Abstract
In recent years, the study of transients of induction motors for diagnosis purposes has gained strength in order to overcome some inherent problems of the classical diagnosis of such machines, which uses Fourier analysis of steady-state quantities.
Novel timefrequency techniques have been applied to these transient quantities, in order to detect the characteristic evolutions of fault-related harmonic components. The detection of these patterns, which are usually specific for each type of fault, enables reliable diagnostic of the corresponding failures. In this context, most of the works hitherto developed have been based on analysis of currents. However, in some applications, vibration measurements are also available. The goal of this work is to validate the applicability of this transient-based diagnosis framework to vibrationmeasurements. A specific timefrequency decomposition tool, the ZhaoAtlasMarks distribution, is proposed. Experimental
results prove the ability of the approach to complement the information obtained from the current analysis. This may be very useful in applications in which the diagnosis via currents is uncertain or in which vibration signals can be easily measured.