CN108767879A - A kind of power system oscillation pattern Fast Identification Method based on stochastic subspace - Google Patents
A kind of power system oscillation pattern Fast Identification Method based on stochastic subspace Download PDFInfo
- Publication number
- CN108767879A CN108767879A CN201810686260.7A CN201810686260A CN108767879A CN 108767879 A CN108767879 A CN 108767879A CN 201810686260 A CN201810686260 A CN 201810686260A CN 108767879 A CN108767879 A CN 108767879A
- Authority
- CN
- China
- Prior art keywords
- matrix
- oscillation
- characteristic value
- frequency
- oscillation modes
- 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
Links
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/24—Arrangements for preventing or reducing oscillations of power in networks
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Complex Calculations (AREA)
Abstract
The invention discloses a kind of control oscillation modes Fast Identification Method based on stochastic subspace, the described method comprises the following steps：The measurement information that electric system is obtained from wide area measurement system, the extension Observable matrix of electric system stochastic subspace is built according to measurement information；According to the state matrix of the extension Observable matrix computing system reduced order state equation of stochastic subspace；Calculate the inverse matrix of the generalized inverse matrix of state matrix；The characteristic value of state matrix and inverse matrix in continuous space is calculated, the characteristic value in low-frequency oscillation of electric power system frequency separation is only retained；Based on the similar quasi- side of mode, the control oscillation modes of system are picked out.The present invention is improved the electric system control oscillation modes recognized based on stochastic subspace and recognizes efficiency under the premise of ensureing control oscillation modes identification precision.
Description
Technical field
The present invention relates to field of power more particularly to a kind of power system oscillation pattern based on stochastic subspace are fast
Fast discrimination method.
Background technology
Regional power grid interconnection scale constantly expands, regenerative resource large-scale grid connection so that Inter-area power oscillation has become
Ability to transmit electricity between restricted area, a key factor for jeopardizing power system security reliability service.Therefore, how quickly, accurately,
Reliably the control oscillation modes of Identification of Power System have highly important engineering real to improving Operation of Electric Systems safety
With value.
Currently, the control oscillation modes identification of electric system is divided into the oscillation mode based on electric system differential-algebraic model
Formula analysis method and Oscillation mode analysis method based on electrical power system wide-area measurement information.Based on electric system differential-algebraically
The Oscillation mode analysis method of model need to build the detail mathematic model of power system device and system and accurate model parameter,
And analysis result is generally only applicable to given system operating point, and when system is far from given operating point, analysis result
The actual oscillation mode of system cannot be accurately reflected, thus based on the Oscillation mode analysis side of electric system differential-algebraic model
Method is mainly used in Power System Off-line safety analysis and Power System Planning.
Based on the Oscillation mode analysis method of electrical power system wide-area measurement information by the correlation theory of pattern-recognition, from electricity
Identification system existing control oscillation modes in the process of running in the wide area measurement information of Force system, since wide area measurement is believed
Breath is directed to the actual motion status information of electric system, can identification result can really reflect system in actual motion
Existing control oscillation modes, thus the Oscillation mode analysis method based on electrical power system wide-area measurement information is mainly used for electricity
The real time on-line monitoring of Force system dynamic stability and control.
According to the difference of electrical power system wide-area measurement information type, the electric system control oscillation modes based on wide area measurement
Discrimination method can be divided into：Control oscillation modes discrimination method based on fault-signal and the leading oscillation mode based on noise-like signal
Formula discrimination method.Control oscillation modes discrimination method based on fault-signal is mainly recognized by the method for oscillating curve fitting
The control oscillation modes of system, such as Prony methods and wavelet transformation；Control oscillation modes discrimination method based on noise-like signal
Mainly slided come the control oscillation modes of identification system, such as autoregression by the transfer function matrix of identification system or subspace method
Movable model method.
The above method is limited by pattern identification theory, is only applicable to handle the electrical power system wide-area amount of above-mentioned a certain type
Measurement information, and the measurement channel of practical power systems may both collect Fault Signal Analyses in HV Transmission or can collect electric system
Noise-like signal, this requires there are it is a kind of can usually handling failure signal and noise-like signal control oscillation modes identification
Method.Stochastic subspace identification have proved to be it is a kind of can simultaneously from the fault-signal and noise-like signal of electric system effectively
Pick out the pattern identification method of system control oscillation modes.But how from stochastic subspace identification result to efficiently separate out electricity
The control oscillation modes of Force system and false oscillation mode, are the electric system control oscillation modes recognized based on stochastic subspace
One of the main problem that discrimination method is faced.
To efficiently separate out the control oscillation modes in stochastic subspace identification result and false oscillation mode, according to leading
Oscillation mode does not change with system order reduction model dimension and is changed, and false oscillation mode changes with system order reduction model dimension and
Acute variation feature isolates the leading oscillation of system by repeating to recognize and calculate the subspace method of multiple and different dimensions
Pattern.And repeat to recognize and calculate the subspace method of multiple and different dimensions, it will necessarily reduce and system is recognized using stochastic subspace
The efficiency of system control oscillation modes, reduces the real-time of power system dynamic stability on-line monitoring.
Invention content
The present invention provides a kind of power system oscillation pattern Fast Identification Method based on stochastic subspace, the present invention exist
Under the premise of ensureing control oscillation modes identification precision, the electric system control oscillation modes recognized based on stochastic subspace are improved
Efficiency is recognized, it is described below：
A kind of control oscillation modes Fast Identification Method based on stochastic subspace, the described method comprises the following steps：
The measurement information that electric system is obtained from wide area measurement system builds electric system with loom according to measurement information
The extension Observable matrix in space；According to the state of the extension Observable matrix computing system reduced order state equation of stochastic subspace
Matrix；
Calculate the inverse matrix of the generalized inverse matrix of state matrix；
The characteristic value of state matrix and inverse matrix in continuous space is calculated, is only retained in low-frequency oscillation of electric power system frequency
The characteristic value in rate section；Based on the similar quasi- side of mode, the control oscillation modes of system are picked out.
Further, the inverse matrix of the generalized inverse matrix for calculating state matrix is specially：
1) state matrix A is calculated0Generalized inverse matrix
2) inverse matrix is calculatedInverse matrix A1：
A1=(O↓TO↑)-1O↓TO↓
Wherein, O↑And O↓For submatrix；T is transposition.
When specific implementation, the characteristic value for calculating state matrix and inverse matrix in continuous space only retains in electricity
The characteristic value of Force system low-frequency oscillation frequency separation is specially：
1) state matrix A is calculated separately0With inverse matrix A1Eigenvalue λ in continuous space0,iAnd λ1,i；
2) λ is deleted respectively0,iAnd λ1,iMiddle characteristic value is real number and the minus complex eigenvalue of imaginary part, is then further deleted
Characteristic value of the complex eigenvalue imaginary part except section [0.4 π, 4 π], only retains λ0,iAnd λ1,iMiddle complex eigenvalue imaginary part is in section
Characteristic value within [0.4 π, 4 π]；
3) step 2) will be passed through treated that characteristic value is stored in λ respectively0And λ1, by state matrix A0In the spy of continuous space
The characteristic value that value indicative remains after step 2) processing is stored in λ0In, by inverse matrix A1Continuous space characteristic value through step
2) the characteristic value deposit λ remained after handling1In.
Further, described to be based on the similar quasi- side of mode, the control oscillation modes for picking out system are specially：
If 1) λ0In complex eigenvalue λ0,iWith λ1In complex eigenvalue λ1,jMeet criterion, then it is assumed that λ0,iAnd λ1,jFor
The same control oscillation modes of system；
2) according to from λ0And λ1In the same control oscillation modes λ of system that isolates0,iAnd λ1,i, the leading of computing system shake
Swing pattern λkAnd corresponding frequency of oscillation fkWith damping ratio ξk。
Further, the criterion is specially：
In formula, f0,iAnd ξ0,iRespectively λ0In the i-th oscillation mode λ0,iFrequency of oscillation and damping ratio；f1,jAnd ξ1,jRespectively
For λ1Middle jth oscillation mode λ0,jFrequency of oscillation and damping ratio；Δ f, Δ ξ and Δ m are respectively λ0,iWith λ1,jFrequency of oscillation it is inclined
Difference, damping ratio deviation and oscillation mode deviation；WithRespectively λ0,iWith λ1,jOscillation frequency deviation, damping ratio deviation
With the threshold value of oscillation mode deviation.
When specific implementation, the control oscillation modes λ of the systemkAnd corresponding frequency of oscillation fkWith damping ratio ξkSpecially：
In formula, imag (λk) statement take λkImaginary part；real(λk) statement take λkReal part；|λk| statement takes λkModulus value.
The advantageous effect of technical solution provided by the invention is：
1, the present invention realizes quick, the accurate recognition of electric system control oscillation modes, improves and is based on stochastic subspace
Electric system control oscillation modes identification efficiency；
2, the present invention is advantageously implemented the quick early warning of power oscillation of power system, and for operation of power networks, dispatcher provides more
The control strategy of suppression system oscillation of power is formulated for the sufficient time；
3, the present invention needs not rely on the accurate mathematical model of electric system based entirely on electrical power system wide-area metric data
With accurate parameter value, you can realize accurate, the quick calculating of electric system control oscillation modes；
4, the opposite eigenvalue method based on higher-dimension electric system mathematic(al) mode of the present invention, computational efficiency is faster；
5, the present invention can not only pick out the control oscillation modes of system, Er Qieke in the fault-signal of electric system
Pick out the control oscillation modes of system from the noise-like signal of electric system, the present invention is opposite, and leading based on Prony shakes
It swings pattern algorithm and has more versatility；6, it is dynamic can to provide more fast and accurately power grid for operation of power networks dispatcher by the present invention
State stablizes information, the advantageous response speed for promoting power system dynamic stability Situation Awareness.
Description of the drawings
Fig. 1 is a kind of power system oscillation pattern Fast Identification Method flow chart based on stochastic subspace；
Fig. 2 is the active power variation diagram that CHINA SOUTHERN POWER Chu Sui direct current monopoles are latched on interconnection YC；
Wherein, YC is circuit number.
Fig. 3 is state matrix A0Characteristic value result of calculation figure in discrete space；
Fig. 4 is state matrix A1Characteristic value result of calculation figure in discrete space；
Fig. 5 is state matrix A0Characteristic value result of calculation figure in continuous space；
Fig. 6 is state matrix A1Characteristic value result of calculation figure in continuous space；
Fig. 7 is state matrix A0Characteristic value result figure after pretreatment in continuous space；
Fig. 8 is state matrix A1Characteristic value result figure after pretreatment in continuous space.
Specific implementation mode
To make the object, technical solutions and advantages of the present invention clearer, embodiment of the present invention is made below further
It is described in detail on ground.
In order to solve to recognize efficiency about the electric system control oscillation modes based on stochastic subspace in background technology
Deficiency, the embodiment of the present invention propose a kind of electric system control oscillation modes Fast Identification Method based on stochastic subspace, with
The Fast Identification for realizing electric system control oscillation modes, more fast and accurately power grid is provided for operation of power networks dispatcher
Dynamic stability information promotes the response speed of power system dynamic stability Situation Awareness.
Embodiment 1
A kind of control oscillation modes Fast Identification Method based on stochastic subspace, referring to Fig. 1, this method includes following step
Suddenly：
101：From wide area measurement system obtain electric system measurement information, according to measurement information build electric system with
The extension Observable matrix in loom space；
102：According to the state matrix of the extension Observable matrix computing system reduced order state equation of stochastic subspace；
103：Calculate the inverse matrix of the generalized inverse matrix of state matrix；
104：The characteristic value of state matrix and inverse matrix in continuous space is calculated, only retains and shakes in electric system low frequency
Swing the characteristic value of frequency separation；
105：Based on the similar quasi- side of mode, the control oscillation modes of system are picked out.
Wherein, the characteristic value of calculating state matrix and inverse matrix in continuous space in step 104 only retains in electricity
The characteristic value of Force system low-frequency oscillation frequency separation is specially：
1) state matrix A is calculated separately0With inverse matrix A1Eigenvalue λ in continuous space0,iAnd λ1,i；
2) λ is deleted respectively0,iAnd λ1,iMiddle characteristic value is real number and the minus complex eigenvalue of imaginary part, is then further deleted
Characteristic value of the complex eigenvalue imaginary part except section [0.4 π, 4 π], only retains λ0,iAnd λ1,iMiddle complex eigenvalue imaginary part is in section
Characteristic value within [0.4 π, 4 π]；
3) step 2) will be passed through treated that characteristic value is stored in λ respectively0And λ1, by state matrix A0In the spy of continuous space
The characteristic value that value indicative remains after step 2) processing is stored in λ0In, by inverse matrix A1Continuous space characteristic value through step
2) the characteristic value deposit λ remained after handling1In.
Further, in step 105 based on the similar quasi- side of mode, the control oscillation modes for picking out system are specially：
If 1) λ0In complex eigenvalue λ0,iWith λ1In complex eigenvalue λ1,jMeet criterion, then it is assumed that λ0,iAnd λ1,jFor
The same control oscillation modes of system；
2) according to from λ0And λ1In the same control oscillation modes λ of system that isolates0,iAnd λ1,i, the leading of computing system shake
Swing pattern λkAnd corresponding frequency of oscillation fkWith damping ratio ξk。
In conclusion 101- steps 105 are ensureing control oscillation modes identification essence to the embodiment of the present invention through the above steps
Under the premise of degree, improves the electric system control oscillation modes recognized based on stochastic subspace and recognize efficiency, meet and actually answer
A variety of needs in.
Embodiment 2
The scheme in embodiment 1 is further introduced with reference to specific calculation formula, example, it is as detailed below
Description：
201：It is electric from being obtained in wide area measurement system (well known to a person skilled in the art technical terms, and this will not be repeated here)
The measurement information of Force system, by the extension Observable matrix O of wide area measurement information architecture electric system stochastic subspace, packet
It includes：
1) the Hankel matrix Hs of wide area measurement signal y (t) structure electric system stochastic subspaces are utilized：
In formula, YPAnd YFExpression formula is respectively：
In formula, n is the maximal dimension of pre-set system order reduction state equation；B is the length of wide area measurement signal y (t)
Degree.
2) LQ decomposition is carried out to Hankel matrix Hs：
Wherein, Hankel matrix Hs, LQ decomposition are common-sense technology known in those skilled in the art, and the present invention is real
Applying example, this will not be repeated here.
In addition, Q1, Q2Only intermediate variable, without physical meaning.
In formula, L11、L21And L22Expression formula is as follows：
3) according to matrix L11、L21And L22Calculating matrix ∑PP、∑FPAnd ∑FF：
Wherein, ∑PP、∑FPAnd ∑FFOnly intermediate variable, without physical meaning.
4) respectively to matrix ∑PPAnd ∑FFChebyshev's Factorization is carried out, matrix L is obtained1And L2：
5) by matrix L1And L2To matrix ∑FPIt is normalized, matrix is after obtaining normalized
(subscript " -1 " indicates that, to matrix inversion, subscript " T " is indicated to Matrix Calculating transposition in formula).
6) to the matrix after normalizedCarry out singular value decomposition：
Wherein, U is left singular vector matrix；S is singular value matrix；V is right singular vector matrix.
7) the extension Observable matrix O of electric system stochastic subspace is calculated：
In formula, extension Observable matrix O expression formulas are：
202：According to the state matrix A of the extension Observable matrix O computing system reduced order state equations of stochastic subspace0,
Including：
1) the extraction submatrix O from extension Observable matrix O↑And O↓：
2) according to matrix O↑And O↓The state matrix A of computing system reduced order state equation0：
A0=O↑+O↓ (15)
In formula, subscript "+" is indicated to Matrix Calculating pseudoinverse, matrix O↑And O↓Only intermediate variable.
203：According to the state matrix A calculated in step 2020, calculate A0Generalized inverse matrix inverse matrix A1, including：
1) state matrix A is calculated0Generalized inverse matrix
2) inverse matrix is calculatedInverse matrix A1：
A1=(O↓TO↑)-1O↓TO↓ (17)
204：To state matrix A0With inverse matrix A1Characteristic value pre-processed, only retain shake in electric system low frequency
The characteristic value of frequency separation is swung, including：
1) state matrix A is calculated separately0With inverse matrix A1Characteristic value in continuous space：
In formula, μ0,iAnd λ0,iRespectively state matrix A0In the characteristic value corresponding to discrete space and continuous space；μ1,iWith
λ1,iRespectively inverse matrix A1In the characteristic value corresponding to discrete space and continuous space；When t is the sampling of wide area measurement system
Between.
2) λ is deleted respectively0,iAnd λ1,iMiddle characteristic value is real number and the minus complex eigenvalue of imaginary part, is then further deleted
Characteristic value of the complex eigenvalue imaginary part except section [0.4 π, 4 π] (π is pi), only retains λ0,iAnd λ1,iMiddle complex eigenvalue is empty
Characteristic value of the portion within section [0.4 π, 4 π].
Since the common frequency of oscillation of electric system is between 0.2Hz-2Hz, according to poor 2 π times between frequency and angular frequency,
So complex eigenvalue imaginary part is within [0.2*2 π, 2*2 π]=[0.4 π, 4 π] of section.
3) step 2) will be passed through treated that characteristic value is stored in λ respectively0And λ1, wherein by state matrix A0In continuous space
Characteristic value after step 2) processing the characteristic value that remains be stored in λ0In, by inverse matrix A1It is passed through in the characteristic value of continuous space
The characteristic value deposit λ remained after step 2) processing1In.
205：Based on the similar quasi- side of mode, the control oscillation modes of system are picked out, including：
1) it is based on the similar quasi- side of mode, if λ0In complex eigenvalue λ0,iWith λ1In complex eigenvalue λ1,jMeet formula (19), then
Think λ0,iAnd λ1,jFor the same control oscillation modes of system：
In formula, f0,iAnd ξ0,iRespectively λ0In the i-th oscillation mode λ0,iFrequency of oscillation and damping ratio；f1,jAnd ξ1,jRespectively
For λ1Middle jth oscillation mode λ0,jFrequency of oscillation and damping ratio；Δ f, Δ ξ and Δ m are respectively λ0,iWith λ1,jFrequency of oscillation it is inclined
Difference, damping ratio deviation and oscillation mode deviation；WithRespectively λ0,iWith λ1,jOscillation frequency deviation, damping ratio deviation
With the threshold value of oscillation mode deviation, in the embodiment of the present inventionWithValue be respectively 0.02,0.03 and 0.05, specifically
Value according in practical application set, and the embodiment of the present invention is without limitation.
2) according to from λ0And λ1In the same control oscillation modes λ of system that isolates0,iAnd λ1,i, the leading of computing system shake
Swing pattern λkAnd corresponding frequency of oscillation fkWith damping ratio ξk：
In formula, imag (λk) statement take λkImaginary part；real(λk) statement take λkReal part；|λk| statement takes λkModulus value.
In conclusion 201- steps 205 realize the electricity based on stochastic subspace to the embodiment of the present invention through the above steps
The Fast Identification of Force system control oscillation modes avoids and detaches control oscillation modes using estimating system state matrix is repeated
And the influence of identification efficiency is reduced, the electric system control oscillation modes identification efficiency based on stochastic subspace is improved, is realized
Electric system control oscillation modes Fast Identification based on wide area measurement information is conducive to the dynamic stability for promoting electric system
The response speed of Situation Awareness.
Embodiment 3
With reference to specific example, Fig. 2-Fig. 8 and table 1- tables 3, feasibility is carried out to the scheme in Examples 1 and 2
Verification, it is described below：
This example is to verify the embodiment of the present invention 1 and 2 by taking the control oscillation modes Fast Identification of CHINA SOUTHERN POWER as an example
Validity.
Based on CHINA SOUTHERN POWER running mode data in 2013, during time-domain-simulation, Chu Suizhi is set
It flows monopole and is latched failure, the monopole in locking failure is out of service after duration 0.1s, 0.1s.Selection is in Yunnan and extensively
The active power of research object of the active power of interconnection YC between western power grid as this method, disturbed rear interconnection YC becomes
Change as shown in Figure 2.
Input using the active power emulation data of 1s~20s on interconnection YC as the present embodiment, first setting system
The maximal dimension n of system reduced order state equation is 200, builds the Hankel matrix Hs of stochastic subspace；According to Hankel matrix Hs point
Other calculating matrix L11、L21L22、∑PP、∑FPAnd ∑FF；Then to matrix ∑PPAnd ∑FFChebyshev's Factorization is carried out, is obtained
Matrix L1And L21；By matrix L1And L21To matrix ∑FPIt is normalized, gets matrix after normalizedTo matrixSingular value decomposition is carried out, matrix U is obtained1And S1；Further according to L1、 U1And S1By formula
(11) extension Observable matrix O, and the extraction matrix O from extension Observable matrix O are calculated↑And O↓；Then according to formula (15) and
(16) state matrix A is calculated separately0With inverse matrix A1。
State matrix A is calculated separately on the basis of aforesaid operations0With inverse matrix A1Characteristic value μ in discrete space0,i
And μ1,i, as a result as shown in Figure 3 and Figure 4；Further according to formula (18) by state matrix A0With inverse matrix A1In discrete space
Characteristic value μ0,iAnd μ1,iIt is converted into state matrix A0With inverse matrix A1The corresponding eigenvalue λ in continuous space0,iAnd λ1,i, as a result
As shown in Figure 5 and Figure 6；Characteristic value in Fig. 5 and Fig. 6 is pre-processed, λ is only retained0,iAnd λ1,iMiddle complex eigenvalue imaginary part is
Positive number, and characteristic value of the imaginary values within section [0.4 π, 4 π], as a result as shown in Figure 7 and Figure 8.
As seen from Figure 7, to state matrix A0After characteristic value in continuous space is pre-processed, two spies are only remained
Value indicative, respectively λ0,1=-0.3066+i5.1094 and λ0,2=-0.1554+i2.5279；As seen from Figure 8, to inverse matrix A1?
After characteristic value in continuous space is pre-processed, two characteristic values are only remained, respectively：
λ1,1=-0.3185+i5.0956 and λ1,2=-0.1556+i2.5282.
1 mode Similarity measures result of table
Further, state matrix A is calculated separately according to formula (19)0With inverse matrix A1Oscillation frequency between the characteristic value of reservation
Rate deviation delta f, damping ratio deviation delta ξ and oscillation mode deviation delta m, the results are shown in Table 1.Then according to given frequency of oscillation
Deviation thresholdDamping ratio deviation thresholdWith oscillation mode deviation thresholdλ can be obtained according to formula (19)0,1=-0.3066+
I5.1094 and λ1,1=-0.3185+i5.0956 corresponds to a control oscillation modes of CHINA SOUTHERN POWER；λ0,2=-
0.1554+i2.5279 and λ1,2=-0.1556+i2.5282 corresponds to another control oscillation modes of CHINA SOUTHERN POWER.
According to the above results, by λ0,1And λ1,1With λ0,2And λ1,2Substitute into formula (20) respectively, two groups of computing system leading to shake
Pattern is swung, is as a result respectively -0.1555+i2.5280 and -0.3125+i5.1025.Wherein oscillation mode -0.1555+i2.5280
Frequency of oscillation be 0.4024Hz, damping ratio 6.1394%；The frequency of oscillation of oscillation mode -0.3125+i5.1025 is
0.8121Hz, damping ratio 6.1140%.
Table 2 further compared traditional Stochastic subspace identification method, Eigenvalues analysis, Prony, automatic returning sliding average mould
Type and the identification result for predicting the theory of error, and based on the result of calculation of Eigenvalues analysis method, calculate separately each method
The relative error of estimated frequency of oscillation and damping ratio.
As shown in Table 2：The frequency of oscillation and damping ratio error that this method is picked out in control oscillation modes 1 are relatively special
Value indicative analysis method is respectively 1.5649% and 2.2715 (the two while being smaller numerical value)；The institute in control oscillation modes 2
The frequency of oscillation that picks out and damping ratio error relative characteristic value analysis method are respectively that 1.3843% and 2.7285% (the two is same
The numerical value of Shi Wei little), by the relative error of above two control oscillation modes frequency of oscillation and damping ratio, demonstrate this method
Can accurate recognition go out the control oscillation modes of system.
The identification result of 2 distinct methods of table compares
Table 3 further compared this method, Stochastic subspace identification method, Prony and automatic returning moving average model
Efficiency is recognized, as seen from the results in Table 3：The computational efficiency (0.1264s) of this method is substantially better than conventional iterative stochastic subspace side
Method (1.8941s), Prony (0.2078s), automatic returning moving average model (1.8420s) and the prediction theory of error (1.8532s)
Identification efficiency, demonstrate the identification efficiency that electric system control oscillation modes can be improved in this method.
The identification of 3 distinct methods of table takes comparison
To the model of each device in addition to doing specified otherwise, the model of other devices is not limited the embodiment of the present invention,
As long as the device of above-mentioned function can be completed.
It will be appreciated by those skilled in the art that attached drawing is the schematic diagram of a preferred embodiment, the embodiments of the present invention
Serial number is for illustration only, can not represent the quality of embodiment.
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent replacement, improvement and so on should all be included in the protection scope of the present invention.
Claims (6)
1. a kind of control oscillation modes Fast Identification Method based on stochastic subspace, which is characterized in that the method includes with
Lower step：
The measurement information that electric system is obtained from wide area measurement system builds electric system stochastic subspace according to measurement information
Extension Observable matrix；According to the state square of the extension Observable matrix computing system reduced order state equation of stochastic subspace
Battle array；
Calculate the inverse matrix of the generalized inverse matrix of state matrix；
The characteristic value of state matrix and inverse matrix in continuous space is calculated, only retains and is in low-frequency oscillation of electric power system frequency zones
Between characteristic value；Based on the similar quasi- side of mode, the control oscillation modes of system are picked out.
2. a kind of control oscillation modes Fast Identification Method based on stochastic subspace according to claim 1, feature
It is, the inverse matrix of the generalized inverse matrix for calculating state matrix is specially：
1) state matrix A is calculated0Generalized inverse matrix
2) inverse matrix is calculatedInverse matrix A1：
A1=(O↓TO↑)-1O↓TO↓
Wherein, O↑And O↓For submatrix；T is transposition.
3. a kind of control oscillation modes Fast Identification Method based on stochastic subspace according to claim 1, feature
It is, the characteristic value for calculating state matrix and inverse matrix in continuous space only retains and is in low-frequency oscillation of electric power system
The characteristic value of frequency separation is specially：
1) state matrix A is calculated separately0With inverse matrix A1Eigenvalue λ in continuous space0,iAnd λ1,i；
2) λ is deleted respectively0,iAnd λ1,iMiddle characteristic value is real number and the minus complex eigenvalue of imaginary part, then further deletes multiple spy
Characteristic value of the value indicative imaginary part except section [0.4 π, 4 π], only retains λ0,iAnd λ1,iMiddle complex eigenvalue imaginary part section [0.4 π, 4
π] within characteristic value；
3) step 2) will be passed through treated that characteristic value is stored in λ respectively0And λ1, by state matrix A0In the characteristic value of continuous space
The characteristic value remained after step 2) processing is stored in λ0In, by inverse matrix A1Continuous space characteristic value through at step 2)
The characteristic value deposit λ remained after reason1In.
4. a kind of control oscillation modes Fast Identification Method based on stochastic subspace according to claim 3, feature
It is, described to be based on the similar quasi- side of mode, the control oscillation modes for picking out system are specially：
If 1) λ0In complex eigenvalue λ0,iWith λ1In complex eigenvalue λ1,jMeet criterion, then it is assumed that λ0,iAnd λ1,jFor system
Same control oscillation modes；
2) according to from λ0And λ1In the same control oscillation modes λ of system that isolates0,iAnd λ1,i, the leading oscillation mode of computing system
Formula λkAnd corresponding frequency of oscillation fkWith damping ratio ξk。
5. a kind of control oscillation modes Fast Identification Method based on stochastic subspace according to claim 4, feature
It is, the criterion is specially：
In formula, f0,iAnd ξ0,iRespectively λ0In the i-th oscillation mode λ0,iFrequency of oscillation and damping ratio；f1,jAnd ξ1,jRespectively λ1In
Jth oscillation mode λ0,jFrequency of oscillation and damping ratio；Δ f, Δ ξ and Δ m are respectively λ0,iWith λ1,jOscillation frequency deviation, resistance
Buddhist nun is than deviation and oscillation mode deviation；WithRespectively λ0,iWith λ1,jOscillation frequency deviation, damping ratio deviation and oscillation
The threshold value of model deviation.
6. a kind of control oscillation modes Fast Identification Method based on stochastic subspace according to claim 4, feature
It is, the control oscillation modes λ of the systemkAnd corresponding frequency of oscillation fkWith damping ratio ξkSpecially：
In formula, imag (λk) statement take λkImaginary part；real(λk) statement take λkReal part；|λk| statement takes λkModulus value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810686260.7A CN108767879A (en) | 2018-06-28 | 2018-06-28 | A kind of power system oscillation pattern Fast Identification Method based on stochastic subspace |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810686260.7A CN108767879A (en) | 2018-06-28 | 2018-06-28 | A kind of power system oscillation pattern Fast Identification Method based on stochastic subspace |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108767879A true CN108767879A (en) | 2018-11-06 |
Family
ID=63974263
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810686260.7A Pending CN108767879A (en) | 2018-06-28 | 2018-06-28 | A kind of power system oscillation pattern Fast Identification Method based on stochastic subspace |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108767879A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110703013A (en) * | 2019-09-26 | 2020-01-17 | 内蒙古电力(集团)有限责任公司电力调度控制分公司 | Online identification method and device for low-frequency oscillation mode of power system and electronic equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102185324A (en) * | 2011-04-25 | 2011-09-14 | 东北电力大学 | Measured-information-based power system low-frequency oscillation analysis method |
| CN103592530A (en) * | 2013-09-27 | 2014-02-19 | 华北电力大学（保定） | Method for discriminating type of low frequency oscillation mechanism based on envelope fitting |
| CN107749621A (en) * | 2017-10-25 | 2018-03-02 | 东北电力大学 | A kind of power system dynamic stability cooperative identification method |
-
2018
- 2018-06-28 CN CN201810686260.7A patent/CN108767879A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102185324A (en) * | 2011-04-25 | 2011-09-14 | 东北电力大学 | Measured-information-based power system low-frequency oscillation analysis method |
| CN103592530A (en) * | 2013-09-27 | 2014-02-19 | 华北电力大学（保定） | Method for discriminating type of low frequency oscillation mechanism based on envelope fitting |
| CN107749621A (en) * | 2017-10-25 | 2018-03-02 | 东北电力大学 | A kind of power system dynamic stability cooperative identification method |
Non-Patent Citations (1)
| Title |
|---|
| TAO JIANG 等: "Estimating electromechanical oscillation modes from synchrophasor measurements in China Southern Power Grid", 《ELECTRIC POWER SYSTEMS RESEARCH》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110703013A (en) * | 2019-09-26 | 2020-01-17 | 内蒙古电力(集团)有限责任公司电力调度控制分公司 | Online identification method and device for low-frequency oscillation mode of power system and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110035090B (en) | False data injection attack detection method for smart grid | |
| CN108767880A (en) | A kind of iteratively faster Random Subspace Method of electric system control oscillation modes identification | |
| Zhu et al. | Dynamic-memory event-based asynchronous dissipative filtering for TS fuzzy singular semi-Markov jump systems against multi-cyber attacks | |
| Wu et al. | Online identification of bad synchrophasor measurements via spatio-temporal correlations | |
| CN108767879A (en) | A kind of power system oscillation pattern Fast Identification Method based on stochastic subspace | |
| CN113191485B (en) | Power information network security detection system and method based on NARX neural network | |
| CN113742669A (en) | User authentication method based on twin network | |
| CN105373805A (en) | A multi-sensor maneuvering target tracking method based on the principle of maximum entropy | |
| Wang et al. | Spectrum prediction method for ISM bands based on LSTM | |
| CN114285545B (en) | Side channel attack method and system based on convolutional neural network | |
| Kang et al. | Classification method for network security data based on multi-featured extraction | |
| CN114841063A (en) | Aero-engine residual life prediction method based on deep learning | |
| CN110750848A (en) | Method for estimating remaining life of software-hardware degradation system by considering software operation | |
| Gao et al. | Protocol-based stability analysis of stochastic hybrid systems under DoS attacks | |
| ALmutairy et al. | Identification and correction of false data injection attacks against AC state estimation using deep learning | |
| Esnaola et al. | Data-injection attacks | |
| Guo et al. | An improved hybrid ARIMA and support vector machine model for water quality prediction | |
| Fang et al. | A non-intrusive security estimation method based on common attribute of iiot systems | |
| CN112418398A (en) | Safety monitoring method for power information equipment | |
| Zhang et al. | A method for detecting abnormal behavior of ships based on multi-dimensional density distance and an abnormal isolation mechanism | |
| Wang et al. | Quantitative Evaluation of Sensor Fault Diagnosability of F-16 High Maneuvering Fighter | |
| CN105007057A (en) | Uniformly dense clutter sparse method aiming at finite set tracking filter | |
| CN112543428B (en) | Indoor positioning method based on ELM-AdaBoost classification learning | |
| Wu et al. | Adaptive Distributed Graph Model for Multiple-Line Outage Identification in Large-Scale Power System | |
| CN108197750A (en) | A kind of wind power combination forecasting method based on LMD-Elman |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181106 |
|
| WD01 | Invention patent application deemed withdrawn after publication |