Machine learning-based estimation method of seismic response of building’s unobserved floor
Daiki Kakehashi, Takenori Hida
Abstract. There is a need for technology that can automatically and immediately identify and evaluate the structural integrity of a building. This study aims to propose a method that utilizes machine learning to estimate the building responses of unobserved floors based on strong motion records. Numerical experiments were performed using a seismic response analysis model of RC super high-rise buildings to assess the estimation accuracy. As a result, the estimation accuracy was improved by increasing the amount of data used for training.
Keywords
Seismic Response, Unobserved Floor, Machine Learning, Time Series Prediction, LSTM
Published online 3/25/2025, 7 pages
Copyright © 2025 by the author(s)
Published under license by Materials Research Forum LLC., Millersville PA, USA
Citation: Daiki Kakehashi, Takenori Hida, Machine learning-based estimation method of seismic response of building’s unobserved floor, Materials Research Proceedings, Vol. 50, pp 331-337, 2025
DOI: https://doi.org/10.21741/9781644903513-39
The article was published as article 39 of the book Structural Health Monitoring
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 license. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
References
[1] National Institute for Land and Infrastructure Management, Ministry of Land, Infrastructure, Transport and Tourism; Building Research Institute: Report on the Damage Investigation of the 2011 off the Pacific coast of Tohoku Earthquake, March 2012. (in Japanese)
[2] Kozo system: SNAP Ver. 8, https://www.kozo.co.jp/index.php (accessed 2024.9.9) (in Japanese)
[3] National Research Institute for Earth Science and Disaster Resilience: Strong-motion seismograph networks (K-NET, KiK-net), https://www.kyoshin.bosai.go.jp/
