Visual Stabilization of Wheeled Mobile Robots with Unknown Visual Parameters

For a monocular vision-based wheeled mobile robot system, an image-based visual servoing control strategy is proposed. The lack of depth information and nonholonomic constraints are inherent problems of the system. In addition, the uncertain intrinsic and extrinsic camera parameters are considered which makes the visual servoing of mobile robot more challenging. With this in mind, a two-degrees-of-freedom control structure is designed to separately deal with the nonholonomic constraints and the system uncertainties. An extended state observer is constructed to estimate the uncertainties and compensate to the controller. Then a switching controller is established for the compensated system to dispose of the nonholonomic constraints. The simulation results show the good performance of the proposed scheme.