Fundamental study on probabilistic generative modeling of earthquake ground motion time histories using generative adversarial networks
Abstract This study proposes a probabilistic model for earthquake ground motion prediction, named ground motion generation model, which can generate ground motion time history data directly. The ground motion generation model is based on a data‐driven technique called generative adversarial networks...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Wiley
2023-01-01
|
Series: | Japan Architectural Review |
Subjects: | |
Online Access: | https://doi.org/10.1002/2475-8876.12392 |
_version_ | 1797556608327221248 |
---|---|
author | Yuma Matsumoto Taro Yaoyama Sangwon Lee Takenori Hida Tatsuya Itoi |
author_facet | Yuma Matsumoto Taro Yaoyama Sangwon Lee Takenori Hida Tatsuya Itoi |
author_sort | Yuma Matsumoto |
collection | DOAJ |
description | Abstract This study proposes a probabilistic model for earthquake ground motion prediction, named ground motion generation model, which can generate ground motion time history data directly. The ground motion generation model is based on a data‐driven technique called generative adversarial networks, allowing generation of ground motion time history data without making assumptions about physical or statistical models. A method to quantitatively and qualitatively evaluate the performance of constructed model is also proposed and the ground motion generation model is optimized for high performance from earthquake engineering and deep learning perspectives. Numerical experiments show that our proposed model is probabilistic, approximating the probabilistic distribution of the dataset of observed records and generating realistic ground motion time histories with various characteristics in the time and frequency domains. |
first_indexed | 2024-03-10T17:05:14Z |
format | Article |
id | doaj.art-c57ca822b26e4554b7b35f87e0a07eab |
institution | Directory Open Access Journal |
issn | 2475-8876 |
language | English |
last_indexed | 2024-03-10T17:05:14Z |
publishDate | 2023-01-01 |
publisher | Wiley |
record_format | Article |
series | Japan Architectural Review |
spelling | doaj.art-c57ca822b26e4554b7b35f87e0a07eab2023-11-20T10:50:42ZengWileyJapan Architectural Review2475-88762023-01-0161n/an/a10.1002/2475-8876.12392Fundamental study on probabilistic generative modeling of earthquake ground motion time histories using generative adversarial networksYuma Matsumoto0Taro Yaoyama1Sangwon Lee2Takenori Hida3Tatsuya Itoi4Department of Architecture, Graduate School of Engineering The University of Tokyo Tokyo JapanDepartment of Architecture, Graduate School of Engineering The University of Tokyo Tokyo JapanDepartment of Architecture, Graduate School of Engineering The University of Tokyo Tokyo JapanMajor in Urban and Civil Engineering, Graduate School of Science and Engineering Ibaraki University Ibaraki JapanDepartment of Architecture, Graduate School of Engineering The University of Tokyo Tokyo JapanAbstract This study proposes a probabilistic model for earthquake ground motion prediction, named ground motion generation model, which can generate ground motion time history data directly. The ground motion generation model is based on a data‐driven technique called generative adversarial networks, allowing generation of ground motion time history data without making assumptions about physical or statistical models. A method to quantitatively and qualitatively evaluate the performance of constructed model is also proposed and the ground motion generation model is optimized for high performance from earthquake engineering and deep learning perspectives. Numerical experiments show that our proposed model is probabilistic, approximating the probabilistic distribution of the dataset of observed records and generating realistic ground motion time histories with various characteristics in the time and frequency domains.https://doi.org/10.1002/2475-8876.12392acceleration time historygenerative adversarial networksground motion predictionperformance‐based earthquake engineeringprobabilistic model |
spellingShingle | Yuma Matsumoto Taro Yaoyama Sangwon Lee Takenori Hida Tatsuya Itoi Fundamental study on probabilistic generative modeling of earthquake ground motion time histories using generative adversarial networks Japan Architectural Review acceleration time history generative adversarial networks ground motion prediction performance‐based earthquake engineering probabilistic model |
title | Fundamental study on probabilistic generative modeling of earthquake ground motion time histories using generative adversarial networks |
title_full | Fundamental study on probabilistic generative modeling of earthquake ground motion time histories using generative adversarial networks |
title_fullStr | Fundamental study on probabilistic generative modeling of earthquake ground motion time histories using generative adversarial networks |
title_full_unstemmed | Fundamental study on probabilistic generative modeling of earthquake ground motion time histories using generative adversarial networks |
title_short | Fundamental study on probabilistic generative modeling of earthquake ground motion time histories using generative adversarial networks |
title_sort | fundamental study on probabilistic generative modeling of earthquake ground motion time histories using generative adversarial networks |
topic | acceleration time history generative adversarial networks ground motion prediction performance‐based earthquake engineering probabilistic model |
url | https://doi.org/10.1002/2475-8876.12392 |
work_keys_str_mv | AT yumamatsumoto fundamentalstudyonprobabilisticgenerativemodelingofearthquakegroundmotiontimehistoriesusinggenerativeadversarialnetworks AT taroyaoyama fundamentalstudyonprobabilisticgenerativemodelingofearthquakegroundmotiontimehistoriesusinggenerativeadversarialnetworks AT sangwonlee fundamentalstudyonprobabilisticgenerativemodelingofearthquakegroundmotiontimehistoriesusinggenerativeadversarialnetworks AT takenorihida fundamentalstudyonprobabilisticgenerativemodelingofearthquakegroundmotiontimehistoriesusinggenerativeadversarialnetworks AT tatsuyaitoi fundamentalstudyonprobabilisticgenerativemodelingofearthquakegroundmotiontimehistoriesusinggenerativeadversarialnetworks |