Pattern Recognition of Power Quality Disturbances Based on Continuous Wavelet Transform.

A methodology to recognize and classify electric power quality (PQ) disturbances is proposed as a part of a solution for improving the diagnostic and making decision processes of planning and operation engineers in the distribution networks. This methodology used the continuous wavelet transform (CWT) for the PQ recognition and the fuzzy sets for the PQ classification which was implemented in a virtual instrument (VI) in LABVIEW. This allows for reducing the uncertainty when variations in the reference waveforms appear. The developed VI has been thoroughly tested using signals of ATP-EMTP simulations with PQ events such as harmonics, voltage transients, sags and swells. Results show that the methodology proposed can achieve more accurate and robust recognition and classification process of PQ disturbance. [ABSTRACT FROM AUTHOR]