Research On Path Planning And Task Scheduling Of AGV System

In recent years,with the in-depth development of China Manufacturing "2025" plan,the traditional manufacturing sector has set off a transformation wave aiming at improving digitalization and intelligence.In 2021,the "14th Five-Year Plan" of the National People's Congress once again put forward that "intelligent manufacturing" is the only way for the transformation and upgrading of manufacturing industry,and one of its core and key elements is the realization of intelligent logistics system.Intelligent AGV(Automatic Guided Vehicle)plays an important role in intelligent logistics system,which not only needs to be combined with production technology,but also requires the ability of thinking,perception,learning,reasoning and decision-making.Facing the demand of highly flexible manufacturing,the scheduling problem of AGV has always been a technical problem to be solved urgently in industrial field.In this paper,the problem of path redundancy and collision caused by single AGV in operation and the problem of task allocation and task sequencing in multi-AGV system are studied respectively.The main research contents are:Firstly,aiming at the selection of map modeling methods and scheduling algorithms in AGV path planning,three different modeling methods are analyzed,and a single AGV path model based on grid map is established.A* algorithm and genetic algorithm commonly used in AGV scheduling system are analyzed.Based on the good scalability of genetic algorithm,an improved genetic algorithm is proposed to solve the scheduling problem of AGV system.Secondly,aiming at the problem of single AGV path redundancy and collision,this paper adds obstacle avoidance module into traditional genetic algorithm,and adds smoothness function into fitness function to set penalty value for AGV turning angle.While ensuring the good global searching ability and robustness of genetic algorithm,it makes the searched path shorter and the turning angle larger,which makes the walking process of AGV smooth,safe and efficient.Next,aiming at the problem that single AGV can not meet the production demand in medium and large factories,the paper further studies,adding multiple AGVS into the model to form AGVS system,aiming at the shortest total walking distance of AGV trolley replenishment task,and combining the double standards of path selection and task sorting,a double-layer coding mode is proposed;At the same time,in order to avoid the clustering of genes on chromosomes in a small neighborhood,an improved genetic algorithm is proposed,which adds a variety of mutation processes.Compared with the traditional genetic algorithm,it enlarges the understanding space and prevents the generation of local optimal solutions.Then the environment is modeled and simulated by MATLAB,and compared with the basic genetic algorithm.Finally,combined with the above research contents and digital twinning technology,this paper applies Plant Simulation software to simulate virtual modeling.Through the open interface of the software,you can see all the information of the actual production in the future from the virtual environment.The change of simulation parameters can be closer to the logistics and transportation links of real factories.While verifying the optimization algorithm proposed in this paper,we can also directly observe the specific situation of the model from the two-dimensional or three-dimensional framework,so that we can fully understand the actual operation process of AGV system.