A neuro-fuzzy controller for reactive navigation of a behaviour-based mobile robot

In this paper, a novel neuro-fuzzy controller is proposed for reactive navigation control of a mobile robot in complex environments with uncertainties. A fuzzy logic system is designed with three behaviours: target seeking, obstacle avoidance, and barrier following. A learning algorithm based on neural network technique is developed to tune the parameters of membership functions, which smooths the trajectory generated by the fuzzy logic system. Under the control of the proposed neuro-fuzzy model, the mobile robot can preferably avoid static and moving obstacles, and generate smooth trajectories toward the target in various situations. The effectiveness and efficiency of the proposed approach are demonstrated by simulation studies.