Vibration-Based Wear Condition Estimation of Journal Bearings Using Convolutional Autoencoders

Abstract

Predictive maintenance is considered a proactive approach that capitalizes on advanced sensing technologies and data analytics to anticipate potential equipment malfunctions, enabling cost savings and improved operational efficiency. For journal bearings, predictive maintenance assumes critical significance due to the inherent complexity and vital role of these components in mechanical systems. The primary objective of this study is to develop a data-driven methodology for indirectly determining the wear condition by leveraging experimentally collected vibration data. To accomplish this goal, a novel experimental procedure was devised to expedite wear formation on journal bearings. Seventeen bearings were tested and the collected sensor data were employed to evaluate the predictive capabilities of various sensors and mounting configurations. The effects of different downsampling methods and sampling rates on the sensor data were also explored within the framework of feature engineering. The downsampled sensor data were further processed using convolutional autoencoders (CAEs) to extract a latent state vector, which was found to exhibit a strong correlation with the wear state of the bearing. Remarkably, the CAE, trained on unlabeled measurements, demonstrated an impressive performance in wear estimation, achieving an average Pearson coefficient of 91% in four different experimental configurations. In essence, the proposed methodology facilitated an accurate estimation of the wear of the journal bearings, even when working with a limited amount of labeled data.

The Prediction of the Remaining Useful Life of Rotating Machinery Based on an Adaptive Maximum Second-Order Cyclostationarity Blind Deconvolution and a Convolutional LSTM Autoencoder

“…The proposed method consists of a transformer theory that can transform the convolutional layers and LSTM architecture into encoded features. The model is scalable to the encoded features by a mixed model of convolutional layers and LSTM [ 21 , 22 , 23 , 24 ].…”

“…This method can extract sensing data to monitor health states, preserve these benefits, overcome the overfitting of spatial fluctuations, and achieve efficient and accurate health monitoring [ 17 , 18 , 19 , 20 ]. In addition, some autoencoder theories are applied to the CNN and LSTM to improve the prediction performance [ 21 , 22 , 23 , 24 ]. A gated recurrent unit (GRU) is derived from an LSTM with a smaller number of gates, which can improve the speed of the RUL prediction, and is a dual-thread GRU (DTGRU), which can utilize parallel GRU layers for stronger prediction [ 25 ].…”

QPWS Feature Selection and CAE Fusion of Visible/Near-Infrared Spectroscopy Data for the Identification of Salix psammophila Origin

“…Convolutional autoencoder (CAE) [31][32][33] is a deep learning neural network model commonly used for feature learning and data dimensionality reduction. It combines the concepts of convolutional neural networks (CNNs) and the structure of autoencoders.…”

Marine Diesel Engine Fault Detection Based on Xilinx ZYNQ SoC

“…Myung-Kyo Seo and Won-Young Yun utilized convolutional neural networks (CNNs) to achieve over 95% accuracy in monitoring gearbox conditions, demonstrating the capability of CNNs in fault diagnosis [5]. Additionally, Cihan Ates, Tobias Höfchen, and others utilized Convolutional Autoencoders for predictive maintenance of rolling bearings, exhibiting impressive performance [6]. Jin Yan, Jian-bin Liao, and their colleagues combined second-order convolutional neural networks (QCNN) with audio and vibration signals from bearings, indicating the improved CNNs' ability to diagnose complex parameters such as vibration signals [7].…”