Study On Modeling And Simulation Of Manoeuvring And Control For Large Container Ship

The container ship is an important kind of transportation tools, and is toward large-scale and high-speed tendency. Ship maneuverability is closely related to navigational safety and economy, to understand the Ship maneuverability is the base for navigational safety; In order to reduce costs and decrease attrition of steering gear is required as straight line navigation and playing fewer rudders. Hence, the maneuverability and the course control problem of large container ship should be given adequate attention. This paper, based on a 5446 TEU container ship and a 9572 TEU container ship, makes a preliminary study on large container ship maneuverability with different loaded conditions and different wind conditions. On this base of it,Two Intelligent Controller based on Gauss function are designed.This paper adopts MMG modeling ideas, and calculates separately the force and torque including hull, main power plant and external disturbance. The model is also considering the low-speed field with low-speed and large drift angle, and the shallow water which is encountered in entering or leaving the port. The models are verified via turning simulation calculation compared with the data from the trial voyage condition. The result indicates that the data of the simulation is consistent with the ones of the trial voyage condition and the models are satisfied. Under different loaded conditions, some parameters, including ship draft, effective rudder area, block coefficient, wind area and so on, are obviously different. This paper makes a preliminary study on large container ship maneuverability with different loaded conditions and different wind conditions. The result of simulation indicates the turning resistance increases, the turning period is longer, the wind causes the ship turning offset, drift angle and speed ups and downs, the heavier the wind is, the more obvious the impact is.Two Intelligent Controllers based on Gauss function are designed in order to study ship course control. Firstly, it is the variable parameter PID which is able to adjust PID parameters online to improve the control performance through non-linear function according to the bias; Secondly, It is the Rbfpid, combining RBF neural network with PID, which is able to on-line identify controlled object to achieve PID parameter tuning. The design of two Controllers is based on Matlab for the use of S function to establish the relevant control module, which docks with the ship model. The results of simulation indicate the control performance and robustness of Vapid and Rbfpid is significantly better than conventional PID, and the output plot of the rudder angle is smaller and gentler. However, the introduction of nonlinear function of Vapid is just based on qualitative analysis, its control performance and robustness is a little worse than Rbfpid.