Navigation strategy of a mobile robot using hierarchical fuzzy-neural network in dynamic environment

This paper proposes a hierarchical fuzzy-neural algorithm for navigation of a mobile robot in dynamic environment, tl-tat is, fuzzy algorithm for its speed control and fuzzy-neural network for collision avoidance. In addition, by predicting the positions of the robot and moving obstacles, we reduces the possible collision of them. Some computer simulation results show the effectiveness of the suggested navigation algorithm. To enhance the practical feasibility, random noise within reasonable scale to consider the slip, of the wheels, backlash of the gear and inaccuracy of information obtained by ultrasonic sensors is applied to show the robustness of the proposed algorithm.