CN112557753A - Online detection method for low-frequency oscillation of Hanning self-convolution window in frequency division band - Google Patents

Online detection method for low-frequency oscillation of Hanning self-convolution window in frequency division band Download PDF

Info

Publication number
CN112557753A
CN112557753A CN202011224055.2A CN202011224055A CN112557753A CN 112557753 A CN112557753 A CN 112557753A CN 202011224055 A CN202011224055 A CN 202011224055A CN 112557753 A CN112557753 A CN 112557753A
Authority
CN
China
Prior art keywords
frequency
low
oscillation
frequency oscillation
window
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202011224055.2A
Other languages
Chinese (zh)
Inventor
王莉
余高旺
石欣
朱云峰
李旭
王智勇
许云龙
蔺立
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xuji Group Co Ltd
XJ Electric Co Ltd
Xuchang XJ Software Technology Co Ltd
Original Assignee
Xuji Group Co Ltd
XJ Electric Co Ltd
Xuchang XJ Software Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xuji Group Co Ltd, XJ Electric Co Ltd, Xuchang XJ Software Technology Co Ltd filed Critical Xuji Group Co Ltd
Priority to CN202011224055.2A priority Critical patent/CN112557753A/en
Publication of CN112557753A publication Critical patent/CN112557753A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R23/00—Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
    • G01R23/16—Spectrum analysis; Fourier analysis
    • G01R23/165—Spectrum analysis; Fourier analysis using filters
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R23/00—Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
    • G01R23/16—Spectrum analysis; Fourier analysis
    • G01R23/177—Analysis of very low frequencies

Landscapes

  • Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Measuring Frequencies, Analyzing Spectra (AREA)

Abstract

The invention discloses a method for detecting Hanning self-convolution window low-frequency oscillation on line in different frequency bands, which is characterized in that aiming at different requirements of different low-frequency oscillation alarm time, an active power oscillation curve calculated by three-phase voltage and three-phase current collected by a broadband oscillation device is divided into low-frequency oscillation frequency bands, and a method for quickly identifying the low-frequency oscillation is realized by applying fast Fourier transform of the Hanning self-convolution window, so that the low-frequency oscillation in different oscillation periods can be quickly detected and timely alarmed, so that operators can find problems in time, and the stable and reliable operation of a power system can be guaranteed.

Description

Online detection method for low-frequency oscillation of Hanning self-convolution window in frequency division band
Technical Field
The invention relates to the technical field of detection and analysis, in particular to a method for detecting Hanning self-convolution window low-frequency oscillation on line in a frequency division mode.
Background
Low frequency oscillations of the power system are inherent in the interconnected system and are mainly characterized in that when the system is interfered, the rotors of the generators swing relatively to each other, so that active power oscillations are generated on the interconnection line. Accurate identification of low-frequency oscillation signals is an important prerequisite for diagnosing and detecting the operation state of the power grid and inhibiting low-frequency oscillation to ensure stable operation of power grid interconnection.
When the low-frequency oscillation occurs in the power system, the online low-frequency oscillation detection program of the broadband measurement device needs to perform rapid spectrum analysis on the active power data measured by the phasor measurement unit PMU, and when the active power oscillation peak value exceeds a preset threshold PoscAnd the broadband oscillation device should perform low-frequency oscillation alarm for X cycles continuously. Common low-frequency oscillation monitoring methods include FFT, Prony, HHT, wavelet analysis and the like. The FFT has the problems of barrier effect, frequency spectrum leakage and the like, Prony is easily influenced by noise, HHT has the problems of end effect and mode aliasing, and the problem of low resolution ratio exists when multi-frequency component signals are extracted through wavelet transformation. Meanwhile, Prony, HHT and wavelet transformation have large calculated amount, which is not beneficial to realizing the function of on-line monitoring of low-frequency oscillation on the device side.
Generally, the oscillation frequency of the low-frequency oscillation is not high, and is usually in the range of 0.1 to 2.5Hz, and the period is 10s to 0.4 s. At least 10s are required to distinguish low frequency oscillations of 0.1Hz, while 0.4s are required to detect low frequency oscillations of 2.5 Hz. If the window is 10s by adopting a fixed window spectrum analysis method, although the oscillation of 0.1Hz can be detected and the oscillation of 2.5Hz can be identified, the oscillation of 2.5Hz may have oscillated for dozens of cycles, which causes low-frequency oscillation alarm delay.
Disclosure of Invention
Objects of the invention
The invention aims to provide a method for detecting Hanning self-convolution window low-frequency oscillation on line in a frequency-division manner, which is used for identifying 0.1-2.5 Hz low-frequency oscillation on line, has high speed, high precision and small calculated amount and meets the requirement of a broadband measuring device on quick alarm of low-frequency oscillation.
(II) technical scheme
In order to solve the above problems, an aspect of the present invention provides a method for on-line detecting hanning self-convolution window low-frequency oscillation in a sub-band, comprising the following steps:
step 1: filtering out direct current components in active power;
step 2: performing frequency band division on the low-frequency oscillation frequency band;
and step 3: and aiming at different frequency bands, performing spectrum analysis by respectively adopting Hanning self-convolution windows FFT with different window lengths.
According to one aspect of the present invention, in step 1, the dc component in the active power is filtered out using the following formula
Figure RE-GDA0002928745520000021
Where M is the power frequency weekly sampling rate.
In step 2, the low frequency oscillation is divided into 5 frequency bands according to one aspect of the invention.
According to one aspect of the invention, the 5 frequency bands are respectively a data window of 1.28s, and low-frequency oscillation of the frequency band of 1.35-2.6 Hz is identified; identifying low-frequency oscillation of a frequency band of 0.7-1.4 Hz by a data window of 2.56 s; identifying low-frequency oscillation of a frequency band of 0.33-0.72 Hz by a data window of 5.12 s; identifying low-frequency oscillation of 0.17-0.35 Hz by a data window of 10.24 s; the 20.48s data window identifies low frequency oscillations of 0.08-0.18 Hz.
According to one aspect of the invention, in step 3, for the oscillation of the high frequency band, a hanning self convolution FFT with a short data window is used; for oscillation in the low frequency band, a long data window hanning self-convolution FFT is used.
According to one aspect of the invention, for different low frequency oscillation frequency bands, up to 5 oscillation cycles of the respective frequency band are required to identify low frequency oscillations.
(III) advantageous effects
The technical scheme of the invention has the following beneficial technical effects:
aiming at different requirements of different low-frequency oscillation alarm time, the invention divides the low-frequency oscillation frequency band of the active power oscillation curve calculated by the three-phase voltage and the three-phase current collected by the broadband oscillation device, and realizes the quick identification method of the low-frequency oscillation by utilizing the quick Fourier transform applying the Hanning self-convolution window, thereby realizing the quick detection and the timely alarm of the low-frequency oscillation with different oscillation periods.
Drawings
Fig. 1 is a hanning self-convolution window interpolation calculation diagram according to an embodiment of the present invention.
Fig. 2 is a diagram of implementation of criteria for low frequency oscillation identification, in accordance with one embodiment of the present invention.
Fig. 3-7 are graphs comparing the results of different frequency band hanning self-convolution FFTs with the same window long hanning FFT, according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings in conjunction with the following detailed description. It should be understood that the description is intended to be exemplary only, and is not intended to limit the scope of the present invention. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present invention.
The following describes a method for detecting low-frequency oscillation of an FFT power system based on hanning self-convolution window according to the present invention with reference to the accompanying drawings.
The invention provides a method for rapidly identifying low-frequency oscillation by utilizing fast Fourier transform applying Hanning self-convolution window to divide low-frequency oscillation frequency bands of an active power oscillation curve calculated by three-phase voltage and three-phase current collected by a broadband oscillation device aiming at different requirements of different low-frequency oscillation alarm time, thereby rapidly detecting and timely alarming the low-frequency oscillation with different oscillation periods.
When the power system has low-frequency oscillation, the expressions of three-phase voltage and current are respectively
Figure RE-GDA0002928745520000031
Figure RE-GDA0002928745520000032
Figure RE-GDA0002928745520000041
Figure RE-GDA0002928745520000042
Figure RE-GDA0002928745520000043
Figure RE-GDA0002928745520000044
Wherein ω is0Is the fundamental angular frequency, ω' is the disturbance frequency; u and I are respectively the voltage amplitude and the current amplitude of the fundamental frequency; u 'and I' are respectively the voltage amplitude and the current amplitude of the disturbance frequency;
Figure RE-GDA0002928745520000045
the initial phases of fundamental frequency A phase voltage and A phase current are respectively; phi is au、φiRespectively A of the disturbance frequencyAnd delta t is the discrete sampling interval of the initial phase of the phase voltage and the phase A current.
Then the instantaneous active power is
Figure RE-GDA0002928745520000046
Wherein
Figure RE-GDA0002928745520000047
Thus, the active power analyzed by the broadband measuring device can be expressed as the sum of the direct current component and the low frequency oscillation component, i.e.
P(n)＝P0+Pmcos(ωmnΔt+φm) (formula 3)
The method comprises the following steps:
step 1: filtering out DC component in active power by using following formula
Figure RE-GDA0002928745520000048
Where M is the power frequency weekly sampling rate.
Step 2: performing frequency band division on the low-frequency oscillation frequency band;
and step 3: for oscillation of a higher frequency band, a Hanning self convolution FFT with a shorter data window is adopted; for oscillation of a lower frequency band, a longer Hanning self convolution FFT is adopted, so that the broadband measuring device can rapidly alarm under different low-frequency oscillation frequencies.
In the step 2, the low-frequency oscillation is divided into 5 frequency bands which are respectively a data window of 1.28s, and the low-frequency oscillation of the frequency band of 1.35-2.6 Hz is identified; identifying low-frequency oscillation of a frequency band of 0.7-1.4 Hz by a data window of 2.56 s; identifying low-frequency oscillation of a frequency band of 0.33-0.72 Hz by a data window of 5.12 s; identifying low-frequency oscillation of 0.17-0.35 Hz by a data window of 10.24 s; the 20.48s data window identifies low frequency oscillations of 0.08-0.18 Hz.
In step 3, the division of all frequency bands is performed on the basis of hanning self-convolution window FFT. The following provides a method for computing hanning self-convolution window FFT.
Hanning self-convolution window FFT of
Figure RE-GDA0002928745520000051
Wherein whn(n) is a Hanning window, whose expression is
Figure RE-GDA0002928745520000052
The DTFT of the Hanning self-convolution window is
Figure RE-GDA0002928745520000053
Let the expression of the active power after differential filtering be
xm(n)＝Amcos(2πfmnΔt+φm) (formula 7)
Wherein A ism、fm、
Figure RE-GDA0002928745520000054
Respectively, amplitude, frequency and initial phase of the band-resolved signal, Δ t is a discrete interval, f s1/Δ t is the sampling rate.
Taking 2N points as the length of the data window, the DTFT of the Hanning self-convolution window applied by the signal is
Figure RE-GDA0002928745520000061
Wherein k is more than or equal to 0 and less than or equal to N-1.
FIG. 1 shows an embodiment of the present invention of a Hanning self-convolution window interpolation calculation graph, which can be used to find the low-frequency oscillation spectral line k to be resolved by searching the spectral line maximummBetween l-1 and l, lines
Figure RE-GDA0002928745520000062
Located in the spectral line
Figure RE-GDA0002928745520000063
And
Figure RE-GDA0002928745520000064
at this time
Figure RE-GDA0002928745520000065
Figure RE-GDA0002928745520000066
Wherein δ ═ (l-k)m)·Δf，
Figure RE-GDA0002928745520000067
fsDelta is more than or equal to 0 and less than or equal to 1.
Then the low frequency oscillation frequency to be resolved is
fmBecomes (l-1- δ) · Δ f (equation 10)
Wherein
Figure RE-GDA0002928745520000068
Corresponding spectral line amplitude:
Figure RE-GDA0002928745520000069
phase position:
Figure RE-GDA00029287455200000610
fig. 2 shows a criterion implementation diagram of low-frequency oscillation identification, and it can be seen from fig. 2 that for different low-frequency oscillation frequency bands, only 5 oscillation cycles of the corresponding frequency band are required at most to identify low-frequency oscillation.
Fig. 3-7 are graphs showing the comparison of the results of different frequency band hanning self-convolution FFTs and the same window long hanning FFT.
When the input signal is
Figure RE-GDA0002928745520000071
And comparing the calculation results of the Hanning self-convolution FFT of different frequency bands with the FFT of the Hanning window with the same window length.
As can be seen from fig. 3, different window lengths identify low frequency oscillations in different frequency bands; meanwhile, under the condition of the same window length, the accuracy of Hanning self-convolution window FFT is higher than that of Hamming window FFT because the side lobe attenuation of the Hanning convolution window is fast. Therefore, the low-frequency oscillation is subjected to frequency band division, and the low-frequency oscillation can be rapidly identified and alarmed; meanwhile, the Hanning convolution window side lobe is attenuated quickly, so that the influence of frequency spectrum leakage of signals can be effectively reduced, the calculation precision is greatly improved, the problem can be found quickly by operators, and the safe and stable operation of a power system is guaranteed.
In summary, the invention provides a method for detecting low-frequency oscillation of a hanning self-convolution window in different frequency bands on line, which divides the low-frequency oscillation frequency bands of an active power oscillation curve calculated by three-phase voltage and three-phase current collected by a broadband oscillation device according to different requirements on different low-frequency oscillation alarm time, and realizes a method for quickly identifying low-frequency oscillation by applying fast fourier transform of the hanning self-convolution window, thereby realizing quick detection and timely alarm of low-frequency oscillation in different oscillation periods. By segmenting the low-frequency oscillation frequency band, the Hanning self-convolution window FFT which reduces the fence effect and the frequency spectrum leakage is adopted, and the low-frequency oscillation component of each segment is quickly identified in an oscillation period of no more than 5 weeks, so that the low-frequency oscillation is quickly alarmed, the problem can be timely found by operating personnel, and the stable and reliable operation of a power system is guaranteed.
It is to be understood that the above-described embodiments of the present invention are merely illustrative of or explaining the principles of the invention and are not to be construed as limiting the invention. Therefore, any modification, equivalent replacement, improvement and the like made without departing from the spirit and scope of the present invention should be included in the protection scope of the present invention. Further, it is intended that the appended claims cover all such variations and modifications as fall within the scope and boundaries of the appended claims or the equivalents of such scope and boundaries.

Claims (6)

1. A method for on-line detecting Hanning self-convolution window low-frequency oscillation with different frequency bands comprises the following steps:
step 1: filtering out direct current components in active power;
step 2: performing frequency band division on the low-frequency oscillation frequency band;
and step 3: and aiming at different frequency bands, performing spectrum analysis by respectively adopting Hanning self-convolution windows FFT with different window lengths.
2. The Hanning self-convolution window low-frequency oscillation online detection method according to claim 1, characterized in that in step 1, the DC component in the active power is filtered out by using the following formula
Figure FDA0002763059760000011
Where M is the power frequency weekly sampling rate.
3. The hanning self-convolution window low-frequency oscillation on-line detection method according to claim 1, wherein in step 2, the low-frequency oscillation is divided into 5 frequency bands.
4. The Hanning self-convolution window low-frequency oscillation online detection method according to claim 3, characterized in that the 5 frequency bands are respectively a data window of 1.28s, and low-frequency oscillation of a 1.35-2.6 Hz frequency band is identified; identifying low-frequency oscillation of a frequency band of 0.7-1.4 Hz by a data window of 2.56 s; identifying low-frequency oscillation of a frequency band of 0.33-0.72 Hz by a data window of 5.12 s; identifying low-frequency oscillation of 0.17-0.35 Hz by a data window of 10.24 s; the 20.48s data window identifies low frequency oscillations of 0.08-0.18 Hz.
5. The Hanning self-convolution window low-frequency oscillation online detection method according to claim 1, characterized in that in step 3, for oscillation in a high frequency band, Hanning self-convolution FFT with a short data window is adopted; for oscillation in the low frequency band, a long data window hanning self-convolution FFT is used.
6. The hanning self-convolution window low-frequency oscillation online detection method according to claim 1, characterized in that for different low-frequency oscillation frequency bands, at most 5 oscillation cycles of the corresponding frequency band are required to identify low-frequency oscillation.
CN202011224055.2A 2020-11-05 2020-11-05 Online detection method for low-frequency oscillation of Hanning self-convolution window in frequency division band Pending CN112557753A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011224055.2A CN112557753A (en) 2020-11-05 2020-11-05 Online detection method for low-frequency oscillation of Hanning self-convolution window in frequency division band

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011224055.2A CN112557753A (en) 2020-11-05 2020-11-05 Online detection method for low-frequency oscillation of Hanning self-convolution window in frequency division band

Publications (1)

Publication Number Publication Date
CN112557753A true CN112557753A (en) 2021-03-26

Family

ID=75042875

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011224055.2A Pending CN112557753A (en) 2020-11-05 2020-11-05 Online detection method for low-frequency oscillation of Hanning self-convolution window in frequency division band

Country Status (1)

Country Link
CN (1) CN112557753A (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101408567A (en) * 2008-11-28 2009-04-15 北京四方继保自动化股份有限公司 Large scale electric network low-frequency oscillation frequency division section detection method based on empirical mode decomposition
CN104062528A (en) * 2014-07-04 2014-09-24 武汉大学 Signal harmonic analysis method and system based on Hanning product window
CN108872402A (en) * 2018-05-08 2018-11-23 天津大学 Ultrasonic wave Butterworth, Hanning window combination with hinder filtering method
CN111398679A (en) * 2020-03-09 2020-07-10 华北电力大学 Sub-synchronous oscillation identification and alarm method based on PMU (phasor measurement Unit)

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101408567A (en) * 2008-11-28 2009-04-15 北京四方继保自动化股份有限公司 Large scale electric network low-frequency oscillation frequency division section detection method based on empirical mode decomposition
CN104062528A (en) * 2014-07-04 2014-09-24 武汉大学 Signal harmonic analysis method and system based on Hanning product window
CN108872402A (en) * 2018-05-08 2018-11-23 天津大学 Ultrasonic wave Butterworth, Hanning window combination with hinder filtering method
CN111398679A (en) * 2020-03-09 2020-07-10 华北电力大学 Sub-synchronous oscillation identification and alarm method based on PMU (phasor measurement Unit)

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
温和;滕召胜;卿柏元;: "Hanning自卷积窗及其在谐波分析中的应用", 电工技术学报, no. 02, pages 5 *

Similar Documents

Publication Publication Date Title
Hassan et al. Induction motor broken rotor bar fault detection techniques based on fault signature analysis–a review
EP2790028B1 (en) Broken rotor bar detection based on current signature analysis of an electric machine
CN106769033B (en) Variable speed rolling bearing fault recognition methods based on order envelope time-frequency energy spectrum
Chow et al. Induction machine fault diagnostic analysis with wavelet technique
Climente-Alarcon et al. Application of the Wigner–Ville distribution for the detection of rotor asymmetries and eccentricity through high-order harmonics
CN103245832B (en) Based on harmonic wave time-frequency characteristic method for parameter estimation and the analyser of quick S-transformation
CN113748326B (en) Method for estimating bearing fault severity for induction motor
US7409854B2 (en) Method and apparatus for determining an operating status of a turbine engine
JP6945371B2 (en) Diagnostic equipment for rotating machine systems, power converters, rotating machine systems, and diagnostic methods for rotating machine systems.
Tang et al. Fractional frequency band entropy for bearing fault diagnosis under varying speed conditions
Liu et al. An online bearing fault diagnosis technique via improved demodulation spectrum analysis under variable speed conditions
JP2023513641A (en) System and method for fault detection based on robust attenuation signal separation
JP7198089B2 (en) POWER CONVERTER, ROTATING MACHINE SYSTEM, AND DIAGNOSTIC METHOD
CN113361331B (en) Power Frequency Interference Elimination Method, System and Medium Based on Windowed Interpolation FFT
CN110849645A (en) Initial diagnosis method for GIS mechanical fault
Gouws Active magnetic bearing condition monitoring
CN114325740A (en) Speed measurement method and system applied to laser Doppler velocimeter
CN112557753A (en) Online detection method for low-frequency oscillation of Hanning self-convolution window in frequency division band
Zhu et al. Fault diagnosis method for rolling element bearings under variable speed based on TKEO and Fast-SC
CN105044458B (en) A kind of trend of harmonic detection method of power based on TT conversion
Liu et al. A new motor fault detection method using multiple window S-method time-frequency analysis
CN117289087A (en) Series fault arc detection method based on CZT conversion
CN116706852A (en) Differential current-based protection method, system, equipment and storage medium
Desheng et al. Time-frequency analysis based on BLDC motor fault detection using Hermite S-method
Huang et al. Short-time kurtogram for bearing fault feature extraction under time-varying speed conditions

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination