Visual Docking Control of an AUV with a Moving Station Using a Cross-Tracking Approach with Unscented Kalman Filtering

In this paper, we propose an unscented Kalman filterer (UKF)-based visual docking control system for an un-deractuated autonomous underwater vehicle (AUV) by using a position-based visual servoing approach. At first, the underwater vehicle's equations of motion and the visual system model are derived. Based on the system models, the relative position and orientation of the underwater vehicle with respect to a moving docking station are estimated by a UKF with the visual measurements of point features on the station. Then, adopting the cross-tracking methodology, the desired yaw and pitch angles are designed with the estimated pose to drive the underactuated underwater vehicle to move along the docking path, which is defined by the straight line along the center of the docking station. Finally, the attitude tracking of underwater vehicles is proposed by using non-singular terminal sliding mode control to overcome the system uncertainties in the rotational dynamic model. The stability of the attitude tracking system is briefly analyzed. Simulation experiments are provided to demonstrate the performances of the proposed visual docking controller.