Research On Modeling And Simulating Of Dynamic Positioning Control System Of The Semi-Submersible Vessel

Along with the more and more demand in the offshore oil industry, the deep-sea explorations and developments are becoming hotspots. Dynamic positioning (DP) is a computer controlled system to automatically maintain a vessel’s position and heading by using her own propellers and thrusters according to the measured signal to resist on the disturbance of wind, wave and current.Due to not dependent on water depth, Dynamic positioning system is substituting traditional anchor systems, such as jack-up barge and anchoring. There are many problems in marine vessel modeling and control, such as big inertia, nonlinearity, coupling between different DOF and delay in actuator and so on. Although traditional PID and LQG have achieved excellent success in general industrial process, they can not provide higher control performance character when applied to DP controller.Model predictive control is an on-line optimization technique which naturally and optimally handles multivariable processes with any set of dynamic and can explicitly impose various constraints in it calculations. The control theory is advanced and popular in modern industrial process for its advantages, such as low requirement on system model, good robustness, and the ability of handling constraints and so on. Based on DP ship "TAI AN KOU", the semi-submersible heavy lift vessels, mathematical model of DP system, observer design and controller design as well as thrust allocation system have been studied in MPC theory thoroughly in this dissertation. The main contributions of this dissertation can be summarized as follows:1. Two mathematical model of ship DP system are studied. Firstly, the concept of ship DP system is introduced in this paper. Then, the descriptions of process model and control model are presented. The process model is rather complex and expect to make real world recur, which is usually used to verify controller and train operator. The control model is comparatively simple to design the controller and tries to catch the essence of system.The control model describing the vessel dynamics is separated into a low frequency model and high frequency model. It is the basic work for further research on DP control method of this dissertation. 2. The paper studies the application of moving horizon filter to the design of dynamic positioning control system based on established mathematical model of dynamic positioning ships in three degree of freedom. Since the state information may not be measurable or it may cost a lot to measure all state variables, filtering and state estimation are important features of a DP system.Optimal moving horizon filter is designed by combining the concept of moving horizon and the structure of FIR filter, and the numerical simulation of a supply ship apart with time-invariant and time-varying mass matrix and damping matrix is implemented. The simulation experiment demonstrates the effectiveness and robustness of the moving horizon filter for plant model with parametric uncertainty by comparing moving horizon filter and Kalman filter.3. The supper controller design of DP system is studied. Several challenges exist with regard to the DP controller design, such as uncertain system model, actuator with saturation hard limitations, state or output (ship position or orientation) constraints. A robust receding horizon dual-mode control method is introduced into the design of DP controller using the property of the invariant ellipsoid. Outside the ellipsoid, the receding horizon control is employed.Once the state enters the ellipsoid, the control is switched to a linear state feedback control, which makes the state stay in the ellipsoid forever and stabilizes the system while satisfying the input, the state and the output constraints. The DP control is obtained by using semi-definite programming (SDP) which can be easily solved by means of linear matrix inequalities (LMI). At last, the simulation study is carried out to verify the feasibility and efficiency of the application of robust receding horizon dual-mode control for dynamic positioning ship.4. A model predictive, dynamic thrust allocation algorithm is developed in this paper for DP ship.The thrust allocation portion is to distribute control power among redundant thruster to meet the desired control objectives under a set of constraints. The thrust allocation portion reallocation control power when some thrusters break down without the need to redesign the overall scheme.Most existing algorithms neglect the actuator dynamics or deal with the actuator dynamic separately, thereby assuming a static relationship between outputs and plant inputs. In this paper, a dynamic control allocation scheme based on MPC that directly takes into account thruster with nonnegligible dynamic and hard constraints. Results are compared with that existing static control allocation schemes.Based on DP ship "TAI AN KOU", the semi-submersible heavy lift vessels, mathematical model of DP system, observer design and controller design as well as thrust allocation system have been studied in MPC theory thoroughly in this dissertation. Large numbers of simulation study are carried out to verify the feasibility and efficiency of the application of model predictive control for dynamic positioning ship.