Multisignal VGG19 Network with Transposed Convolution for Rotating Machinery Fault Diagnosis Based on Deep Transfer Learning

Multisignal VGG19 Network with Transposed Convolution for Rotating Machinery Fault Diagnosis Based on Deep Transfer Learning

To realize high-precision and high-efficiency machine fault diagnosis, a novel deep learning framework that combines transfer learning and transposed convolution is proposed. Compared with existing methods, this method has faster training speed, fewer training samples per time, and higher accuracy....

Full description

Bibliographic Details
Main Authors: Jianye Zhou, Xinyu Yang, Lin Zhang, Siyu Shao, Gangying Bian
Format: Article
Language:English
Published: Hindawi Limited 2020-01-01
Series:Shock and Vibration
Online Access:http://dx.doi.org/10.1155/2020/8863388
id doaj-4c8d4868d0fe4da49a711e0dae204afb
record_format Article
spelling doaj-4c8d4868d0fe4da49a711e0dae204afb2020-12-21T11:41:26ZengHindawi LimitedShock and Vibration1070-96221875-92032020-01-01202010.1155/2020/88633888863388Multisignal VGG19 Network with Transposed Convolution for Rotating Machinery Fault Diagnosis Based on Deep Transfer LearningJianye Zhou0Xinyu Yang1Lin Zhang2Siyu Shao3Gangying Bian4Graduate School of Air Force Engineering University, Xi’an 710051, ChinaAir and Missile Defence College of Air Force Engineering University, Xi’an 710051, ChinaAir and Missile Defence College of Air Force Engineering University, Xi’an 710051, ChinaAir and Missile Defence College of Air Force Engineering University, Xi’an 710051, ChinaAir and Missile Defence College of Air Force Engineering University, Xi’an 710051, ChinaTo realize high-precision and high-efficiency machine fault diagnosis, a novel deep learning framework that combines transfer learning and transposed convolution is proposed. Compared with existing methods, this method has faster training speed, fewer training samples per time, and higher accuracy. First, the raw data collected by multiple sensors are combined into a graph and normalized to facilitate model training. Next, the transposed convolution is utilized to expand the image resolution, and then the images are treated as the input of the transfer learning model for training and fine-tuning. The proposed method adopts 512 time series to conduct experiments on two main mechanical datasets of bearings and gears in the variable-speed gearbox, which verifies the effectiveness and versatility of the method. We have obtained advanced results on both datasets of the gearbox dataset. The dataset shows that the test accuracy is 99.99%, achieving a significant improvement from 98.07% to 99.99%.http://dx.doi.org/10.1155/2020/8863388
collection DOAJ
language English
format Article
sources DOAJ
author Jianye Zhou
Xinyu Yang
Lin Zhang
Siyu Shao
Gangying Bian
spellingShingle Jianye Zhou
Xinyu Yang
Lin Zhang
Siyu Shao
Gangying Bian
Multisignal VGG19 Network with Transposed Convolution for Rotating Machinery Fault Diagnosis Based on Deep Transfer Learning
Shock and Vibration
author_facet Jianye Zhou
Xinyu Yang
Lin Zhang
Siyu Shao
Gangying Bian
author_sort Jianye Zhou
title Multisignal VGG19 Network with Transposed Convolution for Rotating Machinery Fault Diagnosis Based on Deep Transfer Learning
title_short Multisignal VGG19 Network with Transposed Convolution for Rotating Machinery Fault Diagnosis Based on Deep Transfer Learning
title_full Multisignal VGG19 Network with Transposed Convolution for Rotating Machinery Fault Diagnosis Based on Deep Transfer Learning
title_fullStr Multisignal VGG19 Network with Transposed Convolution for Rotating Machinery Fault Diagnosis Based on Deep Transfer Learning
title_full_unstemmed Multisignal VGG19 Network with Transposed Convolution for Rotating Machinery Fault Diagnosis Based on Deep Transfer Learning
title_sort multisignal vgg19 network with transposed convolution for rotating machinery fault diagnosis based on deep transfer learning
publisher Hindawi Limited
series Shock and Vibration
issn 1070-9622
1875-9203
publishDate 2020-01-01
description To realize high-precision and high-efficiency machine fault diagnosis, a novel deep learning framework that combines transfer learning and transposed convolution is proposed. Compared with existing methods, this method has faster training speed, fewer training samples per time, and higher accuracy. First, the raw data collected by multiple sensors are combined into a graph and normalized to facilitate model training. Next, the transposed convolution is utilized to expand the image resolution, and then the images are treated as the input of the transfer learning model for training and fine-tuning. The proposed method adopts 512 time series to conduct experiments on two main mechanical datasets of bearings and gears in the variable-speed gearbox, which verifies the effectiveness and versatility of the method. We have obtained advanced results on both datasets of the gearbox dataset. The dataset shows that the test accuracy is 99.99%, achieving a significant improvement from 98.07% to 99.99%.
url http://dx.doi.org/10.1155/2020/8863388
work_keys_str_mv AT jianyezhou multisignalvgg19networkwithtransposedconvolutionforrotatingmachineryfaultdiagnosisbasedondeeptransferlearning
AT xinyuyang multisignalvgg19networkwithtransposedconvolutionforrotatingmachineryfaultdiagnosisbasedondeeptransferlearning
AT linzhang multisignalvgg19networkwithtransposedconvolutionforrotatingmachineryfaultdiagnosisbasedondeeptransferlearning
AT siyushao multisignalvgg19networkwithtransposedconvolutionforrotatingmachineryfaultdiagnosisbasedondeeptransferlearning
AT gangyingbian multisignalvgg19networkwithtransposedconvolutionforrotatingmachineryfaultdiagnosisbasedondeeptransferlearning
_version_ 1714988467965919232

Similar Items