Abstract: To further improve safety and efficiency of the four-way shuttle storage systems,rational planning of multi-shuttle paths and real-time collision avoidance problems need to be solved urgently.To reduce the collision risk of shuttles,a four-way shuttle collision avoidance path planning algorithm based on improved A*was proposed.By characterizing and predicting the collision risks of path search nodes,the path planning algorithm gained collision avoidance capability.To solve the multi-shuttle path planning problem,a quantitative evaluation model of path quality was established based on the improved A* collision avoidance algorithm to evaluate the path cost,and the traffic conflict diagram to represent the multi-vehicle path collision risk structure.On this basis,with the goal of reducing the collision risk and shortening the path,the multi-shuttle path optimization algorithm based on the multi-objective differential evolution was proposed to solve for the optimal multi-shuttle path with balanced dual-objective.To realize real-time collision avoidance for multi-shuttle operations,a real-time collision avoidance method based on traffic conflict map analysis and collision avoidance path adjustment was proposed.Through random storage task experiments,compared to the A* algorithm without collision avoidance,the proposed method was able to reduce collision risk by 91.6% to 77.3% as the number of shuttles increases from 2 to 10,improve operational efficiency by 2.9% to 26.2%,and achieve multi-shuttle operation collision avoidance with fast response to dynamic environment changes.The experiments and applications demonstrated the effectiveness of the proposed method.

Key words: four-way shuttles, path planning, real-time collision avoidance, collision risk, path quality evaluation, traffic conflict graph, A* algorithm

摘要： 为了进一步提升四向穿梭车仓储系统的安全性和效率,合理规划多车路径以及实时避碰问题亟待解决。为了降低车辆碰撞风险,提出了基于改进A*的四向穿梭车避碰路径规划算法,通过对路径搜索节点的碰撞风险表征与预判,使得路径规划算法获得避碰能力。为了解决多车路径规划问题,基于改进A*避碰算法评价路径成本,以及交通冲突图表示多车路径碰撞风险结构,建立路径质量量化评价模型,在此基础上,以降低碰撞风险与最短路径为目标,提出了基于多目标差分进化的多车路径优化算法,求解均衡双目标的多车最优路径。为了实现多车作业实时避碰,提出基于交通冲突图分析与避碰路径调整的实时避碰方法。通过随机仓储任务试验,与不考虑避碰的A*算法相比,随着车辆数2~10逐渐递增,所提方法能降低碰撞风险91.6%~77.3%,提高作业效率2.9%~26.2%,实现快速响应动态环境变化的多车作业避碰,试验及应用表明了所提方法的有效性。

关键词: 四向穿梭车, 路径规划, 实时避碰, 碰撞风险, 路径质量评价, 交通冲突图, A*算法