After entering the 21st century, especially in recent years, the domestic logistics industry has shown an unprecedented development trend. The traditional logistics industry is undergoing tremendous changes. The status of logistics warehouses in the logistics supply chain is becoming more and more important. Upgrade to intelligence and automation. Therefore, warehousing automation came into being in logistics warehouses. As an important part of warehouse automation, automatic guided vehicles (AGVs) are also playing an increasingly important role. The overall operational performance of the multi-AGV system has a significant impact on the efficiency of the warehousing automation system, which further affects the work efficiency of the logistics warehouse, and the AGV will inevitably have collision and conflict problems during the work process, so the effective and fast AGV scheduling problem has become a problem. A problem that needs to be studied and needs to be continuously broken through. Aiming at the problem of less research on multi-AGV scheduling in warehousing automation at present, this paper mainly carries out the following work on the basis of the existing shortest path search algorithm, task allocation theoretical algorithm and scheduling theoretical algorithm: First, by consulting network materials and papers Literature research, analysis and summary of the current composition elements and development status of warehousing automation multi-AGV system. Secondly, under the premise of applying breadth-first search to solve the shortest path, the traveling salesman problem model is used to solve the shortest selection path problem in the multi-loading and multi-AGV system. At the same time, the shortest selection path is input into the Fuzzy system as a parameter to realize the task allocation optimization. Then, a dynamic locking/unlocking scheduling algorithm based on lock mechanism is proposed, which makes reasonable use of map resources and prevents path conflicts and deadlocks from occurring. At the same time, combined with the actual situation of the project requirements, the theory of the shortest path search and scheduling algorithm is comprehensively used, and the AGV control is realized in the C# .Net Framework environment. Finally, the relevant implementation details and program display interface of the AGV control system are introduced, and the simulation experiments are used to measure the performance of the improved Fuzzy System task allocation algorithm and scheduling algorithm proposed in this paper, and related analysis and further prospects are given.