Research on dynamic task allocation model and method of multi-agent logistics AGV based on improved contract net

In order to improve order picking efficiency, using logistics AGV instead of manual operation has become a main trend. This paper studies the dynamic task allocation problem of Multi-Agent Logistics AGV based on improved contract network algorithm. Aiming at the low efficiency and scalability of task allocation in the traditional contract network, and without considering the dynamic environment, this paper first describes the dynamic task in detail. Then, we propose an improved contract network task allocation model, which connects the trust with the task acquaintance database, sets the time limit for managers to wait for bids, the bidding threshold and the buffer pool, limits the number of bids and reduces the waiting time. Finally, simulation experiments are carried out to verify the effectiveness and correctness of AGV dynamic task allocation method based on improved contract network.