Recursive Singular Spectrum Analysis for Induction Machines Unbalanced Rotor Fault Diagnosis

One of the major challenges diagnosing rotor symmetry faults in induction machines (IMs) is severe modulation fault and supply frequency components. In particular, existing techniques are not able to identify components case low slips. this article, problem tackled by proposing a novel approach. First, new use singular spectrum analysis (SSA), as powerful analyzer, introduced for detection. Our idea treat stator current signature wound IM time series. approach, decomposed into several eigenvalue spectra (rather than spectra) find subspace where component recognizable. Subsequently, detected using some data-driven filters constructed with knowledge about characteristics Then, an inexpensive peak localization procedure applied power exact fault. The detection methods then combined recursive regime further improve diagnosis’ performance particularly at high speeds small faults. proposed approach directly raw signal no suppression or filtration harmonics computational complexity. numerical results obtained real data rotation severities unveil effectiveness real-time feature