Autonomous Vision-based Wheeled Mobile Robot Path Tracking Control

With the platform of AS-R wheeled mobile robot, this paper mainly studied the path following problem based on the self-vision and multi-sensor information fusion.Firstly, using Visual C++ 6.0 development environment, the capture, processing and recognition of the digital images was realized. It was proposed that using the slope of the traced line as the feedback information of the control system to solve the robot's line following problem. And a slope recognition's method was proposed based on the method of least squares (MLS), improving the precision of the slope. And then, aimed at the kinematic model of the two-wheeled mobile robot, based on the error between the given slope and the recognized slope and the PID control algorithm, the control input of the driven motors were calculated. And the velocity feedback was achieved to control the velocity of the driven motors also based on PID algorithm. After then a method fusing the information of vision system and PSD detector was proposed, and was used to overcome disadvantage of the path slope vanishing when the robot moved to the corner. It was solved that the problem of the WMR motion decision-making at the corner. Finally, many experiments were carried out on the AS-R mobile robot, and the effectiveness of the proposed method was shown.