A machine learning approach for classification tasks of ECT signals in steam generator tubes nearby support plate

Roberto Miorelli (1) , Anastassios Skarlatos (1) , Christophe Reboud (1)

1 LSME (CEA, LIST) - Laboratoire de Simulation et de Modélisation Électromagnétique (CEA, LIST)

An automatic classification relying on model based machine learning approach is proposed in the context ofsteam generator tubes inspection by means of eddy current testing. Such tool could realize a first selection of problematicareas and thus potentially considerably reduce the amount of data experts need to analyse. After the generation ofdatabases of signals covering the configurations of interest, a set of classifiers are trained and compared in terms ofperformance. In order to mitigate the size of datasets and enhance classification performance, a classical dimensionalityreduction technique. Results indicate a good potential of such methods for assisting human experts in the task of ECTsignals analysis.

Mots clés: machine learning; eddy current testing; classification; support plat; steam generator tubes

Domaines: Sciences de l'ingénieur [physics]

Roberto Miorelli, Anastassios Skarlatos, Christophe Reboud. A machine learning approach for classification tasks of ECT signals in steam generator tubes nearby support plate. AIP Conference Proceedings, 2019, 38, pp.090004. ⟨10.1063/1.5099822⟩. ⟨cea-04556741⟩