Significant wave height forecasting based on the hybrid EMD-SVM method

摘要

Prediction of significant wave height (SWH) is considered an effective method in marine engineering and prevention of marine disasters. Support vector machine (SVM) model has limitations in processing nonlinear and non-stationary SWH time series. Fortunately, empirical mode decomposition (EMD) can effectively deal with the complicated series. So, the SWH prediction method based on EMD and SVM is proposed by combining the advantages of both methods. A statistical analysis was carried out to compare the results of two models i.e., between the hybrid EMD-SVM and SVM. In addition, two models are used for forecasting SWH with 3, 6, 12 and 24 hours lead times, respectively. A high R value of different prediction times for the hybrid model. Results indicate that SWH prediction of the hybrid EMD-SVM model is superior to the SVM model.