Growing demand for ocean resources has increased the need for ocean mining; ocean development has been moving into deeper water for decades. As station-keeping became a large obstacle to the use of conventional mooring systems in deep water, dynamic positioning (DP) control was introduced. DP control systems have seen many technological advancements since their emergence. Currently, DP control is applied to vehicles that are operated underwater or at the water's surface.
Using active thrusters, DP control systems automatically maintain the position and heading of floating structures subjected to environmental disturbances. DP systems are generally composed of a controller and a thrust allocator. Most DP-controlled ships are overactuated and the allocator decides how to distribute the thrust forces to each actuator device based on mechanical or operational constraints.
In 2014, the Maritime RobotX Challenge competition was held in Singapore. A catamaran surface vessel platform was provided to each participating team; however, propulsion and control systems were to be designed by each team. Some of the challenge missions required maneuvering skills such as station-keeping and weathervaning. Therefore, two thruster configurations were designed and examined so as to achieve DP maneuvering: one approach was to adopt a bow thruster; the other approach was to use azimuth thrusters. The design and experimental demonstration of these two DP systems are separately addressed.
Among various thruster configurations, azimuth thrusters or propeller/rudder pairs tend to make the allocation problem difficult to solve, because these types of propulsion systems include nonlinear constraints. Therefore, numerical optimization is generally required to perform thrust allocation. In addition, if there exists a non-convex constraint, the allocation problem becomes more complex. To solve these types of allocation problems, more elaborate techniques may be required.
In addition, twin-thruster vessels are conditionally underactuated unless the azimuth thrusters cannot rotate a full 360 degrees. This results because the thrust allocation problem does not have a feasible solution because of the angle constraint. This underactuation makes it harder to achieve DP control of a twin-thruster vessel. A path planning or nonlinear control law may be required so as to overcome the underactuation.
This thesis presents a dynamic positioning strategy for a twin-thruster ship that is propelled by two azimuthing thrusters. The geometrical configuration of a twin-thruster vessel makes it possible to simplify the thrust allocation problem. Based on this simplification, a thrust allocation method is proposed; this method does not require a numerical optimization solver. The proposed method directly determines actuator control input, and also indicates whether there is a feasible solution. Using this information, a heuristic approach that overcomes the underactuated property of a twin-thruster vessel is proposed.
Finally, the applicability of the proposed method is demonstrated through a set of experimental studies.