A potential field and bug compound navigation algorithm for nonholonomic wheeled robots

We present a practical trajectory generation algorithm based on the potential field for nonholonomic wheeled robots with obstacle avoidance constraint. The potential field algorithm may be efficient to generate a safe trajectory in realtime. However, it has a drawback to have local minima into which the robot can get trapped. To overcome the drawback, this paper introduces the artificial potential field to force the robot bug-like motion around obstacles. In addition, we discuss the use of the collision cone to generate an efficient repulsive force by reducing effects of previously visited obstacles. The simulation shows the generation of successful trajectories for autonomous robots to satisfy its kinematic constraints.