Bayesian network modeling of correlated random variables drawn from a Gaussian random field

Bensi, Michelle; Kiureghian, Armen Der; Straub, Daniel

doi:10.1016/j.strusafe.2011.05.001

Cited by 48 publications

(22 citation statements)

References 21 publications

(35 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For general statistical dependence among components, most DBN models of systems rapidly become computationally intractable when the number of components in the system or the number of random variables increases. Strategies for reducing the computational efforts when representing random fields in the BN have been proposed [2], but their applicability is still limited. In the proposed approach, the dependence structure is modeled by hierarchical models.…”

Section: Bayesian Network Model Of System Deteriorationmentioning

confidence: 99%

Reliability analysis and updating of deteriorating systems with dynamic Bayesian networks

2016

Self Cite

View full text Add to dashboard Cite

Section: Bayesian Network Model Of System Deteriorationmentioning

confidence: 99%

Reliability analysis and updating of deteriorating systems with dynamic Bayesian networks

2016

Self Cite

View full text Add to dashboard Cite

“…Statistical dependence among random variables can be represented in this way if their correlation matrix is of the Dunnet-Sobel class [23]. The model may be extended to describe more complex correlation structures [24].…”

Section: Spatial Model Of Chloride-induced Reinforcement Corrosionmentioning

confidence: 99%

Assessing and updating the reliability of concrete bridges subjected to spatial deterioration – principles and software implementation

Schneider

Fischer

Bügler

et al. 2015

Structural Concrete

Self Cite

View full text Add to dashboard Cite

Inspection and maintenance of concrete bridges is a major cost factor in transportation infrastructure IntroductionDeterioration processes can reduce the reliability and the safety of engineering structures. Knowledge of the actual condition of the structure is therefore essential in order to comply with element and system target reliabilities (see [1][2][3]). Inspections and structural health monitoring are effective means of acquiring information on the actual condition of an ageing structure and should be utilized to reduce the uncertainty in engineering models and, ultimately, to optimize the management of deteriorating structures. This task is ideally carried out through a Bayesian analysis approach [4][5][6][7][8].As most deterioration processes are spatially distributed in structures and the deterioration progress at different locations in a structure is correlated, such an analysis should be performed considering the structure as a whole (e.g. [9][10][11][12]). However, the integral assessment of spatially distributed deterioration is computationally demanding, and novel computational strategies are required for probabilistic assessment and Bayesian updating of spatial deterioration [13,14]. Ultimately, the modelling approach and the computational methods should lead to software that can be used by engineers who are not experts in reliability analyses. Only in this way can Bayesian methods enhance the integrity management of ageing structures in practice.This paper describes an approach in which the uncertain system deterioration state of an ageing concrete structure is described by a Dynamic Bayesian Network (DBN) model of the underlying stochastic deterioration processes. This approach was motivated by the work of Straub [15] and Qin and Faber [16]. DBN models provide a computational framework that enables the evaluation of the joint distribution function of the system deterioration state based on the prior stochastic model and, in particular, Bayesian updating of the joint distribution function of the system deterioration state with observations of the underlying deterioration process, i.e. inspection and monitoring results.To prove the concept, the approach is implemented in a software prototype employing a DBN model of chloride-induced reinforcement corrosion, which can approximate the spatial correlation of the corrosion process. The prototype is applied to assess and update the deterioration state and structural reliability of a single-cell prestressed concrete box girder. This construction type is typical of long, multi-span highway bridges in Germany. However, the general solution strategy applies to all common concrete bridge structures.Section 2 gives an overview of the underlying system model. The deterioration model and the structural capacity model are described in more detail in sections 3 and 4. The implementation of the prototype is briefly described in section 5. A case study applying the prototype is summarized in section 6. System modelTo model the deterioration state, the box girder i...

show abstract

“…The reader interested in a more thorough introduction to BN is referred to Jensen and Nielsen (2007). Applications of BNs in engineering are presented for example in Straub and Der Kiureghian (2010), Bensi et al (2011), Faber et al (2012 or Codetta-Raiteri et al (2012).…”

Section: /50mentioning

confidence: 99%

Probabilistic assessment of tunnel construction performance based on data

Špačková

Šejnoha

Straub

2013

Tunnelling and Underground Space Technology

Self Cite

View full text Add to dashboard Cite

A probabilistic model for estimating tunnel excavation time is learnt with data from past tunnel projects. The model is based on the Dynamic Bayesian Network technique. The model inputs are determined through an analysis of data from three tunnels built by means of the conventional tunneling method. The data motivate the development of a novel probability distribution to describe the excavation performance. In addition, the probability of construction failure events and the delay caused by such failures are estimated using databases available in the literature. The model is applied to a case study, in which it is demonstrated how observations from the tunnel construction process can be included to continuously update the prediction of excavation time.

show abstract

Bayesian network modeling of correlated random variables drawn from a Gaussian random field

Cited by 48 publications

References 21 publications

Reliability analysis and updating of deteriorating systems with dynamic Bayesian networks

Reliability analysis and updating of deteriorating systems with dynamic Bayesian networks

Assessing and updating the reliability of concrete bridges subjected to spatial deterioration – principles and software implementation

Probabilistic assessment of tunnel construction performance based on data

Contact Info

Product

Resources

About