Ya-Heng Yang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Ya-Heng Yang has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Tracy C. Becker
Assistant Professor, Dept. of Civil Engineering, McMaster Univ., Hamilton, ON, Canada L8S 4L8
2 shared publications - Bozidar Stojadinovic
ETH Zurich
1 shared publication Debra Murphy
Slate Geotechnical Consultants Inc. Oakland California USA
1 shared publicationHarumi Yoneda
1 shared publication- Jinyan Zhao1 shared publication
Maria Camila Lopez Ruiz
Department of Civil Environmental Engineering University of California Berkeley California USA
1 shared publication
Latest Publications
- Open Access
- 19 March 2026
- Published by Wiley in Earthquake Engineering & Structural Dynamics
- Vol. 55 (9) , 1828-1844
Quantifying Model Selection Uncertainty in Structural Analysis: Methodology and Application
With increasing focus on complex engineering systems under rare events, computational models are critical for predictions due to the scarcity or absence of data. However, selecting an appropriate model can be challenging. Using a single model without available test calibration could result in significant bias in performance predictions. A case study of a steel moment frame reveals substantial discrepancies between predictions given different model selections, particularly under increasing ground motion intensities. To address this challenge, this paper proposes a model averaging methodology based on the logic tree approach to incorporate multiple models and capture their full spectrum of potential outcomes. A combination of dimensionality reduction techniques and Bayesian inference is employed to determine the weight of each model candidate. The weight is represented as a Dirichlet prior, which is determined by incorporating expert knowledge and structural component testing results. This approach enables the assignment of a more informative prior even when no observational data is available, while also leaving room for updating the prior if data becomes available. By leveraging model averaging and uncertainty propagation techniques to address model uncertainty in structural analysis, a response interval is then generated for the quantity of interest. This methodology provides a more comprehensive, informative, and feasible assessment to address model selection uncertainty to improve predictive capabilities and enhance the ability of engineers to treat design in a probabilistic manner. In its present formulation, however, the approach is applicable to precollapse response regimes where finite response quantities can be defined and does not explicitly address collapse‐state realizations.
- 19 March 2026
- Published by Wiley in Earthquake Engineering & Structural Dynamics
- Vol. 55 (9) , 1828-1844
Open AccessQuantifying Model Selection Uncertainty in Structural Analysis: Methodology and Application
With increasing focus on complex engineering systems under rare events, computational models are critical for predictions due to the scarcity or absence of data. However, selecting an appropriate model can be challenging. Using a single model without available test calibration could result in significant bias in performance predictions. A case study of a steel moment frame reveals substantial discrepancies between predictions given different model selections, particularly under increasing ground motion intensities. To address this challenge, this paper proposes a model averaging methodology based on the logic tree approach to incorporate multiple models and capture their full spectrum of potential outcomes. A combination of dimensionality reduction techniques and Bayesian inference is employed to determine the weight of each model candidate. The weight is represented as a Dirichlet prior, which is determined by incorporating expert knowledge and structural component testing results. This approach enables the assignment of a more informative prior even when no observational data is available, while also leaving room for updating the prior if data becomes available. By leveraging model averaging and uncertainty propagation techniques to address model uncertainty in structural analysis, a response interval is then generated for the quantity of interest. This methodology provides a more comprehensive, informative, and feasible assessment to address model selection uncertainty to improve predictive capabilities and enhance the ability of engineers to treat design in a probabilistic manner. In its present formulation, however, the approach is applicable to precollapse response regimes where finite response quantities can be defined and does not explicitly address collapse‐state realizations.
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
Identifying Early-Phase Recovery Bottlenecks Through Outcome-Based Metrics in an Integrated Regional Resilience Assessment Platform
Disaster recovery is often governed not by physical damage alone, but by the ability of communities to mobilize and allocate limited recovery resources across space and time. While recent recovery simulation approaches have demonstrated how resource and service constraints shape aggregate recovery trajectories, such “coarse-grained” metrics provide limited guidance for decision-making during the immediate aftermath of a disaster. In particular, they offer little insight into where and why recovery processes stall across localities during the critical early days following an event, when intervention priorities must be set under uncertainty.Given a community’s damage state after a hypothetical earthquake scenario, we examine its recovery process, such as clean-up, inspection, and repair, under a suite of recovery resource allocation scenarios. The analysis uses SimCenter’s R2D tool together with embedded infrastructure system operation simulators and the recovery simulator pyrecodes. For each recovery resource allocation scenario, two complementary indicators, namely resource occupancy and first-passage unfinished metric, are introduced. Resource occupancy is defined as the fraction of components within each locality that are actively engaged in a given recovery stage at a specific time. First-passage unfinished metrics are defined as the fraction of components within each locality that are awaiting initiation of a recovery stage by a given day. When evaluated at early time horizons, these indicators reveal spatially heterogeneous recovery bottlenecks that are not apparent from system-level recovery curves.Recognizing that the availability of recovery resources is difficult to specify before a disaster and may change as additional resources are mobilized in the immediate aftermath of an event, a sensitivity analysis may be necessary for planning optimal recovery resource allocation strategies. To facilitate sensitivity analysis within a reasonable computing time, we combine recovery simulations with a surrogate modelling approach to enable rapid recovery simulations. In particular, polynomial chaos expansion is used as a computationally efficient surrogate to relate alternative resource allocation levels to locality-level recovery indicators. This enables efficient exploration of potential allocation scenarios after a disaster occurs, without requiring repeated high-fidelity recovery simulations, and supports time-constrained decision-making in the early response phase.By emphasizing early-phase (e.g., debris cleaning), locality-resolved diagnostics of resource bottlenecks, rather than aggregate recovery timelines, this study advances the assessment of systemic resilience in cascading risk contexts. The results demonstrate how existing recovery simulation tools, augmented with efficient surrogate modelling, can provide actionable insights for emergency managers and planners to prioritize interventions and allocate limited recovery resources across interconnected urban systems.
- 14 March 2026
- Published by Copernicus GmbH
Open AccessIdentifying Early-Phase Recovery Bottlenecks Through Outcome-Based Metrics in an Integrated Regional Resilience Assessment Platform
Disaster recovery is often governed not by physical damage alone, but by the ability of communities to mobilize and allocate limited recovery resources across space and time. While recent recovery simulation approaches have demonstrated how resource and service constraints shape aggregate recovery trajectories, such “coarse-grained” metrics provide limited guidance for decision-making during the immediate aftermath of a disaster. In particular, they offer little insight into where and why recovery processes stall across localities during the critical early days following an event, when intervention priorities must be set under uncertainty.Given a community’s damage state after a hypothetical earthquake scenario, we examine its recovery process, such as clean-up, inspection, and repair, under a suite of recovery resource allocation scenarios. The analysis uses SimCenter’s R2D tool together with embedded infrastructure system operation simulators and the recovery simulator pyrecodes. For each recovery resource allocation scenario, two complementary indicators, namely resource occupancy and first-passage unfinished metric, are introduced. Resource occupancy is defined as the fraction of components within each locality that are actively engaged in a given recovery stage at a specific time. First-passage unfinished metrics are defined as the fraction of components within each locality that are awaiting initiation of a recovery stage by a given day. When evaluated at early time horizons, these indicators reveal spatially heterogeneous recovery bottlenecks that are not apparent from system-level recovery curves.Recognizing that the availability of recovery resources is difficult to specify before a disaster and may change as additional resources are mobilized in the immediate aftermath of an event, a sensitivity analysis may be necessary for planning optimal recovery resource allocation strategies. To facilitate sensitivity analysis within a reasonable computing time, we combine recovery simulations with a surrogate modelling approach to enable rapid recovery simulations. In particular, polynomial chaos expansion is used as a computationally efficient surrogate to relate alternative resource allocation levels to locality-level recovery indicators. This enables efficient exploration of potential allocation scenarios after a disaster occurs, without requiring repeated high-fidelity recovery simulations, and supports time-constrained decision-making in the early response phase.By emphasizing early-phase (e.g., debris cleaning), locality-resolved diagnostics of resource bottlenecks, rather than aggregate recovery timelines, this study advances the assessment of systemic resilience in cascading risk contexts. The results demonstrate how existing recovery simulation tools, augmented with efficient surrogate modelling, can provide actionable insights for emergency managers and planners to prioritize interventions and allocate limited recovery resources across interconnected urban systems.
- 12 February 2026
- Published by Wiley in Earthquake Spectra
- Vol. 42 (1)
Scaling, Matching, and Variability: Effects of Time Series Selection in Performance of Code‐Compliant Bridges
The modification of time series for the analysis of structures remains a topic of continuous debate. While amplitude scaling is widely accepted, codes often penalize spectral matching. Recent advancements in spectral matching address prior concerns about tight spectral matching by preserving component‐to‐component variability. This study explores and compares the improved spectral matching method with amplitude scaling in estimating the means and distributions of bridge response parameters. Two highway bridges in California were analyzed using suites of scaled and spectrally matched motions. Site and structure specific benchmarks were established by developing hazard curves for engineering demand parameters using the conditional scenario spectra, providing a more objective basis for comparison. Results demonstrate that spectral matching consistently yields lower bias and variability in engineering demand parameter predictions between time series suites compared to amplitude scaling for both bridge models and sites, indicating that spectral matching offers more reproducible and reliable EDP estimates across users and highlighting its strengths for use in bridge design.
- 12 February 2026
- Published by Wiley in Earthquake Spectra
- Vol. 42 (1)
Scaling, Matching, and Variability: Effects of Time Series Selection in Performance of Code‐Compliant Bridges
The modification of time series for the analysis of structures remains a topic of continuous debate. While amplitude scaling is widely accepted, codes often penalize spectral matching. Recent advancements in spectral matching address prior concerns about tight spectral matching by preserving component‐to‐component variability. This study explores and compares the improved spectral matching method with amplitude scaling in estimating the means and distributions of bridge response parameters. Two highway bridges in California were analyzed using suites of scaled and spectrally matched motions. Site and structure specific benchmarks were established by developing hazard curves for engineering demand parameters using the conditional scenario spectra, providing a more objective basis for comparison. Results demonstrate that spectral matching consistently yields lower bias and variability in engineering demand parameter predictions between time series suites compared to amplitude scaling for both bridge models and sites, indicating that spectral matching offers more reproducible and reliable EDP estimates across users and highlighting its strengths for use in bridge design.
- 2 June 2023
- Published by Elsevier in Engineering Structures
- Vol. 290
Lumped mass models for use in predicting collapse of an isolated building
- 2 June 2023
- Published by Elsevier in Engineering Structures
- Vol. 290
Lumped mass models for use in predicting collapse of an isolated building