Research On Ship Fin Roll Reduction Control Strategy

When ships sail across the sea, a variety of sway motions produced by outsiders of the marine environment causes many various shipwreck. Roll motion has a great impact on a sailing ship, such as physical discomfort of staff and device damage. Evermore, it may cause ship crash and casualties. It may also bring huge economic loss. Thus, ship roll reduction control is the hot issues of particular concern. A variety of anti-rolling devices is produced. Fin roll reduction is the best active anti-rolling device with the most widely being used.The traditional anti-rolling control strategy mostly is targeted at the linear systems. The traditional control algorithms are not fully applicable to complex non-linear characteristics of ships. As a result of the uncertainties of the ship model parameters and the disturbances caused by environment, better robust performance is requested for the control algorithms. Therefore, a neural network adaptive control method which can reject the uncertainty of the non-linear mathematical model is used in the fin roll reduction system. The main contents are as follows:1. The mathematical model of ship fin roll reduction is the essential issue of ship roll reduction simulation and control. A more complex and accurate non-linear mathematical model of the fin roll reduction system and disturbance torque model which produced by wave are analyzed. It is the basic for the design for the ship roll reduction.2. Aimed at unpredictable roll angular velocity in the ship fin roll reduction uncertain non-linear systems, neural network adaptive observer is designed by using the stability theory of linear time-invariant system and RBF neural network which has the ability to approximate nonlinear function. Then, neural network adaptive output feedback controller is designed by using backstepping method. It is proved that all error signals in closed-loop system are uniformly and ultimately bounded by using the Lyapunov stability theory. Finally, the effectiveness of the algorithm is verified by using MATLAB simulation. It is proved that the algorithm is effective and has good control effect.