Open AccessClamping diodes failure identification based on the discrete wavelet decomposition of the magnetic near‐field
Author(s) -
Lahouar Ali,
Hamouda Mahmoud,
Abari Ibtissem,
Slama Jaleleddine Ben Hadj
Publication year - 2019
Publication title -
iet power electronics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.637
H-Index - 77
eISSN - 1755-4543
pISSN - 1755-4535
DOI - 10.1049/iet-pel.2018.5043
Subject(s) - clamping , fault (geology) , electronic engineering , converters , power (physics) , reliability (semiconductor) , computer science , wavelet , engineering , voltage , electrical engineering , artificial intelligence , physics , mechanical engineering , quantum mechanics , seismology , geology
Diagnosis of static power converters for distributed power generation systems and electric machines is nowadays a common concern of researchers interested in the reliability of power electronic systems. The conventional diagnosis approaches such as those based on voltage, current, and flux analysis have already been well developed in recent decades. Recently, it was shown that the magnetic near‐field measured above the power converter may carry substantial information about the health state of power components. This study proposes, therefore, a novel non‐invasive diagnosis method based on the discrete wavelet decomposition of the near‐field metered above a static power converter. It is proved that each faulty component can be identified by its appropriate electromagnetic signature. Accordingly, all faulty cases are easily distinguishable, which makes fault classification possible using very simple tools such as envelope detection. The proposed method is validated through experimental tests carried out on a three‐phase three‐level neutral point clamped inverter operating with faulty clamping diodes. Its advantages compared to previous diagnosis methods are also discussed.