| The coupling effect between ship direction keeping and rolling stabilization is usually considered as non-linear function form according to experiment data. Meanwhile, the ship motion is interfered by the sea wave. So the motion of ship is a coupling, non-linear process with large disturbance. This paper designs a control system of ship direction keeping according to the non-linear motion model and studies the NN-adaptive applied to the ship motion direction keeping. The simulation result shows that this method has nice performance on attenuating the rolling as well as controlling the direction of the ship.First, ship motion equation of rearing, pitching and rolling are founded based on Newton motion rules, and the force and torque exerted by the sea are studied, including hydrodynamic force of ship body, hydrodynamic force of rudder and the disturbance of the sea wave. Modeling the hydraulic servo system of the rudder and adopting dynamic pressure damp feedback method to improve servo system performance. Pitching force, rolling force and rearing torque simulation algorithms are also offeredSecond, this paper summarizes the present development of neural network control and introduces the stabilization problem of neural network control. According to the non-linear model of ship motion direction, this paper designs a NN-PID controller and validates the control effect under the influence of 3 degrees of freedom model. To achieve ship direction/rolling control, this paper adopts NN-adaptive method to design the control system. The NN-adaptive controller includes two neural networks, one is used to identify system model, and the other is used as the controller. Both of the two neural networks are improved BP algorithm, and simulate the algorithm in MATLAB?.At last, this paper presents the simulation curves, which indicate that the control system is achieved expect target, and proves that the rudder has nice control performance on rolling as the working condition of ship direction keeping.
|
PDF Full Text Request