Neural network-based inversion algorithms in magnetic flux leakage nondestructive evaluation

Pradeep Ramuhalli; Lalita Udpa; Satish S. Udpa

doi:10.1063/1.1558693

Neural network-based inversion algorithms in magnetic flux leakage nondestructive evaluation

Pradeep Ramuhalli, Lalita Udpa, Satish S. Udpa

Research output: Contribution to journal › Article › peer-review

89 Scopus citations

Abstract

A neural network forward model based iterative approach for solving inverse problem in electromagnetic nondestructive testing was presented. The proposed model uses two neural networks in feedback configuration-a forward network and an inverse network. It was found that the proposed approach is capable of providing accurate estimates of the flaw profiles, even in the presence of reasonable noise.

Original language	English
Pages (from-to)	8274-8276
Number of pages	3
Journal	Journal of Applied Physics
Volume	93
Issue number	10 3
DOIs	https://doi.org/10.1063/1.1558693
State	Published - May 15 2003
Externally published	Yes

Access to Document

10.1063/1.1558693

Cite this

@article{40a2fb0a358a4f0486ce8fe09dbaa923,

title = "Neural network-based inversion algorithms in magnetic flux leakage nondestructive evaluation",

abstract = "A neural network forward model based iterative approach for solving inverse problem in electromagnetic nondestructive testing was presented. The proposed model uses two neural networks in feedback configuration-a forward network and an inverse network. It was found that the proposed approach is capable of providing accurate estimates of the flaw profiles, even in the presence of reasonable noise.",

author = "Pradeep Ramuhalli and Lalita Udpa and Udpa, {Satish S.}",

year = "2003",

month = may,

day = "15",

doi = "10.1063/1.1558693",

language = "English",

volume = "93",

pages = "8274--8276",

journal = "Journal of Applied Physics",

issn = "0021-8979",

publisher = "American Institute of Physics",

number = "10 3",

}

TY - JOUR

T1 - Neural network-based inversion algorithms in magnetic flux leakage nondestructive evaluation

AU - Ramuhalli, Pradeep

AU - Udpa, Lalita

AU - Udpa, Satish S.

PY - 2003/5/15

Y1 - 2003/5/15

N2 - A neural network forward model based iterative approach for solving inverse problem in electromagnetic nondestructive testing was presented. The proposed model uses two neural networks in feedback configuration-a forward network and an inverse network. It was found that the proposed approach is capable of providing accurate estimates of the flaw profiles, even in the presence of reasonable noise.

AB - A neural network forward model based iterative approach for solving inverse problem in electromagnetic nondestructive testing was presented. The proposed model uses two neural networks in feedback configuration-a forward network and an inverse network. It was found that the proposed approach is capable of providing accurate estimates of the flaw profiles, even in the presence of reasonable noise.

UR - http://www.scopus.com/inward/record.url?scp=0037641215&partnerID=8YFLogxK

U2 - 10.1063/1.1558693

DO - 10.1063/1.1558693

M3 - Article

AN - SCOPUS:0037641215

SN - 0021-8979

VL - 93

SP - 8274

EP - 8276

JO - Journal of Applied Physics

JF - Journal of Applied Physics

IS - 10 3

ER -

Neural network-based inversion algorithms in magnetic flux leakage nondestructive evaluation

Abstract

Access to Document

Other files and links

Fingerprint

Cite this