Assessment of ground motion amplitude scaling using interpretable Gaussian process regression: Application to steel moment frames

Seismic consequences estimation for individual buildings is valuable for various stakeholders, including government entities, building owners, and insurers. The robustness of estimation results in the presence of incomplete input information can typically be investigated through sensitivity analysis. However, the estimation process's complexity and the sensitivity analysis's computational burden hinder its practical application, which requires a more efficient procedure to facilitate broader use. This paper proposes a novel framework for sensitivity analysis of seismic consequences estimation to improve the efficiency and reliability of such analysis. The proposed approach encompasses three key components: (1) stochastic ground motion modeling (SGMM)‐based seismic consequences estimation to evaluate the economic, environmental, and social consequences given specific buildings by considering different hazard levels, (2) the training of surrogate model (Gaussian process model) for structural analysis to reduce the computational cost of the evaluation process, and (3) variance‐based global sensitivity analysis to investigate the importance of parameters of concern in the estimation process. The entire procedure is implemented in Python, adhering to object‐oriented programming, and does not rely on external software. Then, the proposed methodology is applied to two distinct three‐story steel moment‐resistant frames (SMRFs) subjected to four different hazard levels to demonstrate its effectiveness. The SGMM method can generate specific ground motions for each hazard level, mitigating the potential for result bias from using ground motions with unrealistic characteristics. Furthermore, the SGMM method is particularly suitable for automated analysis processes reducing the laborious task of screening ground motions from the database. Comparative analysis with surrogate‐free estimation reveals that the surrogate‐based analysis delivers reliable results with significantly reduced computational cost. The results of analyzing different structures under varying hazard levels reflect the variability in sensitivity analysis of consequences estimation, highlighting the necessity of the proposed flexible and efficient framework. Furthermore, the proposed framework's advantages, limitations, and future research needs are discussed.

Section: Variance-based Global Sensitivity Analysis Methodsmentioning

confidence: 99%

Section: Further Improvement In the Surrogate Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Surrogate model‐aided global sensitivity analysis framework for seismic consequences estimation in buildings

Du,

Wang

2024

“…6,7 Different selection procedures have been proposed in the literature either with structural or geotechnical purposes. [8][9][10][11][12][13][14][15][16][17][18] However, depending on the adopted compatibility criteria and on the characteristics of the target motion, the selection of a suitable number of compatible actual accelerograms might result impossible without applying large scale factors which distort the characteristics of the selected motions. 14,19,20 This may lead to unrealistic inputs and hence to unreliable predictions of the seismic performance of structural and geotechnical systems 14,[19][20][21][22][23] ; for the latter case, Biondi et al 21 and Aliberti et al 23 have shown that the influence of the input motion selection criteria on the prediction reliability is relevant regardless the complexity level of the analyses.…”

Section: Introductionmentioning

confidence: 99%

“…Since the target ground motion is usually defined in terms of code‐prescribed elastic response spectra, the geotechnical properties of soil affecting the non‐linear site response should be accounted for in the definition of the interval of vibration periods adopted for checking the compatibility of the selected records 6,7 . Different selection procedures have been proposed in the literature either with structural or geotechnical purposes 8–18 . However, depending on the adopted compatibility criteria and on the characteristics of the target motion, the selection of a suitable number of compatible actual accelerograms might result impossible without applying large scale factors which distort the characteristics of the selected motions 14,19,20 .…”

Section: Introductionmentioning

confidence: 99%

Energy‐compatible modulating functions for the stochastic generation of fully non‐stationary artificial accelerograms and their effects on seismic site response analysis

Genovese

Biondi

Cascone

et al. 2023

The reliability of non‐linear dynamic analysis aimed to predict the seismic performance of structural and geotechnical systems, as well as their dynamic interaction, is significantly affected by the selection of suitable input motions. Several procedures for selecting actual earthquake records compatible with the seismic hazard at the site of interest and for evaluating synthetic accelerograms consistent with a seismological framework are available in the literature. However, the degree of uncertainty affecting these approaches might lead to unreliable performance predictions especially for strongly non‐linear problems, such as those involving the response of soils to cyclic and dynamic loading. In this vein, the paper presents a procedure for generating fully non‐stationary ground motions ensuring energy compatibility with a target motion. Two different approaches have been introduced: i) the time window method based on the central finite difference approach and ii) the simplified intensity‐compatible approach in which a closed form expression is provided to evaluate a modulating function that depends on the Arias intensity and on the strong motion duration of a target motion. To highlight the reliability and the accuracy of the proposed procedure, several sets of spectrum‐compatible artificial accelerograms, generated starting from a rock outcropping motion, were used as input motions in a series of one‐dimensional site response analyses. The analysis results are presented and discussed in the paper highlighting the influence of the main features of the proposed generation procedure on the variability of the computed site response.

Efficiency and explainability of design‐oriented machine learning models to estimate seismic response, fragility, and loss of a steel building inventory

Zaker Esteghamati,

Baddipalli

2024

Machine learning (ML) has recently been used as an efficient surrogate to estimate different steps of performance‐based earthquake engineering (PBEE), from dynamic structural analysis to fragility and loss assessments. However, due to the varied data, models, and features in existing literature, the relative efficiency of ML models across different PBEE steps remains unclear. Additionally, the black‐box nature of advanced ML algorithms limits their ability to provide design‐oriented insights, hindering the broader application of ML in PBEE‐based design. This study provides a comprehensive comparison of the accuracy and explainability of design‐oriented ML models across different steps of PBEE using a consistent database of 621 steel moment frames with varying designs and geometry. Eight ML algorithms were used in a careful training workflow comprising feature selection, hyperparameter tuning, cross‐validation, and model inference. The sensitivity of model accuracy to representative PBEE outputs—maximum responses, median fragility, and expected annual loss—was assessed using statistical measures. In addition, the explainability of the best models for each step was examined to explore the relationship between design parameters and the corresponding PBEE output. The results show that while ML models can reasonably map design parameters to all different PBEE outputs, models accuracy was higher for drift responses, median fragilities, and component‐based loss metrics. In addition, the optimal algorithm remained the same across different PBEE steps, where support vector machines and random forests provided the highest accuracy with an average R2 of 0.93 and 0.91 over different outputs on the test set. Although the selected feature sets varied across outputs and algorithms, height, number of stories, fundamental period, and the minimum of the beams’ moment of inertia were influential for both models and notably affected different PBEE outputs.