ANN inverse analysis based on stochastic small-sample training set simulation

Novák, Drahomı́r; Lehký, David

doi:10.1016/j.engappai.2006.05.003

Cited by 88 publications

(52 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Since in many engineering problems, measurements are represented by single or several curves usually discretized into discrete points, a number of observed quantities is often very large. Therefore, a global sensitivity analysis is applied to choose only quantities important for a particular model parameter to be identified [24]. A number of neurons in hidden layer is determined by consecutively increasing their number taking the over-training and under-training issues into account.…”

Section: Neural Network Architecturementioning

confidence: 99%

“…Log-sigmoid functions are considered as activation functions in all neurons. In all cases studied in this work, 60 training samples are generated by Latin Hypercube Sampling method and optimized by Simulated Annealing in order to minimize the correlation among all samples [24]. Those two methods are implemented in FREET software [25] that has been used.…”

Section: Experimental Validationmentioning

confidence: 99%

“…The procedure itself is based on work by Novák and Lehký [24] and has been successfully verified against artificially generated sets of data in [18]. Here, we present an independent validation and comparison with expert user guesses.…”

Section: Introductionmentioning

confidence: 96%

See 2 more Smart Citations

Soft computing-based calibration of microplane M4 model parameters: Methodology and validation

Kučerová

Lepš

2014

Advances in Engineering Software

View full text Add to dashboard Cite

Dedicated to Professor Zdeněk Bittnar in occasion of his 70th birthday.Keywords: computational simulation of concrete, microplane model M4, inverse analysis, neural networks, global sensitivity analysis, evolutionary algorithm. AbstractConstitutive models for concrete based on the microplane concept have repeatedly proven their ability to well-reproduce non-linear response of concrete on material as well as structural scales. The major obstacle to a routine application of this class of models is, however, the calibration of microplane-related constants from macroscopic data. The goal of this paper is two-fold: (i) to introduce the basic ingredients of a robust inverse procedure for the determination of dominant parameters of the M4 model proposed by Bažant and co-workers in [4] based on cascade Artificial Neural Networks trained by Evolutionary Algorithm and (ii) to validate the proposed methodology against a representative set of experimental data. The obtained results demonstrate that the soft computing-based method is capable of delivering the searched response with an accuracy comparable to the values obtained by expert users.

show abstract

Section: Neural Network Architecturementioning

confidence: 99%

Section: Experimental Validationmentioning

confidence: 99%

See 1 more Smart Citation

Soft computing-based calibration of microplane M4 model parameters: Methodology and validation

Kučerová

Lepš

2014

Advances in Engineering Software

View full text Add to dashboard Cite

show abstract

“…Note that in our identification methodology we do not suppose to have an expert capable of giving the initial estimate of material parameters values, as in e.g. [10,23,20]. Therefore, the bounds on model parameters were kept rather wide.…”

Section: A Brief Description Of the Identified Modelmentioning

confidence: 99%

Novel anisotropic continuum‐discrete damage model capable of representing localized failure of massive structures

Brancherie

Ibrahimbegović²

2009

Engineering Computations

View full text Add to dashboard Cite

In Part I of this paper we have presented a simple model capable of describing the localized failure of a massive structure. In this part, we discuss the identification of the model parameters from two kinds of experiments: a uniaxial tensile test and a three-point bending test. The former is used only for illustration of material parameter response dependence, and we focus mostly upon the latter, discussing the inverse optimization problem for which the specimen is subjected to a heterogeneous stress field.

show abstract

“…The first one utilizes ANN too, but in a different way: Computational time is reduced by using a small-sample simulation technique called Latin hypercube sampling (LHS) in ANN based inverse problem proposed by Novák and Lehký in [10] and [11] first.…”

Section: Introductionmentioning

confidence: 99%

Inverse Reliability Task: Artificial Neural Networks and Reliability-Based Optimization Approaches

Lehký

Slowik

Novák

2014

Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications

Self Cite

View full text Add to dashboard Cite

ANN inverse analysis based on stochastic small-sample training set simulation

Cited by 88 publications

References 19 publications

Soft computing-based calibration of microplane M4 model parameters: Methodology and validation

Soft computing-based calibration of microplane M4 model parameters: Methodology and validation

Novel anisotropic continuum‐discrete damage model capable of representing localized failure of massive structures

Inverse Reliability Task: Artificial Neural Networks and Reliability-Based Optimization Approaches

Contact Info

Product

Resources

About