Article
A Deep Autoencoder-Based Convolution Neural Network
Framework for Bearing Fault Classification in Induction Motors
Rafia Nishat Toma, Farzin Piltan and Jong-Myon Kim *
Citation: Toma, R.N.; Piltan, F.; Kim,
J.-M. A Deep Autoencoder-Based
Convolution Neural Network
Framework for Bearing Fault
Classification in Induction Motors.
Sensors 2021, 21, 8453. https://
doi.org/10.3390/s21248453
Academic Editors: Hamed Badihi,
Tao Chen and Ningyun Lu
Received: 21 November 2021
Accepted: 15 December 2021
Published: 18 December 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Department of Electrical, Electronics and Computer Engineering, University of Ulsan, Ulsan 44610, Korea;
rafiatoma.eceku@gmail.com (R.N.T.); piltanfarzin@gmail.com (F.P.)
* Correspondence: jmkim07@ulsan.ac.kr; Tel.: +82-52-259-2217
Abstract:
Fault diagnosis and classification for machines are integral to condition monitoring in
the industrial sector. However, in recent times, as sensor technology and artificial intelligence have
developed, data-driven fault diagnosis and classification have been more widely investigated. The
data-driven approach requires good-quality features to attain good fault classification accuracy, yet
domain expertise and a fair amount of labeled data are important for better features. This paper
proposes a deep auto-encoder (DAE) and convolutional neural network (CNN)-based bearing fault
classification model using motor current signals of an induction motor (IM). Motor current signals
can be easily and non-invasively collected from the motor. However, the current signal collected
from industrial sources is highly contaminated with noise; feature calculation thus becomes very
challenging. The DAE is utilized for estimating the nonlinear function of the system with the normal
state data, and later, the residual signal is obtained. The subsequent CNN model then successfully
classified the types of faults from the residual signals. Our proposed semi-supervised approach
achieved very high classification accuracy (more than 99%). The inclusion of DAE was found to not
only improve the accuracy significantly but also to be potentially useful when the amount of labeled
data is small. The experimental outcomes are compared with some existing works on the same
dataset, and the performance of this proposed combined approach is found to be comparable with
them. In terms of the classification accuracy and other evaluation parameters, the overall method can
be considered as an effective approach for bearing fault classification using the motor current signal.
Keywords:
bearing fault diagnosis; condition monitoring; convolution neural network (CNN); deep
autoencoder (DAE); motor current signal; residual signal
1. Introduction
Rotating machinery is among the most pervasive and substantial components of
the industrial sector. Whether the system is mechanical or electro-mechanical, one or
more rotating machines are involved; examples include motors, generators, turbines,
gearboxes, drive trains, automobile, and aircraft engines. Due to rapid industrialization
and automation, the use of complex rotating machinery has increased by a lot, which
increases the chance of multiple and significant faults occurring because of a generating
fault in any single component [
1
]. Among all the various types of rotating machinery,
induction motors (IMs) are the most commonly used because of their vigorous design, high
productivity, reliability, and low cost [
2
]. In general, the IM needs to operate uninterrupted
over a long time and under difficult operating scenarios. The operating conditions and
unfavorable environment in many cases initiate different faults and may eventually lead to
undesirable downtime, huge economic losses, and in the worse case, human causalities [
3
].
To avoid these unwanted situations, the fault diagnosis mechanism has emerged as an
important part of the prognosis and health management (PHM) techniques. Research on
the fault diagnosis of rotating machinery recently became a very popular topic, and many
significant breakthroughs were achieved because of the speedy development of artificial
Sensors 2021, 21, 8453. https://doi.org/10.3390/s21248453 https://www.mdpi.com/journal/sensors
