A Study on Damage Evaluation Method for Box Section Steel Columns Using Autoencoder

Yusuke Fujisaku, Hideki Naito, Tomoyuki Kimoto, Masahide Matsumura, Natsumi Takeshima, Seiji Okada, Kiyoshi Ono

研究成果: 書籍の章/レポート/Proceedings会議への寄与査読

抄録

A convolutional autoencoder (CAE)-based damage evaluation method that uses frequency response functions as features was studied. A compression test was conducted on a box-section steel column to introduce out-of-plane deformation, and vibration was excited in the loading steps. From measured vibrational characteristics, the CAE model was trained, and anomaly scores were obtained. The anomaly scores obtained from the area where local buckling has occurred increased with the progression of the loading test. It was concluded that the proposed method is useful to conduct damage evaluation and structural health monitoring on steel columns.

本文言語英語
ホスト出版物のタイトルProceedings of the 33rd International Ocean and Polar Engineering Conference, 2023
編集者Jin S. Chung, Decheng Wan, Satoru Yamaguchi, Shiqiang Yan, Igor Buzin, Hiroyasu Kawai, Hua Liu, Ivana Kubat, Bor-Feng Peng, Ali Reza, Venkatachalam Sriram, Suak Ho Van
出版社International Society of Offshore and Polar Engineers
ページ3215-3221
ページ数7
ISBN(印刷版)9781880653807
出版ステータス出版済み - 2023
イベント33rd International Ocean and Polar Engineering Conference, ISOPE 2023 - Ottawa, カナダ
継続期間: 2023 6月 192023 6月 23

出版物シリーズ

名前Proceedings of the International Offshore and Polar Engineering Conference
ISSN(印刷版)1098-6189
ISSN(電子版)1555-1792

会議

会議33rd International Ocean and Polar Engineering Conference, ISOPE 2023
国/地域カナダ
CityOttawa
Period23/6/1923/6/23

フィンガープリント

「A Study on Damage Evaluation Method for Box Section Steel Columns Using Autoencoder」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル