Constructing a reliable health indicator for bearings using convolutional autoencoder and continuous wavelet transform

Abstract

Estimating the remaining useful life (RUL) of components is a crucial task to enhance reliability, safety, productivity, and to reduce maintenance cost. In general, predicting the RUL of a component includes constructing a health indicator (H I) to infer the current condition of the component, and modelling the degradation process in order to estimate the future behavior. Although many signal processing and data‐driven methods have been proposed to construct the H I, most of the existing methods are based on manual feature extraction techniques and require the prior knowledge of experts, or rely on a large amount of failure data. Therefore, in this study, a new data‐driven method based on the convolutional autoencoder (CAE) is presented to construct the H I. For this purpose, the continuous wavelet transform (CWT) technique was used to convert the raw acquired vibrational signals into a two‐dimensional image; then, the CAE model was trained by the healthy operation dataset. Finally, the Mahalanobis distance (MD) between the healthy and failure stages was measured as the H I. The proposed method was tested on a benchmark bearing dataset and compared with several other traditional H I construction models. Experimental results indicate that the constructed H I exhibited a monotonically increasing degradation trend and had good performance in terms of detecting incipient faults.