Study On Multi-Dimensional Taylor Network Optimal Control For Ship Roll Stabilization

Ship sailing at sea is subject to a variety of interference of marine environmental such as wind, current, wave, therefore unavoidable swaying motions appear on the hull. Among motions, the impact of rolling motion is the most remarkable. Severe roll makes the crew not work, reduces work performance of weapons and equipment, and has an impact on ship safety seriously. Thus, effectively reducing ship swaying motions is essential, and would also increase the life of the ship. At the same time, the coupling between roll motion and yaw motion is considered in this thesis. Ship movement also has the characteristics of nonlinearity and uncertainty, and the study about control problems of uncertain nonlinear systems is of important theoretical significance and practical need.In this thesis, a control method of ship roll stabilization based on multi-dimensional Taylor network (MTN) optimal control is proposed, and compares with the traditional PID control and sliding model control, by MATLAB simulation to summarize their advantages and disadvantages. The main contents include:1. Dynamics mathematical model of the ship is analyzed, while the coupling between roll motion and yaw motion is considered. Sway, roll and yaw three degrees of freedom simulation model of ship is established, and then wave interference model is analyzed. In order to conform with the actual situation, the limiter of amplitude and speed are joined into steering engine and stabilizer.2. PID controller of ship roll stabilization system is designed, and then MATLAB/ Simulink built-in optimization tool is used to optimize the parameters of PID control. Since sliding model control can play a good control effect to general nonlinear systems, nonlinear variable structure controller of ship roll stabilization system is designed.3. MTN controller of ship roll stabilization system is designed, and then the principle of minimum to get the optimal control signal of the controlled model and the conjugate gradient method to optimize the parameters of MTN controller are introduced. Due to time reason, MTN controller is designed based on the PID control, then the parameters of MTN controller is optimized by the conjugate gradient method with the whole process in this thesis.4. A system simulation platform is designed, and MATLAB Graphical User Interface (GUI) is used to design system user interface. In one wave interference, the parameters of each controller in three degrees of freedom model of ship are optimized, and then the effect of three control methods is compared. Then, the wave interference simulation is finished in other sea conditions. By analyzing the results of each sea state simulation, that MTN control is better in most selected wave interference and has better robustness, but PID control is poor relatively, and sliding model control has the problem of stability for the disturbance is showed.5. The work done is concluded in this thesis, and the issues subject for further research is discussed.