Geotechnics for Sustainable Infrastructure Development pp 1099-1105 | Cite as
Machine Learning Scheme of the Degree of Liquefaction Assessment only from the Health Monitoring Device Installed in Individual Wooden House
Health monitoring devices have been developing in order to estimate the damage to house and foundation ground subjected to an earthquake, in japan. However, the devices cannot estimate the degree of liquefaction because it focuses only on evaluation for damage index of wooden house. In this study, an attempt was made to estimate the degree of ground liquefaction only from the health monitoring device. Concretely, using “GAINET”, a health monitoring device developed by a house builder, is placed on the ground surface of the soil container, and the output data such as acceleration response, damage degree of structure and the pore water pressure in the ground were measured as machine learning data by applying several 3D seismic motions. In this research, a machine learning scheme evaluating the classification of liquefaction damage degree is introduced and the possibility to evaluate the liquefaction damage only from the output data obtained from health monitoring device is shown.
KeywordsDamage evaluation Seismic motion Liquefaction
Unable to display preview. Download preview PDF.
In conducting this research, we received various instructions from Professor Motoki Kazama of the Department of Civil Engineering, Graduate School of Engineering, Tohoku University. Mr. Hiroki Fujimaru who are 4th-year undergraduates of the university and Mr. Hiroaki Kabuki who are technical staff supported the experiment. Also, GAINET was lent to Misawa Home Research Institute. We would like to express my gratitude here.
- Kazama, M., Noda, T., Mori, T., and Kim, J. (2012). Overview of the Geotechnical Damages and the Technical Problems Posed after the 2011 off the Pacific Coast of Tohoku Earthquake. Geotechnical Engineering Journal of the SEAGS & AGSSEA, Vol. 43, Issue 2, pp 49–56.Google Scholar
- Mori, T., Matsushita, S., Sato, S., Uzouka, R., and Kazama, M. (2010). The influence of fill depth and old topographical shape upon the seismic response in the development. The Japan Earthquake Engineering Symposium, G02, pp 268–274.Google Scholar
- Mori, T., Kazama, M., and Sato, S. (2014). Seismic damage survey of large-scale development housing sites in Sendai city due to The Great East Japan Earthquake - The whole area exploration in 5 development housing sites. Japanese Geotechnical Journal, Vol. 9, Issue 2, pp 233–253.CrossRefGoogle Scholar
- Towhata, I., Maruyama, S., Kasuda, K., Koseku, J. Wakamatsu, K., Kiku, H., Kiyota, T., Yasuda, S., Taguchi, Y., Aoyama, S., and Hayashida, T. (2014). Liquefaction in the Kanto region during the 2011 off the pacific coast of Tohoku earthquake. Soils and Foundations, Vol. 54, Issue 4, pp 859–873.CrossRefGoogle Scholar
- Yasuda, S., Hitomo, T., and Hasimoto, T. (2004). A detailed study on the liquefaction-induced settlement of timber houses during the 2000 Tottoriken-Seibu Earthquake. Fifth International Conference on Case Histories in Geotechnical Engineering, pp 13–17.Google Scholar
- Misawa Home Co., Ltd. Outline of “GAINET” < https://www.misawa.co.jp/iot/gainet/>.
- Takada, S., and Ozaki, R. (2000). Real time prediction of liquefaction based on strong ground motion records. Journal of JSCE, No. 640, I-50, pp 99–108.Google Scholar
- Takada, S., and Ozaki, R. (2000). Development of real-time liquefaction monitoring system using neural network. Journal of JSCE, No. 640, I-50, pp 109–118.Google Scholar
- Myhajima, M., Nozu, S., Kitaura, M., and Yamamoto, M. (2000). Study on liquefaction detective method using strong ground motion records. Journal of JSCE, No. 647, I-51, pp 405–414.Google Scholar