Research on AGV Path Based on Optimal Planning

Abstract

Abstract: The ability of path planning is the embodiment of intelligent AGV (Automated Guided Vehicle) system. A* algorithm uses the method of cost consumption estimation to achieve fast computing power among many algorithms, which is widely used in AGV path planning. However, it still has the problem of local optimal path planning, and there are redundant nodes and many unnecessary inflection points in the planned path. In order to reduce the total energy consumption in the transportation path, shorten the total length and reduce the turning times of AGV, the traditional A* algorithm is further optimized by using the splitting and screening scheme, and the improved A* algorithm is proposed to make its search faster and more thoughtful in the actual working environment. On the basis of the traditional A* algorithm, turn weight is added to the heuristic function of the unknown node, and the turning consumption can be considered in the calculation and path planning process, thus reducing turning times. The scheme of task splitting can be used to select as many optimal paths as possible, in which the optimal solution can achieve fewer turns and show a relatively smooth circuit. Simulation based on ROS under Ubuntu system version, the results of contrast experiment show that the improved A* algorithm is better than the traditional algorithm in terms of search time, total path travel and number of right-angle turns, which improved the actual operation efficiency of AGV, reduced the energy consumption of AGV, and it can shorten the time of the path search planning. Improved A* Algorithm is more suitable for AGV requirements in factory environment.