Research On Path Planning Of Multiple AGV

With the development of automated logistics systems and smart factories,Automated Guided Vehicle is one of the key tools of the transportation system and has been used more and more.It mainly uses the sensor to detect the surrounding environment to realize the transport link in the physical space.The autonomous movement of multiple AGV cannot be separated from the path planning technology.Its main content is that AGV can follow a specific path planning strategy to ensure that vehicles can run from the starting point to the target point without collision that includes between the AGV and the shelf,and between the multiple AGV,completing the loading and unloading of items..This paper studies the path planning problem of multiple AGV from three aspects.The main tasks include:(1)This paper introduces the idea of the partitioned path search to solve the AGV environment map problem that were rarely dealt with.First,the serial number method is used to identify the acquired environmental information on the map,and the initial environment map is obtained;then this paper uses Canopy-based K-means clustering algorithm and the regional threshold to improve the initial raster map and get a grid map of the two partitioned structures of shelf area and viable area.(2)In order to solve the collision problem between AGV and shelves,according to the characteristics of different partitions in the environment map,the straightest shortest path is used in the viable area.the fusion algorithm based on A-star algorithm and ant colony algorithm is used in the comparison of different path planning methods in the shelf area.The distance matrix of ant colony algorithm is designed to avoid collisions between AGV and shelves.By comparing the optimal parameters of the experiment,the paths of multiple partitions are consolidated and corrected to obtain the single AGV optimal path.(3)In order to solve the collision between multiple AGV,the ant colony system is used to implement the path planning of multiple AGV in the complex shelf area.The priority-based collision avoidance strategy is designed to improve the mathematical model of the ant colony system so that multiple AGV do not directly communicate with each other through the ant colony pheromone to achieve path planning.In order to further improve the coordination of the method,a collaborative strategy was designed to determine in advance whether each AGV needs to re-plan the path to ensure that the AGV is in a collision-free state.The experiment verifies that this method is feasible and integrates the output of collision-free paths for each AGV.