A Heuristic Motion Planning Algorithm for a Mobile Robot With Nonholonomic Constraints

Abstract

This paper proposes a heuristic motion planning algorithm to enhance optimal path determination and tracking for a differential drive mobile robot within a global environment. The proposed method uses the A-star search algorithm to discover the optimal path between the start and goal vertices. This algorithm employs precise distance values and heuristic distance values to compute and expand the vertex for identifying a suitable path. When the path is generated, the movement of the robot is regulated by an attractive potential field (APF) and a repulsive potential field (RPF). The APF assists the robot in following its designed path, while the RPF is employed to avoid obstacles. Subsequently, the APF plays a role in helping the robot return to its trajectory. In case of obstacles obstruct the path, the A-star algorithm identifies a new alternative route starting from the current position to the goal, and the robot navigates along this path. During a series of scenario evaluation, the algorithm demonstrates the ability to present optimal solutions in complex environments. To validate the effectiveness of the system, the simulation is executed with the Webots R2021a software. Then, experimental verification for the designed robot demonstrates the system’s high accuracy in tracking and obstacle avoidance on a specific map.