Slamming Signal Feature Extraction and Classification Based on EEMD and SVM

This paper proposes a method based on Ensemble Empirical Mode Decomposition (EEMD) and Support Vector Machine (SVM) to classify non-stationary slamming signal. Slamming events often happen on the bottom of the bow, bulge, and flare out bow when the ship sailing in the sea, slamming occurs will lead to the destruction of the hull structural strength, so analyzing slamming signal is of great importance. First of all, the signal is decomposed by EEMD, then effective features would be extracted, at last, feature vectors are sent to SVM for classification. Experiments show that the method can effectively identify slamming and non-slamming signal.