Harmonic Detection in Electric Power System Based on
Wavelet Multi-resolution Analysis
Yu Chen
Zhengzhou Institute of Aeronautical Industry Management
Zhengzhou, China
e-mail: chenyu@zzia.edu.cn
Abstract - Because of widespread uses of nonlinear load,
electric power system network is injected with much
harmonic current, which does great harm to consumer
equipment. In order to prevent harmonic current from
influencing safety of system’s operation, we should know
well how much distorted harmonic wave contained and take
corresponding measurement to control or compensate it. In
this paper, based on a kind of multi-resolution wavelet
method, we use seven levels of restructuring through using
Daubechies wavelet (db24) to decompose the electric current
signal into fundamental wave and higher harmonic. By
experiment analysis of simulation software MATLAB,
separated fundamental wave error is less than 1%, which
realizes harmonic effective tracking and gives accuracy
compensation by making use of total distorted component.
Index Terms -Wavelet transform
;
multi-resolution wavelet
analysis; harmonic detection; electric power quality
I. INTRODUCTION
Electronic power system has produced fundamental wave
and also brought harmonic since its birth. In last few years,
because of widely uses of non-linear equipments, such as
transformer, locomotives, rotating electric machines, ARC
furnace, fluorescent lamp, electric controller and thyristor, etc.
[1] A mass of harmonic has been produced, which will bring
about serious distortion of voltage or electric current for
electronic power system, the increase of electrical component
loss, the overheating of equipments and the loss of its life, the
disturbance of relay protection, controlling and communication
circuit so well as the disturbance of customers’ load, etc. The
above mentioned badly affect the safe operation of electronic
power system. At present, harmonic has become one of serious
damages to electric power system. It seems extremely urgent to
solve the problem of harmonic for electric power system.
Harmonic detection is an important branch of harmonic
problems, which can play guiding role in harmonic restraining.
Hence, the detection and analysis of harmonic is an important
task to analyse and control the electronic power system.
II.
HARMONIC DETCTION METHOD IN ELECTRIC POWER SYSTEM
In general, the method of harmonic detection includes DFT
and FFT. The conversion of time domain signal into frequency
domain signal is its main principle. The principle is easy to
understand, the realization is convenient, and the function is
versatile. But owing to failing effect, namely deviation of
sampling frequency as well as undulation of frequency in
electrical network in practice, all can cause outer-sync of
sampling. As a result, the central frequency of sub-harmonics
can’t happen to locate in the central point of the filter. The
asynchronous of sampling for periodic signals will bring about
spectrum leakage. Thereby, it will become inaccurate to the
detection of high-order harmonic. Usually, window function and
interpolation method are applied to decrease spectrum leakage
to some extent. Gabor brought forward STET which has certain
time-resolution and can make up the insufficiency of DFT and
FFT. But for the invariability of STFT’s time-frequency
window, the contravention between time resolution and
frequency resolution is unable to be overcome. The problem of
time-frequency localization has not completely solved. As to
practical application, there are many limitations. [2]
Wavelet transform is provided with a time-frequency
localization character. That is to say, it has higher frequency
resolution and lower time resolution in low frequency ranges
and vice versa. The width of time window and frequency
window can be adjusted, which can easily detect transient signal
and unsteady signal. Also, it can easily get access to the signal’s
frequency distribution in specific time quantum, so the mixed-
signals, which contain all kinds of interweaved frequencies, will
resolve into signal-block with different frequency. Therefore,
the wavelet transform is widely used in project areas such as
signal processing, fault diagnosis and so on.
Based on the method of orthonormal wavelet transform,
this paper carries on wavelet transform for the current signal
with harmonic component in electronic network of power
system, to pick up higher scale transformed value and obtain the
fundamental wave component. In addition, owing to locality of
wavelet in frequency domain and simultaneously in time
domain, thus, tracking the harmonic effectively can carry out
the real time detection of harmonic.
III.
WAVELET TRANSFORM AND MULT-RESOLUTION
DETECTION ALGORITHM
A. Wavelet Transform
Due to locality of waveform function and its Fourier
transform in time and frequency domains, so that makes it
suitable for decomposition function and time-frequency
2008 International Conference on Computer Science and Software Engineering
978-0-7695-3336-0/08 $25.00 © 2008 IEEE
DOI 10.1109/CSSE.2008.18
1204
2008 International Conference on Computer Science and Software Engineering
978-0-7695-3336-0/08 $25.00 © 2008 IEEE
DOI 10.1109/CSSE.2008.18
1204
Authorized licensed use limited to: Institute for Sci Tech Information of China. Downloaded on November 15, 2009 at 19:23 from IEEE Xplore. Restrictions apply.