The Hilbert Huang Transform (HHT) is a powerful tool for Power Quality (PQ) classifications. One of the main advantages of the HHT is its ability to analyze non-stationary complex waveforms with very good time resolution. However, like other waveform classification techniques, it has difficulty in resolving the instant of sudden changes in the waveform. To overcome the above problems, SAX (Symbolic Aggregate ApproXimation) method is proposed to convert the signal into symbols which facilitates a pattern detector algorithm to identify the border of the stationary signals within a non-stationary signal. Results from simulations will be provided and discussed. © 2012 IEEE.