance cost.This method allocates logistics task sequences and execution orders for each robot in the intelligent warehousing system.During task execution,each robot evaluates its remaining task cost and the path cost of inserting real-time tasks.Without affecting the total completion time of current tasks,robots compete to decide whether to insert real-time logistics tasks generated within the next time threshold into the current task sequence,enabling real-time task insertion.Tasks not inserted in real-time wait for the next overall allocation.Experimental results demonstrate that this method significantly reduces idle waiting time for robots,improves system efficiency,and reduces overall operation time and distance traveled.

关 键 词：智能仓储系统 机器人集群 动态任务分配 自适应多层编码遗传算法 实时插入 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP242[自动化与计算机技术—控制科学与工程]