A Comparison of Two Methods Using R.M.S. Value and Wavelet-Based Mann and Morrison Algorithm for Voltage Dip Characterization

Voltage dips have been identified as the major problem among power quality
disturbance events which affect the industrial customers. The dips involve short
reduction in Root Mean Square (RMS) voltage caused by fuults in electrical supply
system or starting of a large load. It is crucial to measure and analyze the voltage dip
events before considering any mitigation action to reduce and eliminate it. This work
presents comparison of two methods used for voltage dip detection and
characterization. A less complicated method performed is to determine the lowest
magnitude of the RMS voltage during the disturbance. Another method is a
combination of wavelet-based analysis using Mann and Morrison algorithm which
estimates the amplitude and the phase angle to characterize the dips. The
performances of both methods are examined using simulation of system voltage dip,
duration of the dip occurrences and point-on wave at beginning in a sinusoidal
voltage supply. The voltage dips are then being classified into some classes
according to their characteristics. The numerical approach using RMS value results
in simplicity method which is easy to be implemented and understood. Eventhough
the wavelet-based method seems to be more complicated, it provides higher accuracy
in determining the voltage dip characterization.
iii