Underwater robotics are a reliable and efficient mean for exploration and surveys in a submarine environment. Albeit, intervention tasks, e.g. installationand maintenance, require the expansive and hazardous deployment of professional divers. Lightweight unmanned underwater vehicles equipped with a multi-degreeof-freedom manipulator, have been proposed as an alternative. However, the control of these vehicle-manipulator systems is challenging due to their non-linear high-dimensional coupled dynamics. The central problem explored in this thesis, is station keeping of an underwater vehicle under the influence of a moving manipulator. The manipulator is represented by predictable disturbing forces and moments. The proposed control scheme is a Model Predictive Control (MPC) algorithm with preview of the disturbances. In simulation, performance of the MPC scheme is evaluated for different degrees of knowledge about the disturbances. Results are compared to a classical feedback controller.
[thesis]
