Improved A* Algorithm and Dynamic Obstacle Avoidance Method for Path Planning and Self- Positioning of Autonomous Guided Vehicles

Abstract

As Autonomous Guided Vehicles (AGV) are becoming more and more widely used, they demand safe and reliable autonomous navigation solutions to problems including how to plan a safe and efficient path based on known environmental information, how to accurately self-locate, and how to precisely avoid obstacles during the navigation process. These problems have given rise to path planning algorithms that have been the focus of research in the field of intelligent robot control. This paper develops new methods for path planning and self-positioning. Firstly, it develops the improved A* algorithm by constructing a heuristic function and increasing the weight of the heuristic cost. The improved algorithm significantly reduces the computation of the traversal of irrelevant node points and hence the computation time. Further it adds a smoothing process for unsmooth paths using Bessel curves, resulting in sufficiently smooth curve. Simulations of the algorithm show that the improved A* algorithm can significantly reduce the number of nodes traversed while ensuring optimal paths. Secondly, in terms of self-positioning, the distance between the visual tag and the robot is analyzed using the visual tag information acquired by the robot vision system to determine the position and orientation of the robot. To achieve collision avoidance during navigation in dynamic environments, a dynamic obstacle avoidance method is proposed, which combines TOF sensors to acquire local environment information; dynamic obstacle avoidance is achieved by inserting and switching tracking sub-targets. Finally, path planning tests were carried out using the AGV. Results show that the method can meet the needs of AGV applications in complex and dynamic environments.