Intelligent prediction for ship motion based on decomposition strategy

In order to solve accurate and real time forecast problem under poor information and uncertain conditions for traditional single prediction methods, an intelligent forecast model of ship motion is designed based on empirical mode decomposition (EMD) and online least squares support vector machine (OLSSVM). The different characteristics information of time series for ship motion is decomposed by EMD; the OLSSVM prediction model is built for each component; the superposition of the each component is taken as the ultimate forecasting value. The experiments of a ship´s rolling time series prediction are done. The simulation results indicate that the proposed model is able to effectively improve the forecasting accuracy and efficiency, compared with the traditional offline support vector machine forecasting model.