Development of Path Planning Approach Using Improved A-star Algorithm in AGV System

Abstract

Automated guided vehicles (AGV) has been widely applied in industry, especially in warehousing, logistics and port transportation, etc. The most important issue in the AGV control system is how to determine the shortest path. Although A-star algorithm is usually used to solve this problem due to its fast computational time, it may suffer from some drawbacks, for instance, there are too many unnecessary inflection points and redundant nodes during its moving process. For this reason, the improved A-star algorithm is proposed to optimize the motion path, including the reduction of path length, number of AGV turns and path planning time. In this scheme, a key point selection strategy is used for the secondary planning based on the initial path obtained by the A-Star algorithm. Therefore, redundant inflection points and nodes in the path can be effectively removed. Additionally, the rotation direction and rotation angle of AGV at the inflection point can be thus determined. This improved A-star method can provide a more efficient path planning with shorter routes, less turn times and shorter operation time compared with previous methods, such as the A-star algorithm and the ant colony algorithm.