首页|离散型制造车间AGV小车自主避障路径选择研究

离散型制造车间AGV小车自主避障路径选择研究

扫码查看
离散型制造车间分工区域多,导致工件运输环境复杂,自动导向车(Automated Guided Vehicle,简称AGV小车)极易发生碰撞事故,提出离散型制造车间AGV小车自主避障路径选择方法.采用栅格地图法将离散型制造车间划分为若干栅格,构建路径长度、转角和、危险率、避障等待时间等目标函数,建立AGV小车多目标路径规划评价函数.利用熵权法确定各目标的权重系数,设置动态环境中AGV小车避障约束条件.通过粒子群优化算法和鸽群优化算法进行AGV小车运行路径联合寻优,得到最优路径集合.将评价值最低的路径作为AGV小车运行路径,实现AGV小车自主避障路径选择.实验结果表明,所提方法选择出来的AGV小车运行路径更为平滑、长度更短且规划用时更少.
Research on Autonomous Obstacle Avoidance Path Selection of AGV in Discrete Manufacturing Workshop
There are many division areas in discrete manufacturing workshops,resulting in complex workpiece transportation en-vironment,and automatic guided vehicle(AGV for short)is very prone to collision accidents.An autonomous obstacle avoidance path selection method of AGV in discrete manufacturing workshop is proposed.The discrete manufacturing workshop is divided into several grids by grid map method.Objective functions such as path length,angle sum,hazard rate and obstacle avoidance waiting time were constructed,and the multi-objective path planning evaluation function of AGV is established.The entropy weight method is used to determine the weight coefficients of each target,and the constraint conditions for AGV obstacle avoid-ance in dynamic environment are set.Through particle swarm optimization algorithm and pigeon swarm optimization algorithm,the AGV running path is jointly optimized,and the optimal path set is obtained.The path with the lowest evaluation value is taken as the running path of the AGV to realize the autonomous obstacle avoidance path selection of the AGV.The experimental results show that the AGV path selected by the proposed method is smoother,shorter and takes less time to plan.

Discrete Manufacturing WorkshopAGVAutonomous Obstacle AvoidancePath SelectionPigeon Swarm Optimization Algorithm

付涛、景兴淇、李正雄、任涛

展开 >

机械工业第六设计研究院有限公司,河南 郑州 450000

合肥工业大学,安徽 合肥 230009

离散型制造车间 AGV小车 自主避障 路径选择 鸽群优化算法

国家重点研发计划"网络协同制造和智能工厂"重点专项项目(2020)国家重点研发计划子课题

2022YFB17084002022YFB1708402

2024

机械设计与制造
辽宁省机械研究院

机械设计与制造

CSTPCD北大核心
影响因子:0.511
ISSN:1001-3997
年,卷(期):2024.398(4)
  • 10