Design of two-stage fuzzy controller for mobile robot using vision navigation

Abstract

Aiming at the requests of path planning in dynamic environment, real-time obstacle avoidance and trajectory tracking, a new twin-stage fuzzy controller is proposed. The controller uses the three-dimensional coordinate information obtained by robot's built-in binocular vision system, and finally achieves the two goals of path planning and trajectory tracking. First, establish the dynamic model of robots, and propose a new idea to locate the robot; Second, design the algorithm of path planning and the control discipline of robot's trajectory tracking based on fuzzy logic, and then complete the path planning and tracking meanwhile; Third, realize the robot's real-time obstacle avoidance by the method of fuzzy logic. The simulation analyzes the effectiveness and stableness of the fuzzy control logic used in the two sections above. Finally, verify the stableness of the controller by the self-made binocular moving robot platform. The result demonstrates that, the proposed tracking algorithm and control method can achieve the goal of dynamic path planning and real-time obstacle avoidance very well.