Prediction of Ship Rolling Base on Ensemble Empirical Mode Decomposition
Due to nonlinear and non-stationary of ship rolling, its difficult to use a single model for prediction. The ensemble empirical mode decomposition (EEMD) is put forward into prediction of ship rolling. First, time series is decomposed into some relative stable intrinsic mode functions (IMFs) by using EEMD, which can effectively reduce interference and coupling of different information. Second, IMFs are reconstructed as three components-the high frequency component, low frequency component and trend component. Support vector machines (SVM), auto-regressive and moving average (AR) and linear regression model are respectively used according to the features of components. Finally, the superposition of forecasting results of the three components is taken as the ultimate forecasting value. Time series of some ship rolling is considered as research object. The simulation results show that relative mean error (RME) is 0.0110 and root mean square error (RMSE) is 0.0400, which illuminate the effectiveness of the proposed predictive method for the ship rolling.
ship rolling ensemble empirical mode decomposition(EEMD) prediction
Zhen-Yang Sheng-Liu Dongyang-Wang
College of Automation Harbin Engineering University Harbin 150001, China No.91404 Troops of PLA Qinghuangdao 066001, China
国际会议
秦皇岛
英文
491-494
2010-11-05（万方平台首次上网日期，不代表论文的发表时间）