Edinburgh Napier University

  Induction motor condition monitoring has primarily
relied on the analysis of currents during steady-state operation
through the Fast Fourier Transform (FFT). Nonetheless, this
conventional approach has many constraints, most of them
related to the usual operation of most motors under situations
differing from a pure stationary regime, or directly under
transient conditions. Indeed, these situations are the most
common in many industrial processes, a fact that reveals the
necessity of developing techniques suited for the analysis of
machine quantities during non-stationary operation. A vast work
has been developed during these recent years in this area, raising
many transient-based techniques being able to diagnose machines
under transient conditions which furthermore overcome some of
the drawbacks of conventional stationary analysis. Most of these
techniques are based on the application of proper signal
processing tools, especially adapted to analyze transient signals
(time-frequency decomposition (TFD) tools). This paper carries
out a review of the most significant techniques sustained on
transient-analysis, grouping them in accordance to the nature of
the TFD tool used in each case. The review intends to emphasize
the most relevant contributions of each work and to serve as a
useful reference to all authors involved in the area