Diagnosis of stator short-circuits through Wigner-Ville transient-based analysis

Diagnosis of stator inter-turn and inter-coil short-circuits is a matter of increasing concern in the industry. The identification of the specific components amplified by these faults in the Fourier spectrum of the steady-state current has been a reason for controversy between different groups of authors. Alternative diagnosis of these faults based on the analysis of transient processes of the machine has led to interesting results, showing characteristic evolutions of fault-related components able to be detected by modern time-frequency decomposition (TFD) tools such as wavelet or Hilbert transforms. These evolutions arise as reliable indicators of the presence of the fault in the machine. In this paper, the Wigner-Ville and the Choi-Williams distributions are applied for the tracing of stator short-circuit-related components during the startup transient. This tool might enhance some of the features of other alternative TFD tools such as the discrete wavelet transform (DWT), with regards to the diagnosis of these faults. Simulation and experimental results prove the ability of this tool, providing interesting conclusions on how the components related to these faults evolve during the transient.