作者: M. Hernandez-Vargas , E. Cabal-Yepez , A. Garcia-Perez
DOI: 10.1016/J.COMPELECENG.2013.12.020
关键词:
摘要: Novel methodology for induction motor condition monitoring is proposed.Singular value decomposition, statistical analysis and an artificial neural network.One-broken-bar, unbalance, faulty bearing their combinations are identified.One-phase of the startup-transient current analyzed in real time. Early detection faults has been a main subject investigation many years. Several approaches have proposed identifying one or more treated isolated way. Multiple combined on motors represent big challenge since reliable diagnosis under presence two simultaneous really difficult. This work introduces novel that merges singular analysis, networks multiple fault identification. Obtained results demonstrate its high effectiveness detecting bearings, broken rotor bars, all possible combinations. The developed field programmable gate array-based implementation offers portable low-cost solution online classification rotating machine Thanks to generalized nature, introduced approach can be extended different working conditions by proper calibration.