Citation: Xu, Q.; Liu, C.; Yang, E.;
Wang, M. An Improved
Convolutional Capsule Network for
Compound Fault Diagnosis of RV
Reducers. Sensors 2022, 22, 6442.
https://doi.org/10.3390/s22176442
Academic Editors: Kim Phuc Tran,
Athanasios Rakitzis and
Khanh T. P. Nguyen
Received: 8 July 2022
Accepted: 23 August 2022
Published: 26 August 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 Improved Convolutional Capsule Network for Compound
Fault Diagnosis of RV Reducers
Qitong Xu
1
, Chang Liu
1,2,
* , Enshan Yang
1
and Mengdi Wang
1
1
Key Laboratory of Advanced Equipment Intelligent Manufacturing Technology of Yunnan Province,
Kunming University of Science & Technology, Kunming 650500, China
2
Faculty of Mechanical & Electrical Engineering, Kunming University of Science & Technology,
Kunming 650500, China
* Correspondence: liuchang3385@gmail.com; Tel.: +86-159-2523-5670
Abstract:
In fault diagnosis research, compound faults are often regarded as an isolated fault mode,
while the association between compound faults and single faults is ignored, resulting in the inability
to make accurate and effective diagnoses of compound faults in the absence of compound fault
training data. In an examination of the rotate vector (RV) reducer, a core component of industrial
robots, this paper proposes a compound fault identification method that is based on an improved
convolutional capsule network for compound fault diagnosis of RV reducers. First, one-dimensional
convolutional neural networks are used as feature learners to deeply mine the feature information of
a single fault from a one-dimensional time-domain signal. Then, a capsule network with a two-layer
stack structure is designed and a dynamic routing algorithm is used to decouple and identify the
single fault characteristics for compound faults to undertake the diagnosis of compound faults of RV
reducers. The proposed method is verified on the RV reducer fault simulation experimental bench,
the experimental results show that the method can not only diagnose a single fault, but it is also
possible to diagnose the compound fault that is composed of two types of single faults through the
learning of two types of single faults of the RV reducer when the training data of the compound
faults of the RV reducer are missing. At the same time, the proposed method is used for compound
fault diagnosis of bearings, and the experimental results confirm its applicability.
Keywords: compound fault diagnosis; convolutional neural network; capsule network; RV reducer
1. Introduction
Industrial robots are at the core of intelligent manufacturing [
1
]. As a core component
of industrial robots, the health of rotate vector (RV) reducers is an important factor affecting
the long-term stable operation of the industrial robots [
2
,
3
]. Different from the single
fault setting in the laboratory, the different faults are interrelated in an actual operation
environment, and compound faults are more common [
4
] and are the main reason for
the failure of the RV reducer [
5
,
6
]. The coupling of different types of single faults into
a compound fault makes them more difficult to identify [
7
] and more dangerous than a
single fault [
8
]. Therefore, it is highly significant in the realm of engineering to research
compound fault diagnoses of RV reducers.
The RV reducer is composed of a front stage of a planetary gear reducer and a rear stage
of a cycloid pinwheel reducer [
9
]. Due to its complex structure, it is a complicated process
to diagnose and identify the damaged parts. Ferrography analysis, acoustic emission
analysis, and vibration analysis are the most commonly used methods to monitor the
health status of RV reducers [
10
]. In a ferrography analysis, Peng [
11
] designed a neural
network to classify the wear particles in oil to determine the wear mode of an RV reducer.
Although this method determines the wear mode inside the RV reducer, the method is
time-consuming and cannot determine the location of a wear failure. In acoustic emission
Sensors 2022, 22, 6442. https://doi.org/10.3390/s22176442 https://www.mdpi.com/journal/sensors
