An automated guided vehicle conflict-free scheduling approach considering assignment rules in a robotic mobile fulfillment system

Abstract

Automated guided vehicle (AGV), as a material handling equipment, has been widely used in various applications, especially in the automated warehouse of manufacturing. The working efficiency of the warehouse is greatly influenced by the organization of the order picking process. As the picking tasks of AGVs increase in a certain period, conflicts and deadlocks of simultaneously working AGVs occur more frequently, which are also defecting factors of efficiency. In this paper, a multi-AGV scheduling approach considering assignment rules (MASA) is proposed, and a stage conflict avoidance method is designed to generate a conflict-free working route for AGVs in an automated warehouse with multiple workstations. Different from the random assignment rule of workstations in previous studies, the crowding-based assignment rule (CAR) and opening-time-based assignment rule (OTAR) are proposed to adjust the AGV working route and alleviate the congestion of AGV at the workstations. The scheduling model of AGVs is constructed to minimize the total completion time based on a space–time network, and a cooperative optimization algorithm is designed to solve the AGVs conflict-free scheduling problem. Numerical experiments verify the effectiveness and stability of the proposed algorithm. With the same working loads, CAR and OTAR result in a 5.55% and 9.56% decrease in the total completion time than the traditional random assignment rule, and the completion time is reduced by 8.01% and 12.13% in class-based tasks, respectively. MASA has great potential to improve the system efficiency of the warehouse when there are massive picking tasks and workstations.