A new local path planning approach for mobile robot with blind zone

Path planning for mobile robot with blind zone, caused by limited sensing capability, is a difficult and practical problem, from which most local path planning approach is suffering. In this study, entry point has introduced to represent the free road which may guide the robot to find the gaps between obstacles. So that, the priory detected free road which has fallen in to blind zone, could be memorized and a new local path planning approach is proposed. By the memorizing, the historical sensor information, is infact partly memorized. By tracking the entry point in blind zone and estimating the probability distribution using uncented kalman filter, the influence of blind zone is reduced. All entry points are then evaluated using a evaluate function. So that both the current sensor information and the historical sensor information are making used. Compared with the traditional local path planning approaches, this approach avoid the trap problem and the hover problem came with the blind zone. Simulations have proved the effect. © 2013 Asian Network for Scientific Information.