04-08-2012, 03:47 PM
Automatic Power Quality Disturbance Measurement in Distribution Systems
Automatic Power Quality Disturbance.pdf (Size: 206.13 KB / Downloads: 35)
INTRODUCTION
A classic and comprehensive definition of power quality [1]
describes it as the generated power compatibility to the
electrical/electronic equipment requirements and standards.
Over the past few years, the power engineers’ community has
witnessed a surge in power quality disturbance studies, using
just about every tool available to the designer, ranging from
Fourier analysis to Wavelets and filter theory. All these years
of effort on power quality studies is concentrated on making
the existing and future power distribution systems more and
more reliable and thereby achieve better customer satisfaction.
The Fourier and Wavelet transform method of analyzing [2] –
[6] the power quality disturbance issues have pretty much
evolved into a part and parcel of a power quality engineer’s
daily routine. Such is its awesome compatibility as regard to
analyzing non stationary signals; the Wavelet transform has
nearly dominated the signal processing sector for almost a
decade and is still continuing to stand head and shoulders
above any other signal processing tool/technique.
PROBABILISTIC ENTROPY
The probabilistic entropy used in this paper is based on
Shannon’s entropy. As defined eloquently in [1], the
Shannon’s entropy for a finite event scheme is given by the
formula C (p1, p2, p3, …… pn p pk
the assigned probability with the kth
Automatic Power Quality Disturbance
Measurement in Distribution Systems
event. To express the
waveform deviation of any perfect sinusoidal voltage cycle
with rms that is different from the ideal sine wave necessitated
the application of the probabilistic entropy.
EQUIVALENT DISTURBANCE FACTOR
The equivalent disturbance factor is a new type of power
quality factor, proposed in this paper. The amount of
magnitude deviation, the duration of the deviation and the
corresponding probabilistic entropy have been combined using
Mamdani based fuzzy inference system to produce the
equivalent disturbance factor. The magnitude deviation and the
duration of the deviation serve as the inputs to the fuzzy
system. The output of the fuzzy inference system is defined as
the equivalent disturbance factor. Fig. 3 shows the architecture
of the fuzzy logic based EDF system and Fig. 4 shows
membership functions of the output EDF. Triangular form of
membership functions have been used here to represent the
two input variables and three linguistic variables low (L),
medium (M) and high (H). The output is the EDF which is
represented by five linguistic variables; Very low (VL), low
(L), Medium (M), High (H) and Very high (VH).
CONCLUSION
A new type of power quality factor, the equivalent
disturbance factor (EDF) has been proposed in this paper
through which, automatic measurement/recognition of short
duration disturbances in distribution systems can be
achieved using an artificial neural network. Over 250
different combinations/possibilities of disturbances have
been taken into consideration in this paper by varying the
magnitude of the disturbance (sag and swell) and the
duration of the disturbance event, relevant to distribution
systems. And for each case of disturbance event the
probabilistic entropy has been calculated. The results
obtained from the ANN gives good approximations
regarding the pattern of the disturbance event.