Efficient routing for multi-AGV based on optimized Ant-agent

AGV (Automated guided vehicle) systems have been widely used in intelligent logistics due to the efficient transportation. The routing for multi-AGV, one of the key factors influence the efficiency of AGV systems, remains a challenge if several AGVs travel in a complex environment. In this paper, we develop an efficient routing for multi-AGV by combining centralized control with decentralized control, which is based on the Ant-agent optimized by repulsive potential field (Rf-Ant-agent). The congestion of node in the Ant-agent is optimized by visibility, and a guidance factor is introduced to the path node selection model to determine the travel direction of the AGV in nodes with centralized control. Then, a new repulsive potential field function is proposed, and a scheme of varying velocity is designed to adjust the velocity of AGVs with decentralized control. Furthermore, a new transition rule is established to determine the motion state of AGVs with the combination of centralized control and decentralized control. Finally, Netlogo is used to simulate and analyze the performance of the Rf-Antagent. The simulation results indicate that the perception distance and safe distance influence the efficiency and stability of Rf-Ant-agent, respectively; relative to the Ant-agent, the collision avoidance, transportation distance and transportation efficiency of the Rf-Ant-agent improved by 16.3%, 4.4% and 21.8%, respectively.