A machine learning framework for assessing post-earthquake structural safety. (May 2018)

• Record Type: Journal Article
• Authors: Zhang, Yu; Burton, Henry V.; Sun, Han; Shokrabadi, Mehrdad
• Is Part Of: Structural safety. Volume 72(2018)
• Page Start: 1
• Page End: 16
• Publication Date: 2018-05
• Subjects: CART Classification and regression tree -- CP Complexity parameter -- DM Damage measure -- EDP Engineering demand parameter -- IDA Incremental dynamic analysis -- NRHA Nonlinear response history analysis -- PBEE Performance-based earthquake engineering -- RC Reinforced concrete -- ROC Receiver operating characteristic -- SMF Special moment frame; Machine learning -- Random forests -- Damage pattern -- Post-earthquake structural safety -- Performance-based assessment -- Seismic resilience; Structural stability -- Periodicals; Safety factor in engineering -- Periodicals; Reliability (Engineering) -- Periodicals; Constructions -- Stabilité -- Périodiques; Coefficient de sécurité en ingénierie -- Périodiques; Fiabilité -- Périodiques; 620.86
• DOI: 10.1016/j.strusafe.2017.12.001
• Languages: English
• ISSNs: 0167-4730

Abstract

Highlights: A framework for assessing the post-earthquake structural safety of damaged buildings is presented. The concepts of response and damage patterns are introduced and incorporated into a systematic methodology integrating probabilistic seismic demand analysis, component-level damage simulation and robust assessments of the residual collapse capacity. Machine learning algorithms are used to explicitly link the response and damage patterns to residual collapse capacity of a damaged structure, and are able to probabilistically predict the structural safety states given any available information. A series of predictive models including Classification and Regression Trees and Random Forests are developed and examined in detail to achieve the optimal model which balance multiple performance measurements. In contrast to previously judgement-based methods for the tagging process, this innovative approach provides a solid statistical support for structural safety assessment. High prediction accuracies are observed based on either response and damage patterns. Abstract: A machine learning framework is presented to assess post-earthquake structural safety. The concepts of response and damage patterns are introduced and incorporated into a systematic methodology for generating a robust dataset for any damaged building. Incremental dynamic analysis using sequential ground motions is used to evaluate the residual collapse capacity of the damaged structure. Machine learning algorithms are used to map response and damage patterns to the structural safety state (safe or unsafe to occupy) of the building based on an acceptable threshold of residual collapse capacity. Predictive models including classification and regression tree and Random Forests are used to probabilistically identify the structural safety state of an earthquake-damaged building. The proposed framework is applied to a 4-story reinforced concrete special moment frame building. Distinct yet partially overlapping response and damage patterns are found for the damaged building classified as safe and unsafe. High prediction accuracies of 91% and 88% are achieved when the safety state is assessed using response and damage patterns respectively. The proposed framework could be used to rapidly evaluate whether a damaged building remains structurally safe to occupy after a seismic event and can be implemented as a subroutine in community resilience evaluation or building lifecycle performance assessment and optimization.