Research On Path Planning And Dynamic Obstacle Avoidance Decision Of Multiple Unmanned Vehicles Based On Improved A~* Algorithm

AGV(Automated guided vehicle)is a modern handling tool with many advantages such as automation and intelligence.It is widely used in intelligent workshops and warehouses.It plays an important role in improving the efficiency of logistics and handling and reducing the cost of human resources.The path planning of the AGV system is an important research content in the application of AGV.This paper focuses on path planning and dynamic obstacle avoidance of multi-AGV systems.The main work is as follows:1.For the problem of single AGV path planning,combined with the navigation method of AGV,based on the grid map construction method,the A~* algorithm and the Dijikstra algorithm were simulated and compared.The algorithm performance in five typical environments was compared,and the better algorithm was determined.The A~* algorithm is used as the basic algorithm for path planning.2.Aiming at the problems of many times of bending,large search range and long search time in the traditional A~* algorithm,the corresponding strategies were improved,and an improved A~* algorithm was proposed.The simulation comparison between A~* and the improved A~* algorithm in different environments shows that the improved A~* algorithm has advantages in terms of search time and search range,which verifies the effectiveness of the improved A~* algorithm.3.To solve the problem of path conflicts in multi-AGV path planning,an improved A~*algorithm is used to plan a static optimal path for a single AGV,and the time window method is used to solve the path conflict between AGVs.Aiming at the collision problem between multiple AGVs,a two-stage obstacle avoidance strategy combining online monitoring and AGV active obstacle avoidance was proposed,which provided a solution for the collision problem between multiple AGVs.4.Established a multi-AGV test platform,including AGV car development,test site construction,and scheduling software development.The research and development of AGV trolley includes the main control,circuit design of stabilized voltage section,PCB design,selection of wireless communication module,obstacle avoidance module,navigation and positioning module,etc.The test site is composed of 12 squares,and the site can be flexibly configured according to test requirements.The scheduling software includes the design and development of database,path planning,monitoring interface,and wireless communication modules.5.RFID positioning test,AGV navigation tracking test,single AGV path planning test,and multiple AGV path planning and dynamic obstacle avoidance tests were carried out.The results show that: RFID positioning data is accurate,real-time,and meets positioning requirements;the introduction of PID control and turning prompts during AGV navigation tracking makes the AGV tracking on the guidance path more stable;the improved A~* algorithm can give the best results in the experiments.The optimal path is significantly better than the traditional A~* algorithm in terms of time cost consumption;the time window method and dynamic obstacle avoidance decision can effectively plan non-conflicting paths for multiple AGVs and avoid mutual collisions between AGVs.The experiments verify the effectiveness of the improved A~* algorithm on multi-AGV path planning problems.