Multi-AGV Intelligent Warehouse Optimization Problem Based On Improved Discrete Particle Swarm Algorithm Combined With A* Algorithm

In the development process of Automated Guided Vehicle(AGV),it has become a high-endurance and high-stability logistics handling tool,which can greatly reduce the cost of storage personnel.In today’s logistics and warehousing systems,as AGV is widely used,it is prone to task time imbalances and excessive obstacle avoidance,which can lead to safety accidents and significantly affect the safety index and efficiency of the entire production line.Based on this problem,this master’s thesis focuses on the fundamental problem of task scheduling and excessive obstacle avoidance caused by the replacement of manual labor with AGV.It uses an improved discrete particle swarm algorithm and A* algorithm to solve the problem of uneven task allocation and multiple AGV interaction obstacle avoidance,thereby enhancing the safety and stability of the warehouse.The specific research content of this paper is as follows:Firstly,it introduces the research background and outlines the relevant theories required for this paper,and summarizes the many problems in the current intelligent AGV warehouse.Secondly,it proposes to solve the problem of multiple AGV interaction using an improved A* algorithm,which avoids over 95% of obstacle avoidance behavior in AGV production.The discrete particle swarm algorithm is improved to solve the problem of insufficient adaptability of traditional speed vector to task scheduling.Finally,it uses A* algorithm to decode the particles of the discrete particle swarm algorithm to solve the current problems.Based on this,a representative factory in Shenyang is selected for case analysis,which can provide reference for the optimization of the entire AGV intelligent warehouse.The simulation is used to solve the problems existing in the intelligent AGV warehouse,and the feasibility of the combined algorithm is verified,which can also help the company find the appropriate number of AGV to improve the efficiency of the warehouse.The factory used in this paper is representative,and the combined algorithm can still be adapted to other intelligent AGV warehouses.This case study also proves that the combined algorithm can provide more information to solve the problems of AGV in reality,and has a certain universality.At the same time,under the conditions of using this algorithm,we can solve the problem of multiple AGV interaction and task allocation in most AGV warehouses,and hope that this paper can provide a new solution for the improvement of intelligent AGV warehouses,and provide a new reference and theoretical guidance for subsequent improvements in multiple AGV interactions.