Design and Optimization of AGV Scheduling Algorithm in Logistics System Based on Convolutional Neural Network

The automation coverage of logistics warehousing in China is still at a low level, and the sorting capacity of express delivery is relatively backward. The whole sorting process cannot be completely separated from manual work. The identification, sorting and transportation of express parcels need human assistance. In this regard, the AGV scheduling algorithm design and optimization of the logistics system based on CNN (convolutional neural network) in this paper has better convergence characteristics. Through simulation experiments, AGV2 reprogrammes the running path from the loading point 21 to the unloading point 24 at the 13th moment, avoiding conflicts with AGVu3 at node 15 at the 20th moment. Low priority AGV2 also reprogrammes the running node at the next moment. The conflicts and deadlocks are eliminated in the iterative operation of the algorithm. This experiment verifies the effectiveness of the algorithm after optimization of AGV scheduling algorithm in logistics system. In this paper, the CNN priority tasks are related to the current task waiting time to break the barrier between priority and low priority, and achieve a balance between the two, so that some low priority tasks can be well executed, and improve the quality of service of the entire low priority tasks.