Research On Multi-AGV Scheduling Strategy Based On Time Window

AGV(Automated Guided Vehicle)is a mobile wheeled robot,which belongs to intelligent transportation equipment.AGV is an important part of the current unmanned factory,which can replace manual handling tasks.With the development of the manufacturing industry and the increase of labor costs,the automated assembly line also needs the corresponding AGV scheduling system to meet the increasing task requirements.AGV can realize the efficient transportation of materials and reduce production costs,so the multi-AGV system is gradually being promoted.Based on the above background,higher requirements are put forward for the optimization of multi-AGV path planning,task scheduling and task allocation algorithms.Research on the optimization of AGV routing and scheduling in smart factories is carried out,and the advantages and disadvantages of several commonly used environmental modeling methods are summarized.The environment of the smart factory was analyzed to determine the environment modeling method using topology+raster map,and put forward the constraints of the map.Improve the pheromone feedback mechanism and pheromone update rules of the traditional ant colony algorithm.The heuristic function of the traditional A*algorithm is improved,and the congestion threshold is set;the number of turns and the steering angle are restricted.Aiming at the path planning needs of large-scale factory maps,an improved A-ant colony multi-AGV path planning algorithm is proposed.The characteristics of the task assignment problem of multiple AGVs and several commonly used task assignment algorithms are summarized,and the particle swarm algorithm used to solve the task assignment model is determined.According to the task distribution characteristics of multiple AGV systems,the particle swarm algorithm in the discrete system is improved,and the integer-composite code of the particles is used to determine the order of task completion.After that,a mathematical model of multi-AGV system task allocation was established,which considered the path cost,time cost,and task revenue and other indicators,which can effectively solve the task allocation problem of multi-AGV systems.A two-layer scheduling algorithm based on time windows is designed to solve the scheduling control problem.Analyze the conflict problem of AGV,summarize several common conflict types,and add special circumstances such as human interference.A path planning algorithm based on time windows and re-planning and real-time monitoring are proposed.By monitoring the position status information of the AGV,it can predict and avoid the possible conflicts to ensure that there is no deadlock during the operation.With the AGV simulation software OpenTCS as the platform,a multi-AGV scheduling simulation system was developed to verify the feasibility of the scheduling and path planning algorithms proposed in this paper.Finally,two AGVs were used to conduct collision avoidance and scheduling experiments in a laboratory environment,which proved the feasibility of the online monitoring dynamic path adjustment strategy.