Self-learning Fuzzy Navigation of Mobile Vehicle

This paper describes a self-learning navigation method which utilizes fuzzy logic and reinforcement learning for navigation of a mobile vehicle in uncertain environments. The proposed navigator consists of three modules: Obstacle Avoidance, Move to Goal and Fuzzy Behavior Supervisor. The fuzzy rules of the on-line obstacle avoidance are learnt through reinforcement learning. A new and powerful method is proposed to constructed these rules automatically. The effectiveness of the learning method and the whole navigator are verified by simulation.