Research On Intelligent Algorithm For AGV Path Plannin

Automated guided vehicle(AGV)has been widely used in many aspects of social life.Path planning is an important issue related to the efficient and stable operation of AGV.According to characteristic of path planning,this paper studies the construction method of AGV environment map,the global path planning method based on improved ant colony algorithm,the local path planning method based on improved dynamic window algorithm,and the multi AGV dynamic path planning method based on improved genetic algorithm and Bezier curve.The detail content is as follows.Firstly,the construction method of environmental map.The principles of several map construction methods such as framework space method,free space method,cell decomposition method and grid map method are systematically introduced.Then this paper analyzed their schematic diagrams.Finally,the grid map method based on feature point extraction is selected to solve the problem of AGV environment map construction.Using the known target environment information,a grid map that can simulate the actual environmental conditions is established.Secondly,global path planning.The particle swarm optimization algorithm and ant colony algorithm based on individual weight are proposed.The adaptive adjustment factor of heuristic function is introduced to optimize the pheromone update mode of traditional algorithm.Finally,the designed simulation experiments show that the ant colony algorithm based on the particle swarm optimization algorithm and individual weight in the global path planning of AGV is feasible and effective.Thirdly,local path planning.This paper analyzes the kinematic model of AGV and proposes an improved hybrid algorithm combining the dynamic window algorithm and the gap following method.The specific idea of this method is: in order to determine the optimal safe navigation angle of AGV,this paper introduce a utility function to optimize the traditional gap following method;further,we could determine the optimal running speed by adjusting the distance evaluation function and speed evaluation function and ignoring the navigation angle evaluation function in the evaluation function of the dynamic window method.Based on the above improved method,the AGV can complete the operation safely and efficiently.Finally,the designed simulation experiment shows that this method in AGV local path planning is feasible and effective.Fourthly,multi AGVs dynamic path planning.We propose a multi AGVs dynamic path planning method based on Bezier curve and improved genetic algorithm.This paper puts forward several improvement methods for the basic links of "selection","crossover" and "mutation" of genetic algorithm.Then,Bezier curve is used to optimize the path found based on genetic algorithm.Based on the above improved method,AGV can obtain the best route and efficiency during operation.Aiming at the conflict problem of multiple AGVs,we propose a pre-collision coordination strategy based on dynamic online priority.The priority update module evaluation function is used to dynamically prioritize AGVs and the conflict resolution module is used to deal with the pre collision problem.Finally,the simulation based on dynamic and unknown obstacles shows that the method based on improved genetic algorithm and Bezier curve is feasible and effective in solving the problem of multi AGV dynamic path planning.