AGV Path Planning Based On A~* And DWA Algorithms

Automated Guided Vehicle(AGV),a new type of intelligent handling Vehicle,is widely used in manufacturing,warehousing and other industries because of its high reliability and good flexibility.Route planning technology is one of the key technologies for AGV to realize autonomous navigation and complete transportation tasks.This paper mainly analyzes and studies two common path planning algorithms,A~* algorithm and DWA algorithm,improves and optimizes them,and designs A fusion algorithm combining the advantages of both.The following are the particular research contents of this paper:(1)The global path planning problem of AGV is studied.Aiming at the disadvantages of traditional A~* algorithm in global path planning,such as path slanting through obstacle vertices,low search efficiency,path turning times and unsmoothness,three aspects are improved: optimizing the selection rules of child nodes;Heuristic function of optimization algorithm;The path was optimized based on Floyd algorithm,and redundant collinear nodes and turning points were removed.Simulation experiments under different environment maps show that the improved A~* algorithm not only reduces the search space,improves the search efficiency,but also reduces the path length,improves the smoothness and security of the path,which fully verifies the effectiveness and adaptability of the improved A~* algorithm in global static path planning.(2)The local path planning problem of AGV is studied.Based on DWA algorithm,the influence of parameter value of evaluation function on path planning is analyzed,and the evaluation function of DWA algorithm is improved.Finally,the improved A~* algorithm and the improved DWA algorithm are combined to realize the global dynamic path planning.According to the simulation experiment,the path planned by the fusion algorithm not only fits the global static optimal path,but also avoids the unknown obstacles in the path effectively,realizing the dynamic obstacle avoidance.(3)The fusion algorithm is verified experimentally.Firstly,the hardware and software operating platform of the smart car is built and navigation parameters are set.Then,the path planning experiment is carried out by using the smart car in the real environment.During the experiment,the smart car successfully avoided the known obstacles and unknown obstacles according to the fusion algorithm,and reached the target point accurately.Experimental results show that the fusion algorithm is feasible and effective in theory and practice.The effectiveness and feasibility of the fusion algorithm designed in this paper are verified by simulation experiments under different raster map environment and intelligent vehicle path planning experiment under real environment,and the results are good.