Title :
ECT Characterization of the Extent of Minute Cracks using a Database Based Inversion Procedure
Author :
Rapacchi, S. ; Le Bihan, Y. ; Pávó, J. ; Marchand, C.
Author_Institution :
Lab. de Genie Electrique de Paris, CNRS-SUPELEC-UPS-UPMC, Gif-sur-Yvette
Abstract :
A study on the characterization of minute cracks by using ECT probe signal is presented. The significant parameters of the ECT data are selected using PCA. The inversion is then achieved using a neural network trained with numerical data. Tests realized with experimental data show that the area of the cracks can be estimated with a fair level of accuracy
Keywords :
cracks; eddy current testing; electrical engineering computing; neural nets; principal component analysis; ECT; PCA; database based inversion procedure; minute cracks; neural network; probe signal; Aerospace industry; Databases; Electrical capacitance tomography; Frequency; Impedance; Neural networks; Principal component analysis; Probes; Shape; Uninterruptible power systems;
Conference_Titel :
Electromagnetic Field Computation, 2006 12th Biennial IEEE Conference on
Conference_Location :
Miami, FL
Print_ISBN :
1-4244-0320-0
DOI :
10.1109/CEFC-06.2006.1633040
