Efficient Multi-AGV Real-time Collaborative Operation in Large-scale Intelligent Warehouses

Nowadays, large warehouses operate in an unmanned way by using automated guided vehicles (AGVs). In such a system, large number of AGVs perform their transportation tasks concurrently, leading to high probability of conflicts. It is very challenging to manage the operations of these AGVs. This paper studies the operation problem of a warehouse where an unloaded AGV can travel under the shelves, which is different from the ones studied in the literature. In this paper, the problem is described by a graphic model. Based on the model, we propose a multi-AGV real-time collaborative operation (MARTCO) method. By this method, each AGV operates autonomously and determines its traveling path itself. We design a number of priority rules and an improved A* algorithm. The AGVs collaborate by applying these rules. Each AGV can dynamically adjust its path by using MARTCO to avoid conflicts. In this way, the method is computationally efficient such that it can deal with the operation of warehouses with more than 100 AGVs. Also, with the improved A* algorithm, the number of returns and nodes to be searched is significantly reduced, which improves the throughput of the system. Large number of experiments are done to verify the proposed method.