A multi-AGV fast path planning method based on improved CBS algorithm in workshops
Zhuo Zhou,Liyun Xu,Haoran Qin,Beikun Zhang,Gang Shang,Zhun Xu
DOI: https://doi.org/10.1177/09544062231181838
2023-07-01
Abstract:Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, Ahead of Print. Automatic guided vehicles (AGVs) are an important component of workshop logistics distribution systems. With the continuous increase in human cost, unmanned workshops are becoming increasingly popular, and an increasing number of AGVs are being applied to workshop logistics distribution systems. However, with the increasing number of AGVs in the workshop, planning a path quickly for each AGV is challenging, and the response time of multi-AGV fast path planning cannot meet actual requirements. To solve this problem, a multi-AGV fast path planning method based on an improved conflict-based search (ICBS) algorithm is proposed. First, the path planning problem is analyzed, and the environment expression, task, objective and constraint models are established. Subsequently, a two-layer ICBS algorithm is proposed to achieve fast path planning. At the bottom layer, an improved A* algorithm is adopted, and the selection strategy of the nodes is changed when multiple minimum values are equal. In the upper layer, the conflict number, conflict objective and conflict constraint sets are established. Three priority rules including Most Conflicts First (MCF) rule, Earliest Conflict First (ECF) rule, Single Search for Conflict Point (SSCP) rule are used. And a conflict elimination strategy is proposed. Finally, the superiority of the algorithm-solving is verified based on benchmark and actual workshop maps.
engineering, mechanical