多货箱机器人任务调度优化研究  

Research on Optimization of Task Scheduling for Multi Container Robots

作　　者：李园园 雷斌[1,2,3] 王喜红 Li Yuanyuan;Lei Bin;Wang Xihong(Mechatronics T&R Institute,Lanzhou Jiaotong University,Lanzhou 730070,China;Gansu Provincial Engineering Technology Center for Informatization of Logistics and Transport Equipment,Lanzhou 730070,China;Gansu Provincial Industry Technology Center of Logistics and Transport Equipment,Lanzhou 730070,China)

机构地区：[1]兰州交通大学机电技术研究所,兰州730070 [2]甘肃省物流及运输装备信息化工程技术研究中心,兰州730070 [3]甘肃省物流与运输装备行业技术中心,兰州730070

出　　处：《机电工程技术》2025年第2期100-105,共6页Mechanical & Electrical Engineering Technology

基　　金：国家自然科学基金资助项目(72061021);甘肃省自然科学基金资助项目(21JR7RA284)。

摘　　要：随着电子商务的蓬勃发展,智能仓储系统对提高系统效率和柔性化的需求越来越高。为此,基于智能仓储背景研究了货箱机器人任务调度问题,通过分析货箱机器人的特征以及任务调度的流程,将多货箱机器人任务调度问题分解为先任务分组再指派及排序两个子问题进行求解;综合考虑机器人利用率、时间、距离因素,在栅格化仓储环境中,对多货箱机器人建立了任务调度优化模型;从应对多决策问题的编码方式和提高初始种群质量和算法收敛性3个方向改进了遗传算法,设计结合聚类算法分组策略的改进遗传算法求解模型;设计了不同任务规模下的多种任务分组策略的对比实验,实验数据显示,其他两种分组比策略聚类算法分组策略下的寻优结果高出120%~200%,证明了聚类算法分组策略的优越性,通过不同算法的仿真实验数据发现改进遗传算法寻优结果对比遗传算法降低了12.27%,验证了算法及模型对提高仓储效率的有效性。With the booming development of e-commerce,the demand for intelligent warehousing systems to improve system efficiency and flexibility is increasing.Therefore,based on the background of intelligent warehousing,the task scheduling problem of container robots is studied.By analyzing the characteristics of container robots and the process of task scheduling,the task scheduling problem of multi container robots is decomposed into two sub problems:task grouping,task assignment,and task sorting.Taking into account factors such as robot utilization,time,and distance,a task scheduling optimization model is established for multi container robots in a grid storage environment.The genetic algorithm is improved from three directions:encoding methods for dealing with multi decision problems,improving initial population quality,and improving algorithm convergence.An improved genetic algorithm solution model is designed combining clustering algorithm grouping strategy.A comparative experiment is designed for multiple task grouping strategies under different task scales.The experimental data showes that compared to the strategy clustering algorithm under grouping strategies,the optimization results of the other two grouping strategies are higher by 120%to 200%,proving the superiority of the clustering algorithm grouping strategy.Through simulation experimental data of different algorithms,it is found that the optimization results of the improved genetic algorithm are reduced by 12.27%compared to the genetic algorithm,verifies the effectiveness of algorithms and models in improving warehousing efficiency.

关 键 词：货箱机器人 任务调度 改进遗传算法 聚类算法 任务分组 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]