Attachment | Size | Timestamp |
---|---|---|
phmc_14_079.pdf | 1023.83 KB | September 17, 2014 - 11:23pm |
Non destructive testing methods are often used in order to detect
and classify structural flaws. The detection of structural
flaws is useful for maintenance. In this paper we propose to
classify flaws in ferromagnetic materials by measuring Eddy
currents. Our approach consists of two steps. First, we use
a system identification algorithm to find a dynamical system
which describes the data. Then, we use the parameters of this
dynamical system as a feauture vector and we use support
vector machines in order to classify the various faults. We
test our method on a well-known benchmark data available at : http://wireless.feld.cvut.cz/diagnolab/node/16. This benchmark is particularly challenging, since the flaws might be more difficult to separate due to the variation of the scanning speed. The experimental evaluation shows that our approach gives good results.