Uncertainty quantification considering global sensitivity for seismic risk analysis of existing bridge structural systems

2022 Fiscal Year Final Research Report

• Project/Area Number: 20H02229
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 22020:Structure engineering and earthquake engineering-related
• Research Institution: University of Tsukuba
• Principal Investigator: Nishio Mayuko 筑波大学, システム情報系, 准教授 (00586795)
• Project Period (FY): 2020-04-01 – 2023-03-31
• Keywords: 不確定性定量化 / 事後分布 / 非線形構造振動 / 構造信頼性 / 画像計測 / 地震フラジリティ

Outline of Final Research Achievements

Seismic risk assessment considering deterioration and damage is required for appropriate maintenance and planning for disaster prevention and mitigation of existing bridge systems that form transportation networks. This study presented the data assimilation that derives posterior distributions avoiding the ill-posed condition by extracting parameters with high "sensitivity to uncertainty" that depends on damage conditions and characteristics of input earthquake motions. Its effectiveness was shown by the experimental study that is target to the data assimilation of load capacity analysis of a steel member specimen with locally introduced corrosion using vision-based strain sensing. In addition, the surrogate modeling method for structural reliability analysis dealing with high-dimensional uncertainties of structural properties and seismic motions was constructed. The applicability was shown by performing the seismic system fragility analysis of an elevated bridge structure.

Free Research Field

構造工学

Academic Significance and Societal Importance of the Research Achievements

データ同化において不確定性感度を考慮することで適切なモデルパラメータ事後分布が推定できることを示したこと，そして地震フラジリティ解析の計算コストを大幅に低減する機械学習を活用した代替モデル法を構築した成果は学術的意義が大きい．また，インフラ構造物運用現場において，運用者が最も知りたい保有性能を得るためのデータ活用法と計算手法を示した社会的意義も大きい．