Over-actuated controller design for dynamic positioning system based on model predictive control

The dynamic positioning system (DPS) of the deepwater semi-submersible platform can only control low-frequency motions of surge, sway and yaw. DPS can be considered as an over-actuated system, because the number of propellers is more than the horizontal plane degrees of freedom. Model predictive control (MPC) which integrates the virtual control force calculation and thrust allocation was adopted. The control forces were explicitly expressed by the propellers' thrust forces, which fully considered the propellers' physical performance, and the thrust change rate was considered as the control parameters. In the design, the environment disturbance was predicted firstly based on the current dynamic states, achieving the initial distribution of environmental perturbations by using the least squares method. And then the standard model predictive control was done, which takes full consideration of the thrust loss due to thrust-thrust interaction problems. It effectively avoided thrust loss by setting the thrust angle between the upstream and downstream thrusters. Based on these designs, while the virtual control forces are gotten, they are allocated directly. Finally the effectiveness of the algorithm is verified.