Automatic feature extraction of time-series applied to fault severity assessment of helical gearbox in stationary and non-stationary speed operation

Abstract

Signals captured in rotating machines to obtain the status of their components can be considered as a source of massive information. In current methods based on artificial intelligence to fault severity assessment, features are first generated by advanced signal processing techniques. Then feature selection takes place, often requiring human expertise. This approach, besides time-consuming, is highly dependent on the machinery configuration as in general the results obtained for a mechanical system cannot be reused by other systems. Moreover, the information about time events is often lost along the process, preventing the discovery of faulty state patterns in machines operating under time-varying conditions. In this paper a novel method for automatic feature extraction and estimation of fault severity is proposed to overcome the drawbacks of classical techniques. The proposed method employs a Deep Convolutional Neural Network pre-trained by a Stacked Convolutional Autoencoder. The robustness and accuracy of this new method are validated using a dataset with different severity conditions on failure mode in a helical gearbox, working in both constant and variable speed of operation. The results show that the proposed unsupervised feature extraction method is effective for the estimation of fault severity in helical gearbox, and it has a consistently better performance in comparison with other reported feature extraction methods.

Experimental Research on On-line Monitoring and Compensation Algorithm of 3D Printing Based on Machine Vision

“…The time required for 3D printing changes along with the workpiece size and modeling parameters, usually taking certain time, some even taking dozens of hours. However, the operator couldn't monitor the whole printing process by eyes on site [3]. If any abnormal conditions such as extruder head blockage, filament break, accumulation skew and height error occur during the printing process, existing 3D printers couldn't recognize them in time and will keep running, which will result in bad consequences such as printing failure, material waste and development cycle delay.…”

“…3. Set each layer thickness as t, the theoretical height of the layer n h 0 =t*n, and the actual height is h 1 .…”

Vibration Signals Analysis by Explainable Artificial Intelligence (XAI) Approach: Application on Bearing Faults Diagnosis

“…The statistical features mentioned above also can be applied to build automatic diagnosis models [17][18][19]. Recently, CNNs are applied in some researches for classification and prediction, in which the features can be extracted automatically [20,21]. If a two-dimensional CNN is utilized for vibration signals analysis, the inputs should be chosen as time-frequency spectra, grey level images of signals, or other two-dimensional data or images [22][23][24].…”

Generative Transfer Learning for Intelligent Fault Diagnosis of the Wind Turbine Gearbox

“…To overcome these problems, end-to-end deep learning structures, such as convolutional neural network (CNN) [5,6], echo state network [7][8][9], extreme learning machine [10] and sparse autoencoder [11][12][13], have been proposed and have drawn much attention in machinery fault diagnosis. Moreover, the optimization algorithms [14][15][16] in machine learning also keep pace with the times.…”