Autonomous Landing of an Unmanned Underwater Vehicle using Hybrid Visual Servoing Control with Image Moments and Quaternions

In this paper, a hybrid visual servo (HVS) control method is proposed for autonomous landing of an unmanned underwater vehicle (UUV) on a surface with any attitude angles (e.g., dam or ship bottom) before the cleaning or detecting operation. The hybrid features used in the HVS controller are constructed by image moments and unit quaternion of the vehicle. The desired velocity is generated by a liner feedback control of hybrid visual errors. In the dynamic control, dynamic inversion with adaptive neural network is used to overcome uncertain disturbances. Simulation results verify the effectiveness of the proposed HVS control strategy.