Identification of rail-sleeper-ballast system through time-domain Markov chain Monte Carlo-based Bayesian approach

Lam, Heung-Fai; Alabi, Stephen Adeyemi; Yang, Jie

doi:10.1016/j.engstruct.2017.03.001

Cited by 34 publications

(18 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are several factors influencing the effectiveness and practicability of ballast monitoring methods within the industrial framework, thereby their adoption by railway infrastructure managers. The following performance metrics have been selected to carry out a comparison with other methods proposed in the literature, such as direct methods (DM) and highfidelity model-based methods (HFMM) [7], [9], [19]- [21], as well as with the current industrial practice based on the track geometry measurements (TGM) [56]:…”

Section: Comparative Analysis With Alternative Methodsmentioning

confidence: 99%

“…Employing model-based techniques is the third approach to railway track degradation analysis. A notable research effort has been conducted by Lam et al [19]- [21] for detection of ballast damage based on a combination of high complexity mechanical models and measured vibration data in model updating frameworks. The feasibility of quantifying ballast damage characteristics (location and level of stiffness reduction) in an artificially damaged system was studied using deterministic [19] and Bayesian [20] model updating techniques.…”

Section: A Literature Surveymentioning

confidence: 99%

“…The feasibility of quantifying ballast damage characteristics (location and level of stiffness reduction) in an artificially damaged system was studied using deterministic [19] and Bayesian [20] model updating techniques. A further investigation was carried out in [21] by employing the Markov chain Monte Carlo algorithm to ensure the applicability of the proposed methodology. The receptance function has been widely used in several research studies for the calibration of numerical model properties, such as ballast and rail-pad stiffness and damping coefficients [22]- [24].…”

Section: A Literature Surveymentioning

confidence: 99%

See 2 more Smart Citations

Monitoring of Turnout Ballast Degradation Using Statistical Low-Complexity Behavioral Models

Barkhordari

Galeazzi

Blanke

2021

IEEE Trans. Contr. Syst. Technol.

View full text Add to dashboard Cite

The dependability of the railway infrastructure is paramount to guarantee safety, comfort, and network capacity. Turnouts are the key infrastructure element to enable the maximization of network capacity and minimization of transport delays. Their failure upsets the overall performance of the railway network; hence, infrastructure managers are interested in securing as high as possible uptime. The ballast layer provides the elastic support to the superstructure (sleepers and rail); thereby, it is crucial to ensure both safety and comfort in railway transport. This article presents a novel detection system for the monitoring of ballast degradation throughout its service life. A statistical model based on the generalized extreme value distribution is proposed to describe the behavior of the resonance frequency associated with the ballast. The generalized likelihood ratio test is then adopted to detect when the state of health of the ballast changes over time. The monitoring system is tested exploiting full-scale measurements of train-induced track vertical acceleration collected at a turnout of the Danish railway network over a two-year period, which includes a maintenance event. Results demonstrate the ability of the ballast monitoring system in detecting the progressive degradation of the ballast quality.

show abstract

Section: Comparative Analysis With Alternative Methodsmentioning

confidence: 99%

Section: A Literature Surveymentioning

confidence: 99%

Section: A Literature Surveymentioning

confidence: 99%

See 1 more Smart Citation

Monitoring of Turnout Ballast Degradation Using Statistical Low-Complexity Behavioral Models

Barkhordari

Galeazzi

Blanke

2021

IEEE Trans. Contr. Syst. Technol.

View full text Add to dashboard Cite

show abstract

“…34 Beck and Au 35 used Markov Chain Monte Carlo (MCMC) method to generate samples to approximate the posterior distribution. Lam et al 36 introduced a novel stopping criterion with MCMC-based Bayesian model to ensure the accuracy of the calculated posterior distribution. Pepi et al 37 used a modified version of the MCMC method to estimate the posterior marginal probability density function of the selected model parameters.…”

Section: Introductionmentioning

confidence: 99%

Probabilistic forecast of wind speed based on Bayesian emulator using monitoring data

Ding

Wan

2020

Struct Control Health Monit

View full text Add to dashboard Cite

Wind speed forecasting can serve a wide spectrum of purposes, including scheduling of a power system and dynamic control of structures. A lot of models are widely used to forecast wind speed, consisting of deterministic models (e.g., physical models, statistical models, and artificial intelligence models) and probabilistic models (e.g., Bayesian model). The wind speed has the characteristics of random, nonlinear, and uncertainty, which highlights the importance of using Bayesian model to predict the wind speed. In this study, a Bayesian emulator with Gaussian process prior is adopted for probabilistic forecast of wind speed. The present Bayesian emulator approach not only maintains the data-driven property which guarantees its high flexibility in modeling the complexity of the target system but also allows for the efficient, probabilistic evaluation of the wind speed in terms of the predictive mean and variance. Nevertheless, the modeling performance of the Bayesian emulator directly depends on the selected covariance function. The influence of different types of covariance functions, which include squared-exponential (SE) covariance function, Matern (MA) covariance function, periodic (PE) covariance function, and composite covariance function, on forecasting performance of the wind speed is studied. One-month wind monitoring data collected by structural health monitoring (SHM) system installed on Jiubao bridge are employed to demonstrate the effectiveness of Bayesian emulator for forecasting the wind speed.

show abstract

“…Lam et al solved this problem by proposing a Markov chain Monte Carlo (MCMC)‐based Bayesian model updating method in which samples were drawn from the posterior PDF at multiple levels extended from Beck and Au . By constructing a bridge PDF at each sampling level, such that the PDFs finally converged to the target posterior PDF in a controlled manner, this method was successfully applied in a series of applications . To investigate the uncertainties in Bayesian model updating, the transitional MCMC algorithm is developed in Ching and Chen, and the main difference of transitional MCMC with the method presented in Lam et al is a different strategy of controlling the samples in each sampling level to approach the important regions of posterior PDF.…”

Section: Introductionmentioning

confidence: 99%

Operational modal identification and finite element model updating of a coupled building following Bayesian approach

Lam

Yang

2017

Struct Control Health Monit

Self Cite

View full text Add to dashboard Cite

Summary This paper presents a comprehensive study of the full‐scale ambient vibration test, modal analysis, finite element (FE) modeling, and model updating of a coupled building in Hong Kong. The coupled building comprised a main part and a complementary part. To capture the dynamic properties of the building, a 21‐setup ambient vibration test was designed and conducted. The modal parameters of each setup were identified following a fast Bayesian fast Fourier transform approach, and the partial mode shapes from the different setups were assembled following the least squares method. The identified modal parameters were analyzed and discussed in detail, revealing certain features of the coupling effects between the main and complementary parts. To determine the equivalent Young's moduli of various structural components, an FE model of the coupled building was developed and updated with the identified modal parameters. The Bayesian approach was followed to explicitly handle the uncertainties induced by modeling error and measurement noise. To ensure the model updating method is applicable even in unidentifiable cases, a Markov chain Monte Carlo simulation was employed in the proposed method to generate samples for approximating the posterior probability density functions of uncertain model parameters. The close match between the modal parameters calculated from the updated FE model and those identified from the measured time‐domain data verified the validity of the proposed FE model. This study provides valuable experience and information for the development of structural model updating and structural health monitoring of building systems.

show abstract

Identification of rail-sleeper-ballast system through time-domain Markov chain Monte Carlo-based Bayesian approach

Cited by 34 publications

References 29 publications

Monitoring of Turnout Ballast Degradation Using Statistical Low-Complexity Behavioral Models

Monitoring of Turnout Ballast Degradation Using Statistical Low-Complexity Behavioral Models

Probabilistic forecast of wind speed based on Bayesian emulator using monitoring data

Operational modal identification and finite element model updating of a coupled building following Bayesian approach

Contact Info

Product

Resources

About