Ship Motion Attitude Prediction Based on Empirical Mode Decomposition and Gaussian Process Regression

Under the influence of wind, waves and ocean current, ships sailing at sea will produce random and complex motions. Accurate ship motion attitude prediction offers improvements in safety and efficiency for marine operations. In this paper, a new hybrid prediction model of ship motion attitude is proposed based on empirical mode decomposition (EMD) and Gaussian process regression (GPR), called EMD-GPR model. The EMD-GPR model and the GPR model were implemented using measured rolling angle data and pitch angle data from a real ship sailing at sea. The results show that EMD processing method can effectively suppress the nonlinearity and non-stationarity of ship attitude, and the hybrid EMD-GPR model is better than the GPR model.