Hierarchical planning for multiple AGVs in warehouse based on global vision

Abstract

In automated warehouses, multiple automated guided vehicles (AGVs) are widely used to sort parcels for improving storage efficiency. However, as the number of AGVs increases, the scheduling complexity also increases dramatically. This paper proposes a hierarchical planning algorithm for efficiently coordinating a fleet of AGVs in warehouses based on global vision. We use the camera installed on the top of the warehouse to capture the global image that is used to track AGVs. Based on a network congestion diffusion model, we quantify the road congestion and add it to the evaluation index to form a time-varying dynamic evaluation function. Then the improved A* algorithm and time window algorithm are combined as the hierarchical planning algorithm to search the idle path and avoid collisions. Extensive simulations show that the hierarchical planning based on global vision not only effectively schedules multiple AGVs in the warehouse, but also has lower time complexity than other real-time path planning algorithms.