LAUSR

Blade and winding failure often occur in the marine current turbine (MCT), which is usually subject to marine biofouling effects. Realizing the accurate fault diagnosis of MCT is of great significance. This article proposes a novel fault diagnosis method to identify the fault states of MCT blades and windings by optimization resampled modulus feature (ORMF) and 1-D-convolutional neural networks (1-D-CNNs). First, the raw stator current measurements are transformed into a modulus signal. Then, the ORMF is proposed to tackle the problem of faint characterization due to various marine currents. Afterward, the 1-D-CNN model is adopted to learn the deeper features of the resampled modulus to conduct fault classification. The prediction accuracy of the proposed method is 99.86% for MCT prototype datasets. Meanwhile, experimental results demonstrate the effectiveness of the proposed method in MCT fault diagnosis, highlighting the superiority of our introduced framework.