Research On Path Planning And Monitoring Technology For Multi-AGV System

With the rapid development of automated manufacturing,logistics and warehousing,higher requirements have been placed on the speed of automated transportation.AGV is the key transportation equipment for transportation in automated manufacturing and logistics warehousing.This paper mainly studies the path planning and monitoring technology for multiAGV systems under large-scale application.The main work is as follows:To achieve load balancing and local congestion prevention of AGV running road network under large-scale application environment,a global path planning method using A* algorithm considering load balancing is proposed,and AGV operation cost is used to consider both regional load and running distance.The AGV running road network area load balancing is realized.The unidirectional multiple-input multiple-access network model is used to verify the simulation results.The results show that the improved algorithm can balance the load of the road network and improve the overall operating efficiency of the AGV system.For the problem of multi-AGV collisions and deadlocks that occur in the local open space,a local path planning algorithm based on Q-Learning and a local path planning algorithm based on improved Dijkstra are constructed.In the Q-Learning-based local path planning algorithm,the Q table is reduced in a way that reduces illegal actions.In the improved Dijkstra's local path planning algorithm,the AGV local region state is taken as the node,and the multi-AGV action cost is taken as the edge.The graph is gradually constructed in the algorithm search.The simulation results show that the local path planning algorithm based on Q-Learning is suitable for solving AGV collisions and deadlocks,while the local path planning algorithm based on improved Dijkstra has good performance for simple scenes and relatively complex scenes.The AGV operation strategy is optimized for a typical warehouse logistics system.In the local loop mode,AGV is used to dynamically optimize the special path at the intersection to optimize AGV operation strategy under centralized control.The simulation results show that the optimized AGV operation strategy can reduce AGV operation steps and improve the operating efficiency of multi-AGV systems.The software structure of multi-AGV monitoring system is designed and planned,and the implementation technology of key function modules is discussed.The path planning example of the actual industrial application scenario is verified.