Wind Turbine Gearbox Fault Detection based on Dilated Convolutional Neural Networks

Abstract

Nowadays, wind energy has become one of the most concerning renewable energy sources due to its low damage to nature and low cost. Wind turbines are usually deployed in remote areas and harsh environments, so they (especially gearboxes) are more likely to suffer faults. Therefore, it is necessary to vigorously develop wind turbine gearbox fault detection method to provide more accurate fault alarms. This paper proposes a fault detection method based on dilated convolutional neural network(DCNN). Compared to the traditional convolutional neural network, the dilated convolution can maintain a larger receiving field to speed up the monitoring speed, and is more suitable for real-time fault monitoring. Finally, an experimental wind turbine gearbox is used to verify the effectiveness of the method.