基于动态避障优先级与预测的多AGV无冲突路径规划

Multi-AGVs Conflict-free Path Planning Based on Dynamic Obstacle Avoidance Priority and Prediction

扫码查看

原文链接

NETL
NSTL
万方数据

中文摘要：为实现多自动导引运输车(AGV)在机器人履约系统(RMFS)中的高效低成本运行,提出一种无冲突路径规划策略.对RMFS的仓储区域进行栅格处理,考虑重载和空载运行成本差异,细化不同运行状态包括匀速、加减速、等待、转弯等的成本,建立多AGV运行成本最小的路径规划模型.通过避障成本动态设置避障优先级及通过预测AGV在拣选工作区排队等待状态确定AGV重载出发时刻,结合改进A*算法,实现多AGV无冲突路径求解.仿真结果表明:从AGV的运行成本角度,考虑动态避障优先级和重载出发时刻预测的多AGV路径规划策略,降低了系统运行成本,缩短了任务完成时间,避免了多AGV在RMFS中的碰撞及拥堵,实现了降本增效.

外文摘要：In order to realize the efficient and low-cost operation of multi automated guided vehicle(AGV)in robot mobile fulfillment systems(RMFS),a conflict free path planning strategy is proposed.The storage area of RMFS is rasterized.A path planning model with the minimum operation cost of multi-AGVs is established,considering the differences in operation costs of heavy load and no-load,and the cost of different operation states,including constant speed,acceleration and deceleration,waiting and turning,is refined.The multi-AGVs conflict free path model is solved by dynamically setting the obstacle avoidance priority through the obstacle avoidance cost and by determining the AGV loading departure time after predicting the queue waiting state of AGV in the picking work area,combined with the improved A*algorithm.The simulation results show that,from the perspective of AGV operation cost,the multi-AGVs path planning strategy considering dynamic obstacle avoidance priority and loading departure time prediction reduces the system operation cost,shortens the task completion time,avoids the collision and congestion of multi-AGVs in RMFS,then realizes cost reduction and efficiency increase.

外文关键词：

Multi-AGVs Conflict-free Path PlanningDynamic Obstacle Avoidance Priority StrategyLoading Departure TimeRobotic Mobile Fulfillment Systems

作者：

李腾、丁佩佩、张茹兰

展开 >

作者单位：

哈尔滨商业大学管理学院,黑龙江哈尔滨 150028

关键词：

多AGV无冲突路径规划动态避障优先级重载出发时刻 RMFS

基金：

国家科技支撑项目黑龙江省自然科学基金资助项目黑龙江省博士后科研启动金项目

项目编号：

2018YFB1402500LH2023G009LBH-Q21102

出版年：

2024

系统工程

湖南省系统工程与管理学会

系统工程

CSTPCD北大核心

影响因子：0.721

ISSN：1001-4098

年,卷(期)：2024.42(3)