基于改进A^(*)算法的仓储环境AGV路径规划  

AGV Path Planning for Warehouse Environment Based on Improved A^(*) Algorithm

作　　者：牟德君[1] 初鹏祥 MU De-jun;CHU Peng-xiang(School of Mechanical Engineering,Yanshan University,Qinhuangdao 066004,China)

机构地区：[1]燕山大学机械工程学院,秦皇岛066004

出　　处：《自动化与仪表》2022年第4期40-45,共6页Automation & Instrumentation

摘　　要：针对大规模仓储环境AGV(automated guided vehicle)的规划路径多转向,在局部繁忙路段发生拥堵甚至死锁的问题,以总工作时间最短为目标,提出一种适用于仓储环境的改进A^(*)算法。首先,改变评价函数代价,以时间代价代替距离代价;其次,引入转向代价,减少转向,节省转向消耗,提高路径平滑性;再次,引入拥堵代价,根据不同拥堵程度选择不同系数;最后,采用自适应系数对评价函数进行加权,使前期能快速收敛,后期搜索速度变慢,避免陷入局部最优。将改进A^(*)算法与原算法进行仿真对比,改进后的A^(*)算法规划的路径转向次数减少,能躲避拥堵路段,长路径寻路时间大幅减少。The paths planned in large-scale storage environment exist too much turns and traffic jams or even deadlock in partial busy roads.Therefor,make an improvement to A^(*)algorithm which aims for the shortest total working time.First,the cost of evaluation function is replaced distance cost with time cost.Secondly,the steering cost is introduced to reduce steering,save steering consumption and improve the smoothness of the path.Thirdly,congestion costs are introduced,and different coefficients are selected according to different congestion levels.Finally,the adaptive coefficient is used to weight the evaluation function,so that the early convergence is fast,and the later search speed is slow,so as to avoid falling into local optimum.The simulation comparison between the improved A^(*)algorithm and the original algorithm shows that the improved A^(*)algorithm can reduce the number of path turns planned,avoid congested sections,and greatly reduce the long path finding time.

关 键 词：A*算法 AGV 路径规划 仓储 

分 类 号：TP242.3[自动化与计算机技术—检测技术与自动化装置]