Time Window Based Path Planning of Multi-AGVs in Logistics Center

The prosperous online shopping has greatly changed people's life, so a massive number of express packages have to be sorted in logistics center every day. Therefore, AGVs (automated guided vehicles) are introduced to execute the sorting tasks that are almost impossible done through human labor. In order to achieve higher transportation and sorting throughput within given period using less AGVs, we need to plan conflict-free paths for AGVs. In this paper, we focus on multi-AGV path planning problem for auto-sorting in logistics center. To solve the problem, we first abstract the actual warehouse plant into directed graph. Then, we propose a heuristic algorithm to search for free time windows of the blocks in the selected paths for each AGV, and select the path with earliest arrival time windows as its schedule path. Compared with existing work like HTN and TW-A*, experiment results show that our method has a higher transportation throughput with less failures in path searching.