Deep Transfer Learning and Time-Frequency Characteristics-Based Identification Method for Structural Seismic Response
The cost of dedicated sensors has hampered the collection of the high-quality seismic response data required for real-time health monitoring and damage assessment. The emergence of crowdsensing technology, where a large number of mobile devices collectively share data and extract information of comm...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Frontiers Media S.A.
2021-02-01
|
Series: | Frontiers in Built Environment |
Subjects: | |
Online Access: | https://www.frontiersin.org/articles/10.3389/fbuil.2021.627058/full |
id |
doaj-bf8cb2d256fd41bc85c505905c3d4912 |
---|---|
record_format |
Article |
spelling |
doaj-bf8cb2d256fd41bc85c505905c3d49122021-02-12T04:38:01ZengFrontiers Media S.A.Frontiers in Built Environment2297-33622021-02-01710.3389/fbuil.2021.627058627058Deep Transfer Learning and Time-Frequency Characteristics-Based Identification Method for Structural Seismic ResponseWenjie Liao0Xingyu Chen1Xinzheng Lu2Yuli Huang3Yuan Tian4Beijing Engineering Research Center of Steel and Concrete Composite Structures, Tsinghua University, Beijing, ChinaBeijing Engineering Research Center of Steel and Concrete Composite Structures, Tsinghua University, Beijing, ChinaKey Laboratory of Civil Engineering Safety and Durability of Ministry of Education, Tsinghua University, Beijing, ChinaKey Laboratory of Civil Engineering Safety and Durability of Ministry of Education, Tsinghua University, Beijing, ChinaKey Laboratory of Civil Engineering Safety and Durability of Ministry of Education, Tsinghua University, Beijing, ChinaThe cost of dedicated sensors has hampered the collection of the high-quality seismic response data required for real-time health monitoring and damage assessment. The emergence of crowdsensing technology, where a large number of mobile devices collectively share data and extract information of common interest, may help remove such obstacles and mitigate the seismic hazard. The present study proposes a crowdsensing-oriented vibration acquisition and identification method based on time–frequency characteristics and deep transfer learning. It can distinguish the responses during an earthquake event from vibration under serviceability conditions. The core classification process is performed using a combination of wavelet transforms and deep transfer networks. The latter were pre-trained using finite element models calibrated with the monitored seismic responses of the structures. The validation study confirmed the superior identification accuracy of the proposed method.https://www.frontiersin.org/articles/10.3389/fbuil.2021.627058/fullcrowdsensingdeep transfer learningtime-frequency characteristicswavelet transformstructural seismic responses |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Wenjie Liao Xingyu Chen Xinzheng Lu Yuli Huang Yuan Tian |
spellingShingle |
Wenjie Liao Xingyu Chen Xinzheng Lu Yuli Huang Yuan Tian Deep Transfer Learning and Time-Frequency Characteristics-Based Identification Method for Structural Seismic Response Frontiers in Built Environment crowdsensing deep transfer learning time-frequency characteristics wavelet transform structural seismic responses |
author_facet |
Wenjie Liao Xingyu Chen Xinzheng Lu Yuli Huang Yuan Tian |
author_sort |
Wenjie Liao |
title |
Deep Transfer Learning and Time-Frequency Characteristics-Based Identification Method for Structural Seismic Response |
title_short |
Deep Transfer Learning and Time-Frequency Characteristics-Based Identification Method for Structural Seismic Response |
title_full |
Deep Transfer Learning and Time-Frequency Characteristics-Based Identification Method for Structural Seismic Response |
title_fullStr |
Deep Transfer Learning and Time-Frequency Characteristics-Based Identification Method for Structural Seismic Response |
title_full_unstemmed |
Deep Transfer Learning and Time-Frequency Characteristics-Based Identification Method for Structural Seismic Response |
title_sort |
deep transfer learning and time-frequency characteristics-based identification method for structural seismic response |
publisher |
Frontiers Media S.A. |
series |
Frontiers in Built Environment |
issn |
2297-3362 |
publishDate |
2021-02-01 |
description |
The cost of dedicated sensors has hampered the collection of the high-quality seismic response data required for real-time health monitoring and damage assessment. The emergence of crowdsensing technology, where a large number of mobile devices collectively share data and extract information of common interest, may help remove such obstacles and mitigate the seismic hazard. The present study proposes a crowdsensing-oriented vibration acquisition and identification method based on time–frequency characteristics and deep transfer learning. It can distinguish the responses during an earthquake event from vibration under serviceability conditions. The core classification process is performed using a combination of wavelet transforms and deep transfer networks. The latter were pre-trained using finite element models calibrated with the monitored seismic responses of the structures. The validation study confirmed the superior identification accuracy of the proposed method. |
topic |
crowdsensing deep transfer learning time-frequency characteristics wavelet transform structural seismic responses |
url |
https://www.frontiersin.org/articles/10.3389/fbuil.2021.627058/full |
work_keys_str_mv |
AT wenjieliao deeptransferlearningandtimefrequencycharacteristicsbasedidentificationmethodforstructuralseismicresponse AT xingyuchen deeptransferlearningandtimefrequencycharacteristicsbasedidentificationmethodforstructuralseismicresponse AT xinzhenglu deeptransferlearningandtimefrequencycharacteristicsbasedidentificationmethodforstructuralseismicresponse AT yulihuang deeptransferlearningandtimefrequencycharacteristicsbasedidentificationmethodforstructuralseismicresponse AT yuantian deeptransferlearningandtimefrequencycharacteristicsbasedidentificationmethodforstructuralseismicresponse |
_version_ |
1724273768387313664 |