2009
DOI: 10.1016/j.compstruc.2008.12.001
|View full text |Cite
|
Sign up to set email alerts
|

Computationally efficient seismic fragility analysis of geostructures

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
30
0
1

Year Published

2013
2013
2023
2023

Publication Types

Select...
6
2

Relationship

0
8

Authors

Journals

citations
Cited by 88 publications
(31 citation statements)
references
References 12 publications
0
30
0
1
Order By: Relevance
“…Metamodels are constructed to calibrate the relation between DMs and uncertain inputs of the structural models, including IMs and material parameters. The construction of the metamodels is either achieved by decomposing the nonlinear input-output relation with high-dimensional model representation (HDMR) [13,14], or realized with polynomial regression [15,16,17,18,19] or other more advanced statistical tools, such as artificial neural networks (ANNs) [20,21,22,23,24], LASSO regression [25], Bayesian networks [26], merging multivariate adaptive regression splines, radial basis function network, support vector regression [27], Kriging [9,28], etc. On the other hand, earthquake accelerations are also used directly as inputs of the metamodel in [29] to predict structural response time histories.…”
Section: Introductionmentioning
confidence: 99%
“…Metamodels are constructed to calibrate the relation between DMs and uncertain inputs of the structural models, including IMs and material parameters. The construction of the metamodels is either achieved by decomposing the nonlinear input-output relation with high-dimensional model representation (HDMR) [13,14], or realized with polynomial regression [15,16,17,18,19] or other more advanced statistical tools, such as artificial neural networks (ANNs) [20,21,22,23,24], LASSO regression [25], Bayesian networks [26], merging multivariate adaptive regression splines, radial basis function network, support vector regression [27], Kriging [9,28], etc. On the other hand, earthquake accelerations are also used directly as inputs of the metamodel in [29] to predict structural response time histories.…”
Section: Introductionmentioning
confidence: 99%
“…The ANN that has been used in this study consisted of three layers: (a) the input layer with sixteen nodes (16 structural modeling parameters in Table 1); (b) the hidden layer; and (c) the output layer with two nodes (one for the median value and the other for the standard deviation). It was observed by Lagaros et al (2009) that increasing the number of hidden layers did not alter significantly the performance of the ANN, thus the runs were performed using one intermediate layer. When applying ANN for the simulation of the structural reliability, especially under dynamic loading conditions and including material nonlinearity, it is difficult to find the best configuration of ANN architecture.…”
Section: The Prediction Capabilities Of Annmentioning
confidence: 99%
“…Conventionally, fragility assessment under seismic loading using a MCS-based method is a computationally intensive problem due to the reason that the MCS-based method needs a large number of samples to get an accuracy result (Lagaros et al 2009). Thus, in this study, the ANN models were used for the prediction of fragility curves to reduce the computational cost.…”
Section: Fragility Curves Via Ann Modelsmentioning
confidence: 99%
See 1 more Smart Citation
“…PGA was used as the intensity measure and damage states were based on the computed deformation of the slope crest. A similar approach was employed by Lagaros et al (2009) and Tsompanakis et al (2010), who also used PGA as an intensity measure. However unlike most examples of fragility curves, which typically use a physical damage descriptor to define damage states, the authors used target factors of safety from Greek seismic design codes as damage states.…”
Section: Introductionmentioning
confidence: 99%