A tandem trust-region optimization approach for ill-posed falling weight deflectometer backcalculation

Romeo, Ryan; Davis, R. Benjamin; Lee, Hyung S.; Durham, Stephan A.; Kim, S. Sonny

doi:10.1016/j.compstruc.2022.106935

Cited by 4 publications

(10 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the flat-prior case, starting points were generated using Latin-hypercube sampling to ensure an even starting distribution (Romeo et al, 2023). In the tuned-prior case, F I G U R E 5 Comparison of Bayesian and TTR backcalculation of dynamic modulus using a (a) flat prior and (b) tuned prior.…”

Section: Methodsmentioning

confidence: 99%

“…Tandem trust-region optimizer (TTR) Romeo et al presented the TTR method as a trustregion optimizer that combines the subspace trust-region interior reflexive (STIR) method and the Levenberg-Marquardt algorithm (LMA; Romeo et al, 2023). The effectiveness of the TTR method was demonstrated in thousands of FWD backcalculations across several pavement models.…”

Section: 11mentioning

confidence: 99%

“…Traditional approaches to the inverse FWD problem use gradient‐based optimization (Goktepe et al., 2006), and gradient methods remain a popular choice for modern backcalculation techniques both in research studies and in practice (Y. R. Kim et al., 2021; Lee et al., 2019; Romeo et al., 2023; Wang et al., 2022). In gradient‐based methods, an initial guess is made for the pavement's material properties, and an optimizer iteratively runs the forward model until it converges to a set of parameters minimizing the difference between measured and simulated FWD deflections.…”

Section: Introductionmentioning

confidence: 99%

“…Most modern forward models compute pavement responses in the frequency domain; however, FWD impacts typically excite a narrow range of loading frequencies known as the EFR. As a result, inverse calculation typically has high accuracy within the EFR (approximately 1 to 100 Hz) and poor accuracy otherwise (Romeo et al., 2023). Noise in FWD force and deflection measurements also affects backcalculation (Garbowski & Pożarycki, 2017; Gopalakrishnan & Manik, 2010; Parvini, 1997).…”

Section: Introductionmentioning

confidence: 99%

“…The inverse problem is ill‐conditioned, and many sets of pavement properties can yield nearly identical responses to FWD loading. Thus, the presence of noise makes it difficult to distinguish between true values and erroneous solutions that produce similar responses (Romeo et al., 2023). Traditional backcalculation approaches compute one of many solutions that match the measured response, with no way to quantify uncertainty.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Bayesian backcalculation of pavement properties using parallel transitional Markov chain Monte Carlo

Coletti,

Romeo,

Davis

2023

Computer aided Civil Eng

Self Cite

View full text Add to dashboard Cite

This paper presents a novel Bayesian method for backcalculation of pavement dynamic modulus, stiffness, thickness, and damping using falling weight deflectometer (FWD) data. The backcalculation procedure yields estimates and uncertainties for each pavement property of interest. As a by‐product of the Bayesian procedure, information about measurement error is recovered. The Bayesian method is tested on simulated FWD backcalculations and compared with a state‐of‐the‐art trust‐region optimization algorithm, and it achieves estimation errors that are nearly an order of magnitude lower than the trust‐region solver. Confidence intervals are computed from thousands of simulated backcalculations and are shown to quantify uncertainty in estimated pavement properties. To cope with the computational expense of backcalculation, a fully parallel transitional Markov chain Monte Carlo procedure is developed. The fully parallel algorithm scales well to computation with many processor cores, and it yields up to a 50% reduction in computation time when compared to existing parallel implementations.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: 11mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Bayesian backcalculation of pavement properties using parallel transitional Markov chain Monte Carlo

Coletti,

Romeo,

Davis

2023

Computer aided Civil Eng

Self Cite

View full text Add to dashboard Cite

show abstract

Optimal reaction pathways of carbon dioxide hydrogenation using P-graph attainable region technique (PART)

Tan,

Sitoh,

Foo

et al. 2023

Discov Chem Eng

View full text Add to dashboard Cite

The attainable region interpretation of the thermodynamic principles has indicated that carbon dioxide (CO2) can be either hydrogenated directly to form dimethyl ether (DME) or gasoline. The process that converts CO2 to DME is more thermodynamically favourable at lower temperature. A certain thermodynamic temperature range (25 to 300 °C) is suggested for the conversion of CO2 to DME via a methanol intermediate pathway without addition of work. Optimal synthesis routes derived from P-graph's mutual exclusion solver were compared with reactions reported in literature and showed great correlation. The reactions collectively possess Gibbs free energy of less than zero, and negative enthalpy of reaction. With P-graph attainable region technique, the case studies have demonstrated that the synthesis of DME and gasoline using CO2 hydrogenation via methanol intermediate and carbon monoxide intermediate from Fischer–Tropsch synthesis is feasible with no work and heat requirement. Both case studies have demonstrated visual advantage of P-graph and data-driven applications. The benefit of integrating the P-graph framework with machine learning model like decision tree classifier was also demonstrated in the second case study as it solves topological optimisation problems without scaling constraints.

show abstract

Asphalt-related temperature correction of deflections measured in central FWD tests on a concrete-over-asphalt composite pavement

Díaz Flores,

Donev,

Aminbaghai

et al. 2024

International Journal of Pavement Engineering

View full text Add to dashboard Cite

A tandem trust-region optimization approach for ill-posed falling weight deflectometer backcalculation

Cited by 4 publications

References 37 publications

Bayesian backcalculation of pavement properties using parallel transitional Markov chain Monte Carlo

Bayesian backcalculation of pavement properties using parallel transitional Markov chain Monte Carlo

Optimal reaction pathways of carbon dioxide hydrogenation using P-graph attainable region technique (PART)

Asphalt-related temperature correction of deflections measured in central FWD tests on a concrete-over-asphalt composite pavement

Contact Info

Product

Resources

About