Mobile Robot Path Planning Method Based on an Improved A* Algorithm

Abstract

The traditional A * algorithm has a series of problems such as a blind search direction, several turning points of the planned path, and a discontinuous curvature. To solve these problems, the traditional A * algorithm was improved in this study. First, the search direction was determined according to the connection angle between the expansion node and the target point as well as the difference of the coordinate system to prevent the algorithm from searching unnecessary expansion nodes. In sequence, considering the actual movement of the mobile robot, a diagonal path was processed from the perspective of a safe distance from obstacles. Finally, a combination of the Bezier curve and improved A * algorithm was used to optimize the planned path and eliminate paths with several large corners. A simulation experiment was conducted using the MATLAB 2020b experimental platform. The results revealed that the improved A * algorithm integrating the Bezier curve had fewer expansion nodes and a shorter planned path length, and the generated path did not touch obstacles and was smoother.