CRACK PROPERGATION MODEL OF CONCRETE BRIDGES IN SHIKOKU BASED DATA DRIVEN APPROACH

• OKAZAKI Yuriko

  香川大学 創造工学部

• OKAZAKI Shinichiro

  香川大学 創造工学部

• CHUN Pang-jo

  愛媛大学 大学院理工学研究科生産環境工学専攻

• ASAMOTO Shingo

  埼玉大学 大学院理工学研究科

• OKUBO Kazuaki

  愛媛大学 防災研究情報センター

Bibliographic Information

Other Title

• データ駆動型アプローチによる四国内コンクリート橋梁のひびわれ回帰モデルの構築
• データ クドウガタ アプローチ ニ ヨル シコク ナイ コンクリート キョウリョウ ノ ヒビワレ カイキ モデル ノ コウチク

Abstract

 The purpose of this study is to construct a regression model on crack propagation for concrete girder members of the road bridges in Shikoku region. The model is constructed by a data-driven type machine learning, Gaussian process regression, based the inspection data of 1,344 bridges owned by the government. The data includes the dimensional data of bridges, the environmental conditions, traffic volume, and crack damage ranks. As the result of the factor analysis by the model, it is revealed that the influence of rainfall amount varies at the completion year of construction 1980 and it greatly affects RC girder of the bridges constructed after 1980. It means that the model can extract the local trend.

Journal

• Journal of Japan Society of Civil Engineers, Ser. F4 (Construction and Management)

  Journal of Japan Society of Civil Engineers, Ser. F4 (Construction and Management) 74 (2), I_107-I_118, 2018

  Japan Society of Civil Engineers

Keywords

• concrete crack
• maintenance
• machine learning
• Data-driven approach
• Gaussian Process Regression model
• Shikoku region

Details

• CRID: 1390564238061431552
• NII Article ID: 130007556052
• DOI: 10.2208/jscejcm.74.i_107
• ISSN: 21856605
• NDL BIB ID: 029508361
• Web Site: http://id.ndl.go.jp/bib/029508361
• Web Site: https://ndlsearch.ndl.go.jp/books/R000000004-I029508361
• Web Site: https://www.jstage.jst.go.jp/article/jscejcm/74/2/74_I_107/_pdf
• Text Lang: ja
• Data Source:
  • JaLC
  • NDL
  • Crossref
  • CiNii Articles
  • KAKEN
• Abstract License Flag: Disallowed