Prediction of Ship Pitching Based on Support Vector Machines

Ship pitching influences mostly ship motion, it´s important to study ship pitching modeling and prediction in order to improve ship´s seaworthiness. Based on the random character of ship movement, this paper put forward a method for prediction of ship pitching movement with SVM. Based on the phase-space reconstruction theory, the method, the characteristic, and the selecting of the key parameters in the modeling is discussed. Using support vector machine model to predict ship pitching series, the RBF kernel function is introduced, which simplified the course of solving non-linear problems, It is shown by the study case that the model proposed in the paper has better generalization and prediction accuracy.