Distance-only based navigation of Wheeled Mobile Robots with obstacle avoidance

We consider the problem of wheeled mobile robot navigation towards an unknown stationary or maneuvering target with real time obstacle avoidance using range-only measurements. We study a method for navigation and guidance termed equiangular navigation guidance (ENG) law and give mathematically rigorous proof of convergence and stability of the proposed navigation law. Using the sensory information, ENG is then modified to augmented-ENG (AENG) in order to navigate the wheeled mobile robot (WMR) towards the target and simultaneously avoid the enroute obstacles. AENG enables the robot to approach an unknown stationary or follow an unpredictable maneuvering target in a cramped environment, while preserving a safety margin from the obstacles. The performance of the proposed navigation law is confirmed with computer simulations and experiments with ActivMedia Pioneer 3-DX wheeled robot.