Study On Simulation Of Manoeuvring And Intelligent Control For Large Twin-propeller Twin-rudder Ship

With the rapid development of sea transport and shipbuilding industry, the current domestic and international transportation of large-scale, high-speed trend is clearly on the maneuverability of the ship have become increasingly demanding.2013, in order to achieve a safe, efficient, energy-saving and environmental protection as the goal, to South Korea’s Daewoo Shipbuilding Maersk ordered many twin propeller twin rudder container ships up18000TEU to put into operation. Which is defined as3E class ships, in the pursuit of economic benefits, pay more attention to social and environmental benefits. For such a huge ship with twin propeller twin rudder, the ship may have more flexible autonomous maneuverability. For the modeling and control of this type of ship maneuvering movement, a scientific issue has important theoretical and engineering significance.This study employs modelling mechanism of MMG modelling ideas, researching hull, rudder, propeller and outside distraction respectively, having established mathematics model of movement on twin-propeller twin-rudder ships and simulately calculated test curve of acutual ship to verify the accuracy of the model. In view of ships’ different shipping situation, on one hand, in high speed the author adapts twin-propeller to adjust with constant speed and gang contraol, compared with atual ship on different working conditions of full load and ballast state; on the other hand, in low speed the author utilizes twin-propeller control individually to research the simulation of ship manoeuvring, and compared with another single-propeller single-rudder container ship in similar size called SHANG HAI studying the advantage of using twin-propeller and twin-rudder cooperatively in low speed.About the ships’ course keeping intelligence control, this thesis designs mainly two kinds of intelligence controllers to control ships’ course. The first kind is PID controller which employs non-linear function to adjust PID on-line on the basis of bias. The second one combines the cerebellar model articulation controller (CMAC) and traditional PID as CMAC neural network has the merits of high rate of convergency, better generalization ability, simple structure and easy soft hardware achievement. On account of CAMC neural network’s real time control fuction, there is no need to learn how to get initial weight value ahead for adapting different working condition. Sothis thesis combines CMAC and PD controller and conduct on-line real time learning. The simulation result shows that PID controller and the composite controller of CMAC and PID have good control performance, however, the composite controller of CMAC and PID has better control performance and robustness and the output of rudder angle is smaller. What’s more, the cure is more smooth. All in all, the composite controller of CMAC and PID demonstrates excellent control performance in course intelligence control aspect.