Research On Path Planning Algorithm For Unmanned AGV Delivery In Campus

In recent years,industries such as e-commerce,express delivery,and food delivery have developed rapidly.Unmanned delivery AGVs(Automated Guided Vehicles)have become a key means of achieving smart delivery and have been widely used and developed.In campus environments,unmanned delivery AGVs are gradually popularized and put into use.In order to improve the driving safety and delivery efficiency of AGVs in campus environments,this paper studies the improved intelligent water drop algorithm for global path planning,the improved dynamic window algorithm for local path planning,and the multi-AGV path planning algorithm in dynamic environments.The main research contents of this paper include the following aspects:(1)Studied several commonly used map environment modeling methods including visual map method,grid map method,topological map method,and free space method,analyzed their advantages and disadvantages,and ultimately selected the grid method for map modeling.Analyzed the actual needs of path planning for unmanned delivery AGVs in the campus,established two map models through the grid method,and conducted research on AGV path planning.Finally,several common path planning algorithms were discussed and analyzed.(2)Proposed an improved intelligent water drop algorithm for global path planning.The minimum soil volume of the intelligent water drop algorithm was restricted,and combined with the idea of simulated annealing algorithm,suboptimal solutions were absorbed with a certain probability.The soil update coefficient of the intelligent water drop algorithm was adaptively improved,and the B-spline curve method was used to optimize the curve.Through simulation experiments,it was proved that the improved intelligent water drop algorithm can avoid the problem of easily falling into local optima in traditional intelligent water drop algorithm.(3)Proposed an improved dynamic window algorithm for local path planning.By integrating the improved intelligent water drop algorithm,an adaptive directional angle evaluation function that can change adaptively was designed,and a changing function was designed and added to the evaluation function of the dynamic window algorithm.Through simulation experiments,it was verified that the improved dynamic window algorithm can effectively solve the AGV path planning problem in dynamic environments.(4)Proposed a dynamic priority-based dynamic window algorithm for path planning of multi-AGV systems in dynamic environments.Corresponding solutions were proposed for different conflicts that may occur in multi-AGV systems,and a dynamic task priority update solution was proposed.Through simulation experiments,it was proved that the dynamic priority-based dynamic window algorithm is feasible and effective for path planning of multi-AGV systems in dynamic environments.