Nonparametric density estimates used for multiple AE source detection

0498354 - ÚT 2019 RIV CZ eng C - Konferenční příspěvek (zahraniční konf.)
Gális, P. - Chlada, Milan
Nonparametric density estimates used for multiple AE source detection.
Stochastic and Physical Monitoring Systems. Praha: Czech Technical University in Prague, 2018 - (Franc, J.; Hobza, T.), s. 11-23. ISBN 978-80-01-06501-3.
[SPMS 2018. Dobřichovice (CZ), 18.06.2018-22.06.2018]
Grant CEP: GA MŠMT LM2015068; GA ČR GA16-09848S
Klíčová slova: material health monitoring * micro-cracks * acoustic emission * density estimation * kernel * numerical optimization
Obor OECD: Applied mechanics

Paper deals with the localization of acoustic emission (AE) sources by means of exact geodesic curves on 3D vessels composed of several parametrized surfaces. Precise arrival times of AE events, plugged in DeltaT/DeltaL localization is discussed. Two-dimensional nonparametric (kernel) density estimates are employed to obtain the localization maps on 3-D surfaces and the potential regions of cracks are visibly highlighted. This newly proposed procedure is applied to steam reservoir and other experimental vessels. The results presented are accompanied by thoroughgoing evaluation of their quality and practical usefulness.
Trvalý link: http://hdl.handle.net/11104/0291650