Research on Support Vector Regression Model Based on Different Kernels for Short-term Prediction of Ship Motion

Abstract

Short-term prediction of ship motions is of great significance for improving security and efficiency of the offshore operation. Therefore, many prediction models have been proposed by researchers. In recent years, support vector regression algorithm has been proved to be effective to the prediction of ship motions. In this paper, using six degrees of freedom ships motions date collected from real ships, simulation experiments are carried out to analyze the prediction results of the SVR algorithm based on four commonly used kernel functions, and the validity and practicability of these kernel functions for the prediction of ship motions are compared.