Research On Ship Rolling Modeling And Prediction In Multiple Conditions

Due to the influence of environmental factors such as waves, wind and ocean currents, all kinds of oscillation will appear inevitably, when ships sailing on the sea. The primary influences of ship motion are roll motion. Severe rolling may have negative effects on the seaworthiness of ship, and it is not safe to the personnel or property in ship. Therefore, in order to improve the safety of the ship sailing on the sea, the study of ship rolling motion is critical.The purpose of this dissertation is to study ship rolling, establish rolling motion model of RBF neural network and forecast the rolling motion, in multiple conditions (different heading, speed and sea state), based on a fixed form as the research object. The main research work is as follows:Firstly, a mathematical model of ship rolling is built. The statistical properties of ocean waves are studied before the analysis of wave spectrum. The equal energy sampling method is used to get simulation curve of the wave height and angle, for long-crested wave, combined with the expression of wave spectrum. Then, rolling motion equation in long-crested wave, based on the linear rolling theory, is established.Secondly, the change of ship rolling motion in multiple conditions is analyzed. The simulation of ships rolling motion of different heading, speed and sea state is carried out, based on the mathematical model of ship rolling, which have been established before. The influences of heading, speed and sea state, to the frequency response curve, statistical properties (average roll angle, roll angle of righteousness and3%reliability roll angle) and the torque of ship rolling, are analyzed.Thirdly, RBF neural network is used to build ship rolling model. The data of ship rolling, based on the mathematical model of ship rolling, in all kinds of conditions is used to train the RBF neural network to get ship rolling model of the RBF neural network. Then, test data is chosen to test the model.Finally, the forecast model of ship rolling is built. The time series of ship rolling obtained from the mathematical model are used to establish the forecast model of RBF neural network, and the ship rolling can be forecasted.