Research On Ship Motion Modeling And Prediction Based On Time Series

The ship was disturbed by the environmental factors such as sea wave and ocean current.The swing existed inevitably,especially in the bad sea condition, which would bring about prodigious safe hidden trouble for ship's operation.If people could predict the ship motion attitude or motive trend advancing several seconds,it was of great significance in the instruction and expiation of the taking off and landing of carrier plane,controlling of reduction swing and the launching of guided missile on the ship.At present,many researches have been done about the ship motion attitude predictions at home and abroad.Among of these researches,Time Series Analysis Method has get more and more attention to extreme short prediction of ship motion.The most advantage of this method is that we need not know any prior information and state equation of ship motion,only making use of historical data to find rules and to predict the future data.The main jobs of this thesis are stated as follows:1.Several predicted methods are researched and compared about the ship motion predictions at home and abroad.It is determined that Autoregressive(AR) Model can been taken as the predicted model.Meanwhile the parameter estimation about AR model is researched by using least mean square(LMS) algorithm, lattice recursive least-square(LRLS) algorithm and Kalman filtering algorithm. This paper gives detailed analysis of three algorithms in convergence rate of parameter estimation,predicted precision,robust performance.The simulation results show that adaptive AR model is feasible in real-time prediction of ship motion.2.As the uncertainty of ship motion and chaotic characteristics have a close contact.In view of the nonlinearity and uncertainty of ship motion.The paper introduces phase space reconstruction theory,several methods are given about how to determine the delay time and embedding dimension.The chaotic characteristics of time series about ship motion are proved by using the largest Lyapunov exponent method.3.In view of the nonlinearity and chaotic characteristics of ship motion in random sea wave,The paper gives the second Volterra predicted model combining capability of adaptive technique and nonlinearity of Volterra series.As the truncation order of series equal to the least embedding dimension of chaotic time series,the predicted models of Volterra series based on several adaptive algorithms are compared.In the finally part,the summary to the research works of the full text is given, and the direction of further study is pointed out about the chaotic time series prediction of ship motion.