Affiliations: [a] Institute for Fundamentals and Theory in Electrical
Engineering, Graz University of Technology, Graz, Austria
Correspondence:
[*]
Corresponding author: Alice Reinbacher-Köstinger, Institute
for Fundamentals and Theory in Electrical Engineering, Graz University of
Technology, Kopernikusgasse 24, 8010 Graz, Austria. Tel.: +43 316 873 7255;
Fax: +43 316 873 7761; E-mail: alice.koestinger@tugraz.at
Abstract: The identification of a conductive object known in shape but hidden
by low- or non-conductive material is a special application of non-destructive
eddy current testing. To find the positioning and orientation parameters of an
object from the magnetic flux densities measured in a distance of several
centimeters, a non-linear inverse problem has to be solved. This can be done by
deterministic as well as stochastic optimization methods in principle. In order
to fulfill all demands on the identification process, that are high accuracy,
reliability and low computation time, a combination of a deterministic and a
stochastic method is useful and was therefore applied to the problem at hand.
It is shown that the computation time can be decreased enormously applying a
hybrid method while the required accuracy and reliability hold.
