A global-local path planning for mobile robot with real-time unknown obstacle avoidance

During the execution of given tasks, mobile robot is required to respond to both known and unknown obstacles in real time, but avoiding the unknown obstacles in the cluttered surroundings is difficult. This paper aims to deal with the collision avoidance problem for the mobile robot through real-time path planning via a best-position-based dynamic window approach (BP-DWA). Firstly, the traditional A* algorithm is promoted by removing the redundant path points. Then, the promoted A* algorithm is incorporated into the DWA to prevent the robot from getting trapped into local optima, while avoiding the unknown obstacles. Subsequently, a BP-DWA is developed, which provides the next best position for the mobile robot instead of the next best velocity. This improvement helps to reduce the mobile robot’s computing burden and enhance the velocity stability. Simulation comparisons between the BP-DWA and the existing dynamic window-based approaches demonstrate the effectiveness and feasibility of the proposed algorithm.