Mobile Robot Path Tracking in Unknown Dynamic Environment

The paper proposes a cooperative system of fuzzy neural networks (FNN) and fuzzy logic control (FLC). And this system is used to on-line path tracking of a mobile robot in unknown dynamic environment, instead of the commonly used FLC to research on this problem. Firstly, the application of FNN based on the kinematics models of mobile robot is described in detail, moreover, it uses an improved back propagation (BP) algorithm to complete the network learning and training. Secondly, it utilizes FLC to fulfill the real-time avoiding obstacles. The simulation results show that the proposed cooperative system is effective and robust, meanwhile possesses better abilities of path tracking and real-time avoiding obstacles in unknown dynamic environment.