A hybrid AR-DWT-EMD model for the short-term prediction of nonlinear and non-stationary ship motion

Reliable and accurate short-term prediction of ship motions improves the safety in ship motion related maritime operations, where the empirical mode decomposition (EMD) is widely used for its advantages like the satisfactory nonlinear and non-stationary processing capability. Therefore a hybrid autoregressive AR-DWT-EMD model is proposed for the short-term prediction of non-stationary and nonlinear ship motion. The proposed hybrid model is the combination of AR-EMD model and the discrete wavelet decomposition (DWT). Compared with the traditional AR model, hybrid AR-DWT model, and hybrid AR-EMD model, the hybrid AR-DWT-EMD model is more capable to handle the nonlinear and non-stationary signals. The prediction results suggest that better predictions are obtained by the proposed AR-DWT-EMD model than other models.