Multi-AGV Path Planning Based On Multi-objective Decision Theory

With the progress of society and the development of industrial technology,AGV(Automated Guided Vehicle)has became an important transportation tool in today's industrial production and automated storage system.Path planning technology is a core technology in AGV application,the study of AGV path planning technology has important practical significance.In this paper,the traditional path planning algorithm is improved,the multi-objective decision theory is introduced into the path planning,which improves the efficiency and reliability of AGV transportation.The feasibility of the improved algorithm is verified by simulation.The main research contents are as follows:First of all,the single AGV path planning is studied.The common modeling methods of environment map,the three wheel structure AGV and its motion mode are briefly introduced.The path planning problem of Dijkstra algorithm and A*algorithm in raster map are studied,A*algorithm is chosen as the path planning algorithm in this paper.Aiming at the defect that the path of A*algorithm planning based on Manhattan distance is too close to the obstacles and overturned,a multi-objective A*algorithm is proposed,and increase path risk and path smoothness as path evaluation index.The path evaluation index and the four tree search mechanism in the A*algorithm constitute a multi-objective decision problem.The linear weighting method is used to solve the multi-objective decision making problem,and the weights in the linear weighting method are comprehensively determined by comparison matrix method and entropy weight method.Finally,MATLAB simulation is used to verify the feasibility of A*algorithm based on multi-objective decision theory in single AGV path planning.Secondly,the multi AGV path planning is studied.Based on the multi-objective A*algorithm,the AGV running time is predicted and the path is planned with time information.Tabu list is set up and the planned path is added to tabu list.When multi AGV is used for path planning,tabu table is used to determine whether the path conflicts with other AGV paths.Path conflict is studied in raster map,and divided into two categories according to conflict characteristics.Two adjustment strategies are used to adjust the conflict,the time and distance indicators are used to evaluate the conflict adjustment plan.For multi AGV path planning,there is a problem that the distance between multiple AGV is too close.The conflict regulation strategy is improved,so that the path of each AGV in multi AGV path planning can meet the needs of multi-target,and the feasibility of multi AGV path planning algorithm is verified by MATLAB simulation.Finally,the task allocation and unpredictable conflict handling methods in multi AGV system are analyzed,and the actual process of multi AGV entry and shipment in warehouse is simulated.