Citation: Lei, X.; Lu, N.; Chen, C.;
Wang, C. An AVMD-DBN-ELM
Model for Bearing Fault Diagnosis.
Sensors 2022, 22, 9369. https://
doi.org/10.3390/s22239369
Academic Editor: Jongmyon Kim
Received: 6 November 2022
Accepted: 29 November 2022
Published: 1 December 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 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/).
Article
An AVMD-DBN-ELM Model for Bearing Fault Diagnosis
Xue Lei
1
, Ningyun Lu
1,2,
*, Chuang Chen
3
and Cunsong Wang
4
1
College of Automation Engineering, Nanjing University of Aeronautics and Astronautics,
Nanjing 211106, China
2
State Key Laboratory of Mechanics and Control of Mechanical Structures,
Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China
3
College of Electrical Engineering and Control Science, Nanjing Tech University, Nanjing 211816, China
4
Institute of Intelligent Manufacturing, Nanjing Tech University, Nanjing 210009, China
* Correspondence: luningyun@nuaa.edu.cn
Abstract:
Rotating machinery often works under complex and variable working conditions; the
vibration signals that are widely used for the health monitoring of rotating machinery show extremely
complicated dynamic frequency characteristics. It is unlikely that a few certain frequency components
are used as the representative fault signatures for all working conditions. Aiming at a general
solution, this paper proposes an intelligent bearing fault diagnosis method that integrates adaptive
variational mode decomposition (AVMD), mode sorting based deep belief network (DBN) and
extreme learning machine (ELM). It can adaptively decompose non-stationery vibration signals into
temporary frequency components and sort out a set of effective frequency components for online fault
diagnosis. For online implementation, a similarity matching method is proposed, which can match
the online-obtained frequency-domain fault signatures with the historical fault signatures, and the
parameters of AVMD-DBN-ELM model are set to be the same as the most similar case. The proposed
method can decompose vibration signals into different modes adaptively and retain effective modes,
and it can learn from the idea of an attention mechanism and fuse the results according to the weight
of MIV. It also can improve the timeliness of the fault diagnosis. For comprehensive verification of
the proposed method, the bearing dataset from the University of Ottawa is used, and some recent
methods are repeated for comparative analysis. The results can prove that our proposed method has
higher reliability, higher accuracy and higher efficiency.
Keywords:
bearing fault diagnosis; variable working conditions; adaptive VMD; mode sorting; DBN-ELM
1. Introduction
Bearing fault diagnosis plays a crucial role in rotary machines. Accurate detection and
isolation of early-stage bearing faults will contribute to more safe and more efficient operation
of the rotary machines [1]. An analytic model based approach is an important research issue
that is widely applied on fault diagnosis of rotating machinery [
2
]. At present, data-driven
methods (including signal processing, statistical analysis and various advanced AI-model-
based methods) are widely applied to bearing health monitoring and fault diagnosis as they
can make full use of the vibration information in rotary machines [3–16].
However, the many existing methods are supposed to handle the dataset collected
from constant working conditions (i.e., with constant speed or working load). In practice,
the working conditions of rotary machines are complex and varying, so the vibration
signals are often non-stationary. In recent years, some fault diagnosis methods have been
proposed under variable working conditions [17,18].
As an example, Figure 1a,b illustrates a vibration signal and its envelope spectrum when
an inner raceway fault occurs and the speed setting is a constant value;
Figures 2 and 3
show
the vibration signals for the cases with two different variable speed settings. It is easy to
observe the difference among them. The fault diagnosis model developed for the scenery
with constant working conditions is rarely applicable to other sceneries.
Sensors 2022, 22, 9369. https://doi.org/10.3390/s22239369 https://www.mdpi.com/journal/sensors
