Imbalance Detection in Low Power Turbine Through Vibration Signals and Convolutional Neural Networks

Abstract

The condition monitoring and the fault detection in wind turbines reduce the cost of repairment and maintenance tasks. An early detection of faults allows repairing before the damage is aggravated. In this article, a methodology based on convolutional neural networks and the time-frequency plane of vibration signals for the detection of three different levels of imbalance damage (low, medium, and high) is presented. In general, the methodology consists of the acquisition of vibration signals from three levels of imbalance and the condition with no damage. Then, the spectrogram function is applied to get an image from the time-frequency plane of the vibration signals. This image is segmented and analyzed by the convolutional neural network to detect the level of imbalance damage. Results show the proposal effectiveness as 100 % of accuracy is obtained.