CN105259471B - A kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal - Google Patents
A kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal Download PDFInfo
- Publication number
- CN105259471B CN105259471B CN201510661962.6A CN201510661962A CN105259471B CN 105259471 B CN105259471 B CN 105259471B CN 201510661962 A CN201510661962 A CN 201510661962A CN 105259471 B CN105259471 B CN 105259471B
- Authority
- CN
- China
- Prior art keywords
- mrow
- mtd
- msub
- transient
- mtr
- 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.)
- Active
Links
Landscapes
- Locating Faults (AREA)
Abstract
The present invention relates to a kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal, comprise the following steps:(1) after power distribution network breaks down, the zero-sequence current of each branched line failure is recordedAnd faulted phase currentAnd calculate transient zero-sequence currentWith transient fault phase current(2) respectively to transient zero-sequence currentWith transient fault phase currentCarry out mutative scale bistable state processing;(3) each branched line transient characteristic zero-sequence current is calculatedBetween cross correlation matrix number M and integrated correlation coefficient Mn, select MnMinimum circuit ln;(4) to the transient characteristic phase current of each branched lineCarry out 7 layers of db10 WAVELET PACKET DECOMPOSITIONs, and computational short cut energyWith simplified gravity frequency(5) to simplify gravity frequencyFor abscissa, simplify energyFor ordinate, polarity parameter value is that ordinate forms three-dimensional system of coordinate.Compared with prior art, the present invention has the advantages that reliability is high, is not easy by external interference.
Description
Technical field
The present invention relates to a kind of distribution network fault line selection method, and accidental resonance and transient current are based on more particularly, to one kind
The three-dimensional fault-line selecting method of signal.
Background technology
With the development of China's power system, huge power network inevitably causes power failure.Power failure will give state
People's economy brings about great losses.Reliably, lasting power supply is the target of China's power system development.
Important component of the power distribution network as power system, influence to user the closest with contacting for user
It is the most direct.The widely used neutral non-effective grounding mode of China's 3~60kV power distribution networks, also known as small current neutral grounding system,
The failure overwhelming majority of small current neutral grounding system is singlephase earth fault.During generation singlephase earth fault, earth current very little, line
Voltage is still symmetrical, and fault current is weaker, 1~2 hour can be continued to run with case of a fault, however, single-phase
During earth fault, the rising of non-faulting phase-to-ground voltage can make system insulation be on the hazard, and easily cause single-line to ground fault event
The problems such as barrier develops into phase fault, and route protection jump is noisy, power failure.Further, since fault-signal is faint, operating mode situation is answered
The problems such as miscellaneous, is easily caused route selection difficulty.Therefore, there is important and far-reaching meaning in this field Efforts To Develop research work.
At present, many scholars have done substantial amounts of research work to this, it is proposed that many selection methods.Spy based on institute's foundation
The difference of sign amount, at present, fault line selection method for single-phase-to-ground fault can be divided into 3 classes:Signal injection method, steady-state quantity route selection method and transient
Route selection method.Wherein:Signal injection method needs additional signal device, Project Realization complexity;Steady-state quantity route selection method existing characteristics signal
It is faint, the insecure problem of route selection result;The transient characteristic of transient route selection method institute foundation is more several times greater than steady-state value or even tens
Times, and do not influenceed by arc suppression coil, without adding extras, therefore there is higher reliability and application value.Transient state
Amount route selection method it is also proposed a variety of methods, wherein:Existing document constructs criterion to realize using wavelet transformation extraction characteristic information
Failure line selection, but wavelet transformation is easily affected by noise, and selected feature band is likely to be non-effective fault transient amount;In
State patent application CN103135037B utilizes the transient zero-sequence current signal in the T/4 cycles after Prony piecewise fitting failures, not only
The close saturation of current transformer magnetic is effectively avoided to gathering the influence of signal, and improves Prony entirety to a certain extent
Fitting precision, but this algorithm is computationally intensive, and exponent number determination difficulty is fitted, anti-noise ability is not strong;SVMs is solving
There is advantage in small sample, non-linear and high dimensional pattern identification problem, but recognition capability is easily influenceed by inherent parameters;It is Chinese special
Profit application CN103901323A utilizes WAVELET PACKET DECOMPOSITION fault transient zero-sequence current, and feature band is extracted with energy maximum principle,
And be entered into and improve in Duffing oscillators, changed by system phasor and apart from progress route selection, but in small resistance failure
When, route selection accuracy rate is relatively low, route selection poor reliability;In addition, there will be document also has according to the width between fault wire and non-faulting line
Value, waveform and polarity etc. carry out route selection, but the method is particularly easy to by external interference influence of noise, the accuracy rate of route selection
And reliability is not high.
In recent years, the research of Stochastic Resonance Theory achieves larger progress.So-called accidental resonance, it is exactly using random common
Principle of shaking carries out a kind of new technology with actual application value of small-signal enhancing detection, and it is studied and is related to using oneself
Numerous ambits such as physics, signal transacting, mechanical fault diagnosis, biology, chemistry, but the research in power system is relative
It is less.
The content of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide a kind of reliability is high, is not easy
By the three-dimensional fault-line selecting method based on accidental resonance and transient current signal of external interference.
The purpose of the present invention can be achieved through the following technical solutions:One kind is based on accidental resonance and transient current signal
Three-dimensional fault-line selecting method, comprise the following steps:
(1) as power distribution network residual voltage u0(t) it is more than 0.15 times of bus rated voltage UNWhen, while voltage transformer TV
Broken string does not occur and series resonance does not occur for arc suppression coil, records zero-sequence currentAnd faulted phase currentAnd calculate
Transient zero-sequence currentWith transient fault phase currentWherein, fault moment 0.02s,WithTime
Length is 0s~0.04s, and t is sampled point, and n is circuit number, n=1,2,3,4;(2) respectively to the transient zero-sequence in step (1)
Electric currentWith transient fault phase currentMutative scale bistable state processing is carried out, obtains transient characteristic zero-sequence currentWith
Transient characteristic faulted phase current
(3) each branched line transient characteristic zero-sequence current is calculatedBetween cross correlation matrix number M to synthesis it is related
Coefficient Mn, select MnMinimum circuit ln, and make lnPolarity parameters pnFor 1, the p of All other routesnFor -1;
(4) to the transient characteristic phase current of each branched lineCarry out 7 layers of db10 WAVELET PACKET DECOMPOSITIONs, and computational short cut
EnergyWith simplified gravity frequency
(5) to simplify gravity frequencyFor abscissa, simplify energyFor ordinate, polarity parameter value is ordinate structure
Into three-dimensional system of coordinate, with characteristic pointTo represent the transient characteristic l of circuitn, and calculate characteristic point
Squared-distance d between fixing point (0,1,1)n, and judge dnMinimum circuit is faulty line.
(6) d is cut offnMinimum circuit, and whether failure judgement state eliminates, if being still malfunction, for bus
Failure, otherwise EP (end of program).
Described step (1) is specially:
(101) power distribution network residual voltage u is judged0(t) whether it is more than 0.15 times of bus rated voltage UNIf u0(t) >
0.15UN, then step (102), otherwise, return to step (101) are performed;
(102) judge whether voltage transformer TV breaks, if broken string, send voltage transformer TV broken string warnings, otherwise,
Perform step (103);
(103) judge whether arc suppression coil occurs series resonance, if series resonance occurs, regulation arc suppression coil eliminates string
Join resonance, otherwise, it is determined that power distribution network breaks down, while starts sampling apparatus, record each branched line failure and occur upper one week
Zero-sequence current between next cycle occurs to failure for the phaseAnd faulted phase currentAnd calculate transient zero-sequence currentWith transient fault phase current
(104) reduced according to faulted phase voltage, non-faulting phase voltage rise determination failure phase, while start sampling apparatus,
Record each branched line failure and a upper cycle occurred to the zero-sequence current between failure generation next cycleWith failure phase
Electric currentAnd calculate transient zero-sequence currentWith transient fault phase current
Described transient zero-sequence currentWith transient fault phase currentCalculation formula difference it is as follows:
Wherein, Q is the sampling number of each power frequency period.
Described step (2) is specially:
(201) set it is optimized after mutative scale bistable state in potential function parameter a, b and 4 rank Runge-Kutta equations
Numerical computations step-length hs, wherein, a=-0.7926, b=0.0072, hs=CR/fs(fs=105Hz, is sample frequency, CR=
2000 be frequency compression ratio);
Mutative scale bistable state expression formula is:
Dx/dt=-dV (x)/dx+s (t)
In formula:V (x) is potential function, V (x)=- ax2/2+bx4/4;A and b is potential function parameter;S (t) represents transient zero-sequence
Electric currentOr transient fault phase current
(202) transient zero-sequence current is usedThe s (t) in mutative scale bistable state expression formula is replaced, through 4 rank Runge-Kuttas
Equation solution obtains transient characteristic zero-sequence currentWith transient fault phase currentReplace in mutative scale bistable state expression formula
S (t), obtain transient characteristic faulted phase current through 4 rank Runge-Kutta equation solutions
Described step (3) is specially:Calculate each branched line transient characteristic zero-sequence currentBetween cross correlation
Matrix number M, and calculate integrated correlation coefficient Mn, select MnMinimum circuit ln, and make lnPolarity parameters pnFor 1, All other routes
Polarity parameters pnFor -1, cross correlation matrix number M expression formulas are:
In formula:ρnjRepresent circuit ln'sWith circuit lj'sN, j are circuit number, j=1,2,3,4;
Integrated correlation coefficient MnFor:N=1,2,3,4.
Described step (4) is specially:To the transient characteristic faulted phase current of each branched lineCarry out 7 layers of db10
WAVELET PACKET DECOMPOSITION, after first frequency band is rejected, calculate it and simplify energyWith simplified gravity frequencyWherein, energy is simplified
AmountIt is calculated as follows:
In formula, frn(t) it is circuit ln'sThe wavelet package reconstruction coefficient of r-th of frequency range, r frequency bands after WAVELET PACKET DECOMPOSITION
Numbering, r=2 ..., 128;K is reconstruction coefficients sample length;EsnTo be eachThe total energy of each frequency band after WAVELET PACKET DECOMPOSITION
Amount;
Simplify gravity frequencyIt is as follows:
In formula, fgnFor transient state gravity frequency;R is frequency band number.
Operation principle of the present invention is as follows:
1 transient zero-sequence current fault signature
Transient zero-sequence current analysis is carried out using single-phase earthing zero sequence transient magnetic and zero-sequence network, as shown in Figure 2.
Wherein, C0For circuit zero sequence electric capacity;L0For circuit zero sequence equivalent inductance;RgFor the transition resistance of earth point;RpAnd LpRespectively disappear
The equivalent resistance and inductance of arc coil;E (t) is residual voltage.In the moment of compensation grid collapses, can be obtained by Fig. 2 and Fig. 3
Flow through the transient zero-sequence current i of trouble point0.tFor:
In formula:i0L.tFor the inductance electricity in transient zero-sequence current;i0C.tFor the capacitance current component of transient zero-sequence current;
ILmAnd ICmRespectively initial value (the I of inductive current and capacitance currentCm=Uphmω C, ILm=Uphm/ωL);UphmFor the width of phase voltage
Value;ω is power frequency angular frequency;ωfIt is respectively the vibration angular frequency and attenuation coefficient of transient zero-sequence current capacitive component with δ;τLFor
The damping time constant of inductive current;The initial phase of faulty line phase voltage during to be grounded.
From formula (1), when singlephase earth fault occurs for small current neutral grounding system, transient state capacitance current has periodically
Damped oscillation characteristic.In addition, the free oscillation frequency of general overhead transmission line is 300~1500Hz, the inductance of cable run is remote
Less than overhead transmission line, and direct-to-ground capacitance is compared with the big manyfold of the latter, therefore the frequency of oscillation of capacitance current transient process is very high, continues
Time is very short, and its free oscillation frequency is generally 1500~3000Hz.Therefore, the distribution network failure transient state through grounding through arc
The frequency range of zero-sequence current is about 0~3000Hz, and the zero sequence impedance of circuit is much smaller than the capacitive reactance of electric capacity in actual power distribution network,
It can be ignored, zero-sequence network can be reduced to shown in Fig. 3 during singlephase earth fault, and then can obtain the moulds of faulty line n zero
Electric current, voltage relationship are:
And for perfecting the mould electric currents of circuit j zero, voltage relationship is:
In formula:i0nFor by circuit n zero mould electric current;i0nFor by circuit j zero mould electric current;u0For the mould of bus zero electricity
Pressure;Uf0For pressure drop of the trouble point virtual power supply on zero-sequence network;Rf0It is zero mould transition resistance;L is the mould of arc suppression coil zero electricity
Sense;R is arc suppression coil zero mould resistance of series connection;C0nFor zero mould distribution capacity of nth bar circuit;C0jFor zero mould point of j-th strip circuit
Cloth electric capacity.
The transient zero-sequence voltage derivative of faulty line and zero-sequence current opposite polarity it can be seen from formula (2) and formula (3),
The transient zero-sequence voltage derivative for perfecting circuit is identical with zero-sequence current polarity.In addition, it can be seen from existing document, through arc suppression coil
For ground connection distribution network system when phase voltage crosses peak value generation singlephase earth fault, transient zero-sequence current is mainly high frequency capacitance point
Amount, transient characteristic are obvious;When power distribution network crosses zero failure in phase voltage, transient zero-sequence current is mainly inductive component, and transient state is special
Unobvious are levied, but the faulty line attenuating dc component caused by arc suppression coil inductance is larger, and it is straight to perfect Line Attenuation
Flow component is smaller, and it flows only through fault feeder and without flow through feeder line is perfected, therefore, the difference between faulty line and non-fault line
The different difference being greater than between non-fault line and non-fault line.To sum up, faulty line can be utilized and perfect circuit transient state
Zero-sequence current difference and polarity carry out route selection, but the method is easily affected by noise, need to form synthesis with reference to other method
Criterion improves route selection reliability.
2 transient state phase current fault signatures
When singlephase earth fault occurs for small current neutral grounding system, the rise of non-faulting phase voltage is normal voltageTimes, therefore
Barrier phase voltage is changed into 0, and line voltage is constant.Due to there is substantial amounts of energy-storage travelling wave tube in small current neutral grounding system, during singlephase earth fault
Substantial amounts of transient signal will be produced, will mainly be included:Charge and discharge transient signal to line capacitance caused by phase voltage change, disappear
Transient inductance electric current caused by arc coil.Assuming that single-phase earthing, main transient current circulation path such as Fig. 2 occur for a phases of circuit 2
It is shown.
When singlephase earth fault occurs for small current neutral grounding system, the transient state earth current of trouble point includes transient state capacitance current
With transient inductance electric current, existing document proves that transient fault electric current mainly includes 4 frequency components, i.e.,:2 high fdrequency components, 1
Individual fundamental component and 1 DC component.2 high fdrequency components correspond to discharge and recharge transient signal, and DC component corresponds to arc blowout wire
Inductive current caused by circle.
As shown in Figure 2, charge and discharge transient signal has following characteristics:Discharge signal is directly through failure phase line electric capacity and ground
Flow cycle is formed, loop inductance is small, and signal attenuation speed is fast, and frequency of oscillation is higher (kHz);Charging transient signal need by
After transformer, then through non-faulting phase line electric capacity and ground composition loop, whole loop inductance is big, and signal attenuation is slow, frequency of oscillation
Relatively low (hundred Hz), and the transient signal amplitude that charges is better than discharge signal.In addition, a phase faults transient state of circuit 1 mainly includes non-event
Hinder phase line capacitive charge signal (hundred Hz), failure phase line electric capacity discharge signal (kHz), capacitive fundamental component (50Hz) and
Transient inductance component.Because charging signals account for leading, energy focuses primarily upon hundreds of Hz frequency ranges;The a phase faults transient state of circuit 2 containing only
Fundamental component and the line capacitance discharge signal (kHz), energy focus primarily upon kHz frequency ranges.
In summary, when singlephase earth fault occurs for small current neutral grounding system, faulty line and non-fault line it is temporary
There is difference on state faulted phase current signal energy collection midband, still, when fault resstance is too small, if only being concentrated by energy
Frequency band carries out route selection, and route selection nargin may be insufficient, if in addition, only by the difference and polarity between faulty line and non-fault line
Carry out route selection, noise immunity deficiency.Therefore, can be with reference to the polarity of the energy of transient state phase current, gravity frequency and transient zero-sequence current
To form three-dimensional failure line selection criterion, and then improve route selection nargin and route selection accuracy rate.
3 bistable systems
For studying shown in the bistable system such as formula (2) of accidental resonance:
Dx/dt=-dV (x)/dx+s (t)+Γ (t) (4)
In formula:V (x)=- ax2/2+bx4/4;S (t) represents input signal;Γ (t) represents noise.
If it is simplest single-frequency periodic signal Isin (2 π f to make input signal s (t)0T), noise is white noise, that
, so that it may obtain a simplest model that can produce accidental resonance:
Dx/dt=ax-bx3+Isin(2πf0t)+Γ(t) (5)
For example, it is transient zero-sequence current that the present invention, which makes s (t),And formula (4) is carried out using 4 rank Runge-Kutta methods
Solve, solve to obtain transient characteristic zero-sequence currentWherein n is circuit number.
Although Stochastic Resonance Phenomenon and the intuition of people are variant, its general principle or fairly simple.First, may be used
The general principle of accidental resonance is explained by a simple model.
Single Brownian Particles move in the Symmetric Double Well-potential shown in Fig. 3, and when it is not acted on by any external force, particle will
Finally stay in one of potential well, and will be determined positioned at which potential well by initial position.But when random perturbation be present,
Particle has certain probability in the presence of random force and jumped between two potential wells.When particle is only by Periodic Force, such as
The intensity very little of fruit cycle external force, then Brownian Particles will do the vibration of small range in some potential well, without having across potential well
Grand movement.And when cycle external force and noise act on simultaneously, the above situation will change:The gesture of random force induction
Transition and cycle external force between trap occur synchronously, and particle will do grand movement with outside force frequency between two potential wells, weak
Input periodic signal amplified, then just there occurs accidental resonance.
It is noted that work as initial value x0When=0, namely Brownian Particles be in potential hump position such as Fig. 3 in II position,
System is in this unstable stationary state of x=0, any small disturbance (x0≠ 0) Brownian Particles can all be made away from original position
Put.This small interference can come from the interior noise of system, can be from outer noise.But, either interior noise is still
Outer noise, bistable state can all show colourful content under perturbation action, turn into the system being more of practical significance, then
The inherent law of analysis system is more conducive to using the statistical method of correlation.
4 mutative scale bistable systems
The essence of mutative scale thought:It is low frequency by big frequency conversion so that the frequency of big parameter signal is close or meets
Small parameter condition required by accidental resonance, also it is in the frequency band range that can be detected frequency compression to bistable system.Become
The concrete operation process of yardstick is:According to the frequency of signal and sample frequency fsDetermine a frequency compression scale ratio
(Contraction Ratio, CR), a numerical value material calculation h is then re-defined according to CRs=CR/fs, finally utilize quadravalence
Runge-Kutta equation numerical solution mutative scale bistable system, wherein, f in the present inventionsFor 105Hz, CR 2000.
Traditional signal to noise ratio is estimated has a case that more visible spectral line suitable for the frequency spectrum of input/output signal, and temporary
State zero-sequence current is typically all broadband signal, and signal frequency range is not concentrated in one or several denumerable frequencies, but
It is distributed in very wide one section of frequency band.Therefore, traditional signal to noise ratio, which is estimated, will be difficult to be applicable, it is necessary to develop other measurement indexs.
Although nonlinear Langevin equation is unable to the motion of Accurate Prediction Brownian Particles, but can foretell the statistics of racetrack well
Property, then the present invention mutative scale bistable system is described for faint aperiodic input by the use of cross-correlation coefficient as estimating
Response, wherein, transient zero-sequence current izn(t) input signal is used as, then iznAnd i (t)zj(t) the covariance Cov (i betweenzn
(t),izj) and cross-correlation coefficient ρ (t)ijRepresented with equation below:
In formula:N is circuit number.
At present, many accidental resonance researchs are by parameters such as the parameter of regulating system or noise intensities, believe input
Number and noise act synergistically, allow system output to produce resonance widely known in similar mechanics and export.The present invention will be with event
Based on hindering line feature opposite with the zero-sequence current full dose initial period overall variation trend of non-faulting line, input is laid stress on
In the variation tendency of signal and output signal, and estimated using cross-correlation coefficient.Cross-correlation coefficient be reflection two signals it
Between dependency relation level of intimate statistical indicator, provide two signals, as shown in Fig. 6 (a), Fig. 6 (b), calculate between them
Cross-correlation coefficient is -0.8175.Wherein, negative sign shows that two signals belong to negatively correlated, namely variation tendency is on the contrary, be shown in Fig. 6
Mark, 0.8175 shows that two signals belong to strong correlation and belong to strong correlation.
5 Runge-Kutta equations
In order to avoid calculating higher derivative, Runge-Kutta equation utilizes the linear combination of values of the F (x, y) at some points,
A kind of calculation formula is constructed, after making it by Taylor series expansion, compared with the Taylor expansion of the solution of initial-value problem, exists and to the greatest extent may be used
Item more than energy is identical, so as to ensure that formula has higher precision.This method indirect utilization thought of Taylor expansion, keeps away
The difficulty of calculating higher derivative is exempted from.
The form of in general Runge-Kutta equation is:
In formula:αe、μe,iAnd ceIt is undetermined parameter;hsIt is numerical computations step-length.The principle of these parameters is chosen, is requirement
The 1st formula right-hand member is in (x in formula (7)e,ye) place makees Taylor expansion, and presses hsPower arranged type from low to highWith the Taylor expansion of Solutions of Ordinary Differential Equations
There is item as much as possible to overlap, that is, require meeting formula
(6):
ζ1=Fe,ξ2=F 'e,ζ3=F "e... (8)
In formula:Fe,F′e,F″e... represent y ' (xe)=F (xe,ye), y " (xe), y ' " (xe),….Usual wushu (7) is referred to as
N level Runge-Kutta methods, it is abbreviated as N level Runge-Kutta equations.The Runge-Kutta equation of higher order because amount of calculation is larger,
Do not use typically.The present invention uses 4 rank Runge-Kutta equations, shown in its calculating formula such as formula (9):
6 wavelet packets
Analysis method of wavelet packet is the improvement to multiresolution wavelet analysis method, and transient zero-sequence current signal can be carried out
Comprehensive Time-frequency Decomposition and effectively reflect time-frequency characteristics, be more beneficial for the fault signature of extraction transient zero-sequence current signal.It is small
Ripple bag arthmetic statement is as follows:
Define subspaceIt is function uυ(t) Closure Spaces, thusIt is function u2υ(t) Closure Spaces, and
Make uυ(t) Double-scaling equation of formula (10) is met:
In formula (10), G (β)=(- 1)βH (1- β), i.e. H (β) and G (β) also have orthogonality relation.As υ=0, above-mentioned table
It is up to formula:
For υ=Z+Situation, can promote and drawThe sequence constructed
Referred to as by basic functionThe Orthogonal Wavelet Packet of determination.Due toBy HβUniquely determine, so also known as { uυ(t)}
For on sequence { HβOrthogonal Wavelet Packet.In WAVELET PACKET DECOMPOSITION, each layer WAVELET PACKET DECOMPOSITION is exactly one from the point of view of frequency
Kind band logical or low pass filter.A width of [the f of band of each wave filters(β-1)/2τ,fsβ/2τ], wherein τ is the wavelet decomposition number of plies, and β is
The β node of wavelet decomposition, fsFor the frequency of input signal.
In the figure 7, A represents low frequency, and D represents high frequency, and the sequence number number at end represents level (namely the chi of WAVELET PACKET DECOMPOSITION
Degree).Decomposition has following relation:
S=AAA3+DAA3+ADA3+DDA3+AAD3+DAD3+ADD3+DDD3 (12)
Wavelet packet can be adaptive selected individual features frequency band according to the feature of analyzed signal, be allowed to and signal frequency
Spectrum matches, during so as to improve-frequency division resolution, and have a wide range of applications.Daubechies wavelets have orthogonality, tight branch
The advantages that collection and N-1 rank vanishing moments.Therefore, the present invention extracts the feature of transient current signal letter using db10 Orthogonal Wavelet Packets
Number.
Compared with prior art, the present invention has advantages below:
(1) transient characteristic of the application based on transient route selection method institute foundation is more several times greater than steady-state value or even tens times, and
Do not influenceed by arc suppression coil, without adding extras, therefore there is higher reliability and application value;
(2) present invention utilizes the frequency band energy distribution structure of the difference and transient fault phase current between transient zero-sequence current
Into three-dimensional line-selection criterion, different failure situations is adapted to, and has carried out Multi-information acquisition route selection, route selection accuracy rate is higher than only
According to the distribution of the frequency band energy of the difference between transient zero-sequence current or transient fault phase current;
(3) mutative scale bistable system has preferably identification to transient zero-sequence current, can be effectively reduced noise pair
The waveform of transient zero-sequence current and the influence of polarity, there is stronger anti-noise ability, and noiseproof feature is better than EMD algorithms and tradition
Power frequency integer harmonic route selection diagnostic method.In addition, under noise-free case, it is original to its after the processing of mutative scale bistable system
Frequency band energy distribution influence it is little;In the case where there is noise situations, after the processing of mutative scale bistable system, to the frequency band of its original
Energy distribution has certain facilitation.
Brief description of the drawings
Fig. 1 is three-dimensional fault-line selecting method flow chart of the application based on accidental resonance and transient current signal;
Fig. 2 is single-phase earthing zero sequence transient magnetic;
Fig. 3 is singlephase earth fault zero-sequence network simplification figure;
Fig. 4 is singlephase earth fault model;
Fig. 5 is the schematic diagram of Stochastic Resonance Phenomenon in Symmetric Double Well-potential;
Fig. 6 is the oscillogram of different variation tendencies;
Fig. 7 is that WAVELET PACKET DECOMPOSITION tree decomposes frame diagram;
Fig. 8 is ATP simulation model figures;
Fig. 9 is the transient fault phase current figure of each circuit band -0.5db noises;
Figure 10 is the transient characteristic phase current figure of each circuit;
Figure 11 is route selection three-dimensional system of coordinate.
Embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.
A kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal as shown in Figure 1, including following step
Suddenly:
(1) as power distribution network residual voltage u0(t) it is more than 0.15 times of bus rated voltage UNWhen, mutual induction of voltage is detected respectively
Whether device TV occurs broken string and series resonance with arc suppression coil, if voltage transformer TV breaks, sends voltage transformer
TV broken string warnings;Otherwise, judge whether arc suppression coil occurs series resonance, if series resonance occurs for arc suppression coil, regulation disappears
Arc coil eliminates series resonance, otherwise judges that power distribution network breaks down, while starts sampling apparatus, records each branched line failure
A upper cycle occurred to the zero-sequence current between failure generation next cycleAnd faulted phase currentAnd calculate temporary
State zero-sequence currentWith transient fault phase currentWherein, fault moment 0.02s,WithTime length
Spending for 0s~0.04s, t is sampled point, and n is circuit number, n=1,2,3,4;
(2) respectively to the transient zero-sequence current in step (1)With transient fault phase currentIt is double to carry out mutative scale
Steady state process, obtain transient characteristic zero-sequence currentWith transient characteristic faulted phase current
(3) each branched line transient characteristic zero-sequence current is calculatedBetween cross correlation matrix number M to synthesis it is related
Coefficient Mn, select MnMinimum circuit ln, and make lnPolarity parameters pnFor 1, the p of All other routesnFor -1;
(4) to the transient characteristic phase current of each branched lineCarry out 7 layers of db10 WAVELET PACKET DECOMPOSITIONs, and computational short cut
EnergyWith simplified gravity frequency
(5) to simplify gravity frequencyFor abscissa, simplify energyFor ordinate, polarity parameter value is ordinate structure
Into three-dimensional system of coordinate, with characteristic pointTo represent the transient characteristic l of circuitn, and calculate characteristic pointSquared-distance d between fixing point (0,1,1)n, and judge dnMinimum circuit is faulty line.
Described step (1) is specially:
(101) power distribution network residual voltage u is judged0(t) whether it is more than 0.15 times of bus rated voltage UNIf u0(t) >
0.15UN, then step (102), otherwise, return to step (101) are performed;
(102) judge whether voltage transformer TV breaks, if broken string, send voltage transformer TV broken string warnings, otherwise,
Perform step (103);
(103) judge whether arc suppression coil occurs series resonance, if series resonance occurs, regulation arc suppression coil eliminates string
Join resonance, otherwise, it is determined that power distribution network breaks down, while starts sampling apparatus, record each branched line failure and occur upper one week
Zero-sequence current between next cycle occurs to failure for the phaseAnd faulted phase currentAnd calculate transient zero-sequence currentWith transient fault phase current
(104) reduced according to faulted phase voltage, non-faulting phase voltage rise determination failure phase, while start sampling apparatus,
Record each branched line failure and a upper cycle occurred to the zero-sequence current between failure generation next cycleWith failure phase
Electric currentAnd calculate transient zero-sequence currentWith transient fault phase current
Described transient zero-sequence currentWith transient fault phase currentCalculation formula difference it is as follows:
Wherein, Q is the sampling number of each power frequency period.
Described step (2) is specially:
(201) set it is optimized after mutative scale bistable state in potential function parameter a, b and 4 rank Runge-Kutta equations
Numerical computations step-length hs, wherein, a=-0.7926, b=0.0072, hs=CR/fs(fs=105Hz, is sample frequency, CR=
2000 be frequency compression ratio);
Mutative scale bistable state expression formula is:
Dx/dt=-dV (x)/dx+s (t)
In formula:V (x) is potential function, V (x)=- ax2/2+bx4/4;A and b is potential function parameter;S (t) represents transient zero-sequence
Electric currentOr transient fault phase current
(202) transient zero-sequence current is usedThe s (t) in mutative scale bistable state expression formula is replaced, through 4 rank Runge-Kuttas
Equation solution obtains transient characteristic zero-sequence currentWith transient fault phase currentReplace in mutative scale bistable state expression formula
S (t), obtain transient characteristic faulted phase current through 4 rank Runge-Kutta equation solutions
Described step (3) is specially:Calculate each branched line transient characteristic zero-sequence currentBetween cross correlation
Matrix number M, and calculate integrated correlation coefficient Mn, select MnMinimum circuit ln, and make lnPolarity parameters pnFor 1, All other routes
Polarity parameters pnFor -1, cross correlation matrix number M expression formulas are:
In formula:ρnjRepresent circuit ln'sWith circuit lj'sN, j are circuit number, j=1,2,3,4;
Integrated correlation coefficient MnFor:N=1,2,3,4.
Described step (4) is specially:To the transient characteristic faulted phase current of each branched lineCarry out 7 layers of db10
WAVELET PACKET DECOMPOSITION, after first frequency band is rejected, calculate it and simplify energyWith simplified gravity frequencyWherein, energy is simplified
AmountIt is calculated as follows:
In formula, frn(t) it is circuit ln'sThe wavelet package reconstruction coefficient of r-th of frequency range, r frequency bands after WAVELET PACKET DECOMPOSITION
Numbering, r=2 ..., 128;K is reconstruction coefficients sample length;EsnTo be eachThe total energy of each frequency band after WAVELET PACKET DECOMPOSITION
Amount;
Simplify gravity frequencyIt is as follows:
In formula, fgnFor transient state gravity frequency;R is frequency band number.
Operation principle of the present invention is as follows:
1 transient zero-sequence current fault signature
Transient zero-sequence current analysis is carried out using single-phase earthing zero sequence transient magnetic and zero-sequence network, as shown in Figure 2.
Wherein, C0For circuit zero sequence electric capacity;L0For circuit zero sequence equivalent inductance;RgFor the transition resistance of earth point;RpAnd LpRespectively disappear
The equivalent resistance and inductance of arc coil;E (t) is residual voltage.In the moment of compensation grid collapses, can be obtained by Fig. 2 and Fig. 3
Flow through the transient zero-sequence current i of trouble point0.tFor:
In formula:i0L.tFor the inductance electricity in transient zero-sequence current;i0C.tFor the capacitance current component of transient zero-sequence current;
ILmAnd ICmRespectively initial value (the I of inductive current and capacitance currentCm=Uphmω C, ILm=Uphm/ωL);UphmFor the width of phase voltage
Value;ω is power frequency angular frequency;ωfIt is respectively the vibration angular frequency and attenuation coefficient of transient zero-sequence current capacitive component with δ;τLFor
The damping time constant of inductive current;The initial phase of faulty line phase voltage during to be grounded.
From formula (1), when singlephase earth fault occurs for small current neutral grounding system, transient state capacitance current has periodically
Damped oscillation characteristic.In addition, the free oscillation frequency of general overhead transmission line is 300~1500Hz, the inductance of cable run is remote
Less than overhead transmission line, and direct-to-ground capacitance is compared with the big manyfold of the latter, therefore the frequency of oscillation of capacitance current transient process is very high, continues
Time is very short, and its free oscillation frequency is generally 1500~3000Hz.Therefore, the distribution network failure transient state through grounding through arc
The frequency range of zero-sequence current is about 0~3000Hz, and the zero sequence impedance of circuit is much smaller than the capacitive reactance of electric capacity in actual power distribution network,
It can be ignored, zero-sequence network can be reduced to shown in Fig. 3 during singlephase earth fault, and then can obtain the moulds of faulty line n zero
Electric current, voltage relationship are:
And for perfecting the mould electric currents of circuit j zero, voltage relationship is:
In formula:i0nFor by circuit n zero mould electric current;i0nFor by circuit j zero mould electric current;u0For the mould of bus zero electricity
Pressure;Uf0For pressure drop of the trouble point virtual power supply on zero-sequence network;Rf0It is zero mould transition resistance;L is the mould of arc suppression coil zero electricity
Sense;R is arc suppression coil zero mould resistance of series connection;C0nFor zero mould distribution capacity of nth bar circuit;C0jFor zero mould point of j-th strip circuit
Cloth electric capacity.
The transient zero-sequence voltage derivative of faulty line and zero-sequence current opposite polarity it can be seen from formula (2) and formula (3),
The transient zero-sequence voltage derivative for perfecting circuit is identical with zero-sequence current polarity.In addition, it can be seen from existing document, through arc suppression coil
For ground connection distribution network system when phase voltage crosses peak value generation singlephase earth fault, transient zero-sequence current is mainly high frequency capacitance point
Amount, transient characteristic are obvious;When power distribution network crosses zero failure in phase voltage, transient zero-sequence current is mainly inductive component, and transient state is special
Unobvious are levied, but the faulty line attenuating dc component caused by arc suppression coil inductance is larger, and it is straight to perfect Line Attenuation
Flow component is smaller, and it flows only through fault feeder and without flow through feeder line is perfected, therefore, the difference between faulty line and non-fault line
The different difference being greater than between non-fault line and non-fault line.To sum up, faulty line can be utilized and perfect circuit transient state
Zero-sequence current difference and polarity carry out route selection, but the method is easily affected by noise, need to form synthesis with reference to other method
Criterion improves route selection reliability.
2 transient state phase current fault signatures
When singlephase earth fault occurs for small current neutral grounding system, the rise of non-faulting phase voltage is normal voltageTimes, therefore
Barrier phase voltage is changed into 0, and line voltage is constant.Due to there is substantial amounts of energy-storage travelling wave tube in small current neutral grounding system, during singlephase earth fault
Substantial amounts of transient signal will be produced, will mainly be included:Charge and discharge transient signal to line capacitance caused by phase voltage change, disappear
Transient inductance electric current caused by arc coil.Assuming that single-phase earthing, main transient current circulation path such as Fig. 4 occur for a phases of circuit 2
It is shown.
When singlephase earth fault occurs for small current neutral grounding system, the transient state earth current of trouble point includes transient state capacitance current
With transient inductance electric current, existing document proves that transient fault electric current mainly includes 4 frequency components, i.e.,:2 high fdrequency components, 1
Individual fundamental component and 1 DC component.2 high fdrequency components correspond to discharge and recharge transient signal, and DC component corresponds to arc blowout wire
Inductive current caused by circle.
As shown in Figure 4, charge and discharge transient signal has following characteristics:Discharge signal is directly through failure phase line electric capacity and ground
Flow cycle is formed, loop inductance is small, and signal attenuation speed is fast, and frequency of oscillation is higher (kHz);Charging transient signal need by
After transformer, then through non-faulting phase line electric capacity and ground composition loop, whole loop inductance is big, and signal attenuation is slow, frequency of oscillation
Relatively low (hundred Hz), and the transient signal amplitude that charges is better than discharge signal.In addition, a phase faults transient state of circuit 1 mainly includes non-event
Hinder phase line capacitive charge signal (hundred Hz), failure phase line electric capacity discharge signal (kHz), capacitive fundamental component (50Hz) and
Transient inductance component.Because charging signals account for leading, energy focuses primarily upon hundreds of Hz frequency ranges;The a phase faults transient state of circuit 2 containing only
Fundamental component and the line capacitance discharge signal (kHz), energy focus primarily upon kHz frequency ranges.
In summary, when singlephase earth fault occurs for small current neutral grounding system, faulty line and non-fault line it is temporary
There is difference on state faulted phase current signal energy collection midband, still, when fault resstance is too small, if only being concentrated by energy
Frequency band carries out route selection, and route selection nargin may be insufficient, if in addition, only by the difference and polarity between faulty line and non-fault line
Carry out route selection, noise immunity deficiency.Therefore, can be with reference to the polarity of the energy of transient state phase current, gravity frequency and transient zero-sequence current
To form three-dimensional failure line selection criterion, and then improve route selection nargin and route selection accuracy rate.
3 bistable systems
For studying shown in the bistable system such as formula (2) of accidental resonance:
Dx/dt=-dV (x)/dx+s (t)+Γ (t) (4)
In formula:V (x)=- ax2/2+bx4/4;S (t) represents input signal;Γ (t) represents noise.
If it is simplest single-frequency periodic signal Isin (2 π f to make input signal s (t)0T), noise is white noise, that
, so that it may obtain a simplest model that can produce accidental resonance:
Dx/dt=ax-bx3+Isin(2πf0t)+Γ(t) (5)
For example, it is transient zero-sequence current that the present invention, which makes s (t),And formula (4) is carried out using 4 rank Runge-Kutta methods
Solve, solve to obtain transient characteristic zero-sequence currentWherein n is circuit number.
Although Stochastic Resonance Phenomenon and the intuition of people are variant, its general principle or fairly simple.First, may be used
The general principle of accidental resonance is explained by a simple model.
Single Brownian Particles move in the Symmetric Double Well-potential shown in Fig. 5, and when it is not acted on by any external force, particle will
Finally stay in one of potential well, and will be determined positioned at which potential well by initial position.But when random perturbation be present,
Particle has certain probability in the presence of random force and jumped between two potential wells.When particle is only by Periodic Force, such as
The intensity very little of fruit cycle external force, then Brownian Particles will do the vibration of small range in some potential well, without having across potential well
Grand movement.And when cycle external force and noise act on simultaneously, the above situation will change:The gesture of random force induction
Transition and cycle external force between trap occur synchronously, and particle will do grand movement with outside force frequency between two potential wells, weak
Input periodic signal amplified, then just there occurs accidental resonance.
It is noted that work as initial value x0When=0, namely Brownian Particles be in potential hump position such as Fig. 5 in II position,
System is in this unstable stationary state of x=0, any small disturbance (x0≠ 0) Brownian Particles can all be made away from original position
Put.This small interference can come from the interior noise of system, can be from outer noise.But, either interior noise is still
Outer noise, bistable state can all show colourful content under perturbation action, turn into the system being more of practical significance, then
The inherent law of analysis system is more conducive to using the statistical method of correlation.
4 mutative scale bistable systems
The essence of mutative scale thought:It is low frequency by big frequency conversion so that the frequency of big parameter signal is close or meets
Small parameter condition required by accidental resonance, also it is in the frequency band range that can be detected frequency compression to bistable system.Become
The concrete operation process of yardstick is:According to the frequency of signal and sample frequency fsDetermine a frequency compression scale ratio
(Contraction Ratio, CR), a numerical value material calculation h is then re-defined according to CRs=CR/fs, finally utilize quadravalence
Runge-Kutta equation numerical solution mutative scale bistable system, wherein, f in the present inventionsFor 105Hz, CR 2000.
Traditional signal to noise ratio is estimated has a case that more visible spectral line suitable for the frequency spectrum of input/output signal, and temporary
State zero-sequence current is typically all broadband signal, and signal frequency range is not concentrated in one or several denumerable frequencies, but
It is distributed in very wide one section of frequency band.Therefore, traditional signal to noise ratio, which is estimated, will be difficult to be applicable, it is necessary to develop other measurement indexs.
Although nonlinear Langevin equation is unable to the motion of Accurate Prediction Brownian Particles, but can foretell the statistics of racetrack well
Property, then the present invention mutative scale bistable system is described for faint aperiodic input by the use of cross-correlation coefficient as estimating
Response, wherein, transient zero-sequence current izn(t) input signal is used as, then iznAnd i (t)zj(t) the covariance Cov (i betweenzn
(t),izj) and cross-correlation coefficient ρ (t)ijRepresented with equation below:
In formula:N is circuit number.
At present, many accidental resonance researchs are by parameters such as the parameter of regulating system or noise intensities, believe input
Number and noise act synergistically, allow system output to produce resonance widely known in similar mechanics and export.The present invention will be with event
Based on hindering line feature opposite with the zero-sequence current full dose initial period overall variation trend of non-faulting line, input is laid stress on
In the variation tendency of signal and output signal, and estimated using cross-correlation coefficient.Cross-correlation coefficient be reflection two signals it
Between dependency relation level of intimate statistical indicator, provide two signals, as shown in Fig. 6 (a), Fig. 6 (b), calculate between them
Cross-correlation coefficient is -0.8175.Wherein, negative sign shows that two signals belong to negatively correlated, namely variation tendency is on the contrary, be shown in Fig. 6
Mark, 0.8175 shows that two signals belong to strong correlation and belong to strong correlation.
5 Runge-Kutta equations
In order to avoid calculating higher derivative, Runge-Kutta equation utilizes the linear combination of values of the F (x, y) at some points,
A kind of calculation formula is constructed, after making it by Taylor series expansion, compared with the Taylor expansion of the solution of initial-value problem, exists and to the greatest extent may be used
Item more than energy is identical, so as to ensure that formula has higher precision.This method indirect utilization thought of Taylor expansion, keeps away
The difficulty of calculating higher derivative is exempted from.
The form of in general Runge-Kutta equation is:
In formula:αe、μe,iAnd ceIt is undetermined parameter;hsIt is numerical computations step-length.The principle of these parameters is chosen, is requirement
The 1st formula right-hand member is in (x in formula (7)e,ye) place makees Taylor expansion, and presses hsPower arranged type from low to highWith the Taylor expansion of Solutions of Ordinary Differential Equations
There is item as much as possible to overlap, that is, require meeting formula (6):
ζ1=Fe,ξ2=F 'e,ζ3=F "e... (8)
In formula:Fe,F′e,F″e... represent y ' (xe)=F (xe,ye), y " (xe), y ' " (xe),….Usual wushu (7) is referred to as
N level Runge-Kutta methods, it is abbreviated as N level Runge-Kutta equations.The Runge-Kutta equation of higher order because amount of calculation is larger,
Do not use typically.The present invention uses 4 rank Runge-Kutta equations, shown in its calculating formula such as formula (9):
6 wavelet packets
Analysis method of wavelet packet is the improvement to multiresolution wavelet analysis method, and transient zero-sequence current signal can be carried out
Comprehensive Time-frequency Decomposition and effectively reflect time-frequency characteristics, be more beneficial for the fault signature of extraction transient zero-sequence current signal.It is small
Ripple bag arthmetic statement is as follows:
Define subspaceIt is function uυ(t) Closure Spaces, thusIt is function u2υ(t) Closure Spaces, and
Make uυ(t) Double-scaling equation of formula (10) is met:
In formula (10), G (β)=(- 1)βH (1- β), i.e. H (β) and G (β) also have orthogonality relation.As υ=0, above-mentioned table
It is up to formula:
For υ=Z+Situation, can promote and drawThe sequence constructed
Referred to as by basic functionThe Orthogonal Wavelet Packet of determination.Due toBy HβUniquely determine, so also known as { uυ(t)}
For on sequence { HβOrthogonal Wavelet Packet.In WAVELET PACKET DECOMPOSITION, each layer WAVELET PACKET DECOMPOSITION is exactly one from the point of view of frequency
Kind band logical or low pass filter.A width of [the f of band of each wave filters(β-1)/2τ,fsβ/2τ], wherein τ is the wavelet decomposition number of plies, and β is
The β node of wavelet decomposition, fsFor the frequency of input signal.It is as shown in Figure 7 that WAVELET PACKET DECOMPOSITION tree decomposes framework.
In the figure 7, A represents low frequency, and D represents high frequency, and the sequence number number at end represents level (namely the chi of WAVELET PACKET DECOMPOSITION
Degree).Decomposition has following relation:
S=AAA3+DAA3+ADA3+DDA3+AAD3+DAD3+ADD3+DDD3 (12)
Wavelet packet can be adaptive selected individual features frequency band according to the feature of analyzed signal, be allowed to and signal frequency
Spectrum matches, during so as to improve-frequency division resolution, and have a wide range of applications.Daubechies wavelets have orthogonality, tight branch
The advantages that collection and N-1 rank vanishing moments.Therefore, the present invention extracts the feature of transient current signal letter using db10 Orthogonal Wavelet Packets
Number.
According to above-mentioned theory, the present invention does single-phase earthing emulation experiment using ATP-EMTP, simulation model as shown in figure 8,
The specific electric parameter of model is as follows:
Circuit:The positive order parameter R of overhead transmission line1=0.17 Ω/km, L1=1.2mH/km, C1=9.697nF/km;Zero sequence is joined
Number R0=0.23 Ω/km, L0=5.48mH/km, C0=6nF/km.The positive order parameter R of cable run11=0.193 Ω/km, L11=
0.442mH/km, C11=143nF/km;Zero sequence parameter R00=1.93 Ω/km, L00=5.48mH/km, C00=143nF/km.
Transformer:110/10.5kV;The single-phase neutral point coil resistance 0.40 in high-pressure side, inductance 12.2;Low-pressure side phase line
Enclose resistance 0.006, inductance 0.183;Exciting current 0.672A, magnetic flux 202.2Wb, magnetic circuit resistance 400k Ω.Load:Without exception
Using delta connection, ZL=400+j20 Ω.Arc suppression coil:When arc suppression coil earthing system emulates, arc suppression coil inductance is
LN=1281.9mH, RN=40.5217 Ω.
2 change failure initial phase angle and resistance simulation result and analysis
Simulation model is built by parameter, and l is set1In the failure at bus 5km, change fault initial angle (0 °, 30 °,
60 °, 90 °) and grounding resistance emulated.Then, selected transient zero-sequence current and transient fault phase current are pressed
Route selection is carried out according to selection method, wherein, the parameter of mutative scale bistable system is:A=b=1, CR=1500.It is in addition, of the invention
With (l1, 0 °, 300 Ω) and this form represents circuit l1Initial phase angle is 0 °, failure situation when fault resstance is 300 Ω.As a result
As shown in table 1, wherein representing l1Transient characteristic zero-sequence current and l2The cross-correlation coefficient of transient characteristic zero-sequence current.
1 different initial phase angles of table and resistance simulation result
(a) failure initial phase angle is 0 °
(b) failure initial phase angle is 30 °
(c) failure initial phase angle is 60 °
(d) failure initial phase angle is 90 °
As shown in Table 1, when fault resstance and failure initial phase angle all change, selection method of the present invention can be accurate
Select faulty line.
Simulation result and analysis in the case of 3 Noises
In view of when real system breaks down, adopt signal often with noise, therefore, for the new side of checking this paper route selections
The noise immunity of method, in the case of different faults, to its transient zero-sequence current and transient fault phase current addition -0.5db noise,
Route selection is carried out according to selection method of the present invention, route selection result is as shown in table 2.
The route selection result of different faults situation under table 2-0.5db noise intensities
Choose under the noise background that signal to noise ratio is -0.5db, failure situation (l3,90 °, 1500 Ω) illustrate, provide
Transient characteristic phase current under strong noise background, as shown in Figure 9;Transient characteristic phase current figure after the processing of mutative scale bistable state
Shown in 10, characteristic point is as shown in Figure 10 in the position of route selection three-dimensional system of coordinate.
It was found from Figure 10 and Figure 11, after the processing of mutative scale bistable state, the transient fault phase current under strong noise background
Noise substantially weaken.As can be seen from Figure 11, faulty line l3Energy be concentrated mainly on low-frequency range, energy is big, polarity with it is non-therefore
Hinder circuit on the contrary, and the squared-distance d corresponding to it3Minimum, only 0.0537;The frequency range that the energy of non-fault line is concentrated
Higher than faulty line, energy is small, and the squared-distance corresponding to them is all bigger, typically all 4.2 or so, faulty line with it is non-
This species diversity of faulty line squared-distance is also consistent with theory analysis.In addition.Shown by table 2 and a large amount of simulation results, this hair
Different faults situation of the bright described selection method under strong noise background can realize accurate route selection, reliability height.
In summary, presently preferred embodiments of the present invention is these are only, is not intended to limit the scope of the present invention.
Within the spirit and principles of the invention, any modification, equivalent substitution and improvements made etc., it should be included in the present invention's
Within protection domain.
Claims (6)
1. a kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal, it is characterised in that this method includes
Following steps:
(1) as power distribution network residual voltage u0(t) it is more than 0.15 times of bus rated voltage UNWhen, while voltage transformer TV is not sent out
Series resonance, record zero-sequence current do not occur for raw broken string and arc suppression coilAnd faulted phase currentAnd calculate transient state
Zero-sequence currentWith transient fault phase currentWherein, fault moment 0.02s,WithTime span
For 0s~0.04s, t is sampled point, and n is circuit number, n=1,2,3,4;
(2) respectively to the transient zero-sequence current in step (1)With transient fault phase currentCarry out mutative scale bistable state
Processing, obtains transient characteristic zero-sequence currentWith transient characteristic faulted phase current
(3) each branched line transient characteristic zero-sequence current is calculatedBetween cross correlation matrix number M and integrated correlation coefficient
Mn, select MnMinimum circuit ln, and make lnPolarity parameters pnFor 1, the p of All other routesnFor -1;
(4) to the transient characteristic phase current of each branched lineCarry out 7 layers of db10 WAVELET PACKET DECOMPOSITIONs, and computational short cut energyWith simplified gravity frequency
(5) to simplify gravity frequencyFor abscissa, simplify energyFor ordinate, polarity parameter value is that ordinate forms three
Dimension coordinate system, with characteristic pointTo represent the transient characteristic l of circuitn, and calculate characteristic point
Squared-distance d between fixing point (0,1,1)n, and judge dnMinimum circuit is faulty line.
2. a kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal according to claim 1, its
It is characterised by, described step (1) is specially:
(101) power distribution network residual voltage u is judged0(t) whether it is more than 0.15 times of bus rated voltage UNIf u0(t) > 0.15UN,
Then perform step (102), otherwise, return to step (101);
(102) judge whether voltage transformer TV breaks, if broken string, send voltage transformer TV broken string warnings, otherwise, perform
Step (103);
(103) judge whether arc suppression coil occurs series resonance, if series resonance occurs, it is humorous that regulation arc suppression coil eliminates series connection
Shake, otherwise, it is determined that power distribution network breaks down, while starts sampling apparatus, record each branched line failure and occurred a upper cycle and arrive
Zero-sequence current between next cycle occurs for failureAnd faulted phase currentAnd calculate transient zero-sequence currentWith transient fault phase current
(104) reduced according to faulted phase voltage, the rise of non-faulting phase voltage determines failure phase, while starts sampling apparatus, record
Occurred for each branched line failure to the zero-sequence current between failure generation next cycle a upper cycleAnd faulted phase currentAnd calculate transient zero-sequence currentWith transient fault phase current
3. a kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal according to claim 2, its
It is characterised by, described transient zero-sequence currentWith transient fault phase currentCalculation formula difference it is as follows:
<mrow>
<msub>
<mi>i</mi>
<msub>
<mi>z</mi>
<mi>n</mi>
</msub>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msub>
<mi>i</mi>
<msub>
<mi>Z</mi>
<mi>n</mi>
</msub>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<msub>
<mi>i</mi>
<msub>
<mi>Z</mi>
<mi>n</mi>
</msub>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>-</mo>
<mi>Q</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>i</mi>
<msub>
<mi>&chi;</mi>
<mi>n</mi>
</msub>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msub>
<mi>i</mi>
<msub>
<mi>X</mi>
<mi>n</mi>
</msub>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<msub>
<mi>i</mi>
<msub>
<mi>X</mi>
<mi>n</mi>
</msub>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>-</mo>
<mi>Q</mi>
<mo>)</mo>
</mrow>
</mrow>
1
Wherein, Q is the sampling number of each power frequency period.
4. a kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal according to claim 1, its
It is characterised by, described step (2) is specially:
(201) parameter a, b of potential function and the numerical computations step-length h of 4 rank Runge-Kutta equations in mutative scale bistable state are sets,
Wherein, a=-0.7926, b=0.0072, hs=CR/fs, wherein fs=105Hz, is sample frequency, and CR=2000 is frequency pressure
Contracting ratio;
Mutative scale bistable state expression formula is:
Dx/dt=-dV (x)/dx+s (t)
In formula:V (x) is potential function, V (x)=- ax2/2+bx4/4;A and b is potential function parameter;S (t) represents transient zero-sequence currentOr transient fault phase current
(202) transient zero-sequence current is usedThe s (t) in mutative scale bistable state expression formula is replaced, through 4 rank Runge-Kutta equations
Solve to obtain transient characteristic zero-sequence currentWith transient fault phase currentReplace the s in mutative scale bistable state expression formula
(t), transient characteristic faulted phase current is obtained through 4 rank Runge-Kutta equation solutions
5. a kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal according to claim 1, its
It is characterised by, described step (3) is specially:Calculate each branched line transient characteristic zero-sequence currentBetween cross-correlation
Coefficient matrix M, and calculate integrated correlation coefficient Mn, select MnMinimum circuit ln, and make lnPolarity parameters pnFor 1, other lines
The polarity parameters p on roadnFor -1, cross correlation matrix number M expression formulas are:
<mrow>
<mi>M</mi>
<mo>=</mo>
<mfenced open = "[" close = "]">
<mtable>
<mtr>
<mtd>
<msub>
<mi>&rho;</mi>
<mn>11</mn>
</msub>
</mtd>
<mtd>
<msub>
<mi>&rho;</mi>
<mn>12</mn>
</msub>
</mtd>
<mtd>
<mo>...</mo>
</mtd>
<mtd>
<msub>
<mi>&rho;</mi>
<mrow>
<mn>1</mn>
<mi>n</mi>
</mrow>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>&rho;</mi>
<mn>21</mn>
</msub>
</mtd>
<mtd>
<mn>1</mn>
</mtd>
<mtd>
<mo>...</mo>
</mtd>
<mtd>
<msub>
<mi>&rho;</mi>
<mrow>
<mn>2</mn>
<mi>n</mi>
</mrow>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<mtable>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
</mtable>
</mtd>
<mtd>
<mtable>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
</mtable>
</mtd>
<mtd>
<mtable>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
</mtable>
</mtd>
<mtd>
<mtable>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
<mtr>
<mtd>
<mo>.</mo>
</mtd>
</mtr>
</mtable>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mi>&rho;</mi>
<mrow>
<mi>n</mi>
<mn>1</mn>
</mrow>
</msub>
</mtd>
<mtd>
<msub>
<mi>&rho;</mi>
<mrow>
<mi>n</mi>
<mn>2</mn>
</mrow>
</msub>
</mtd>
<mtd>
<mo>...</mo>
</mtd>
<mtd>
<msub>
<mi>&rho;</mi>
<mrow>
<mi>n</mi>
<mi>n</mi>
</mrow>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
In formula:ρnjRepresent circuit ln'sWith circuit lj'sCross-correlation coefficient;N, j are circuit number, j=1,2,
3,4;
Integrated correlation coefficient MnFor:
6. a kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal according to claim 1, its
It is characterised by, described step (4) is specially:To the transient characteristic faulted phase current of each branched lineCarry out 7 layers
Db10 WAVELET PACKET DECOMPOSITIONs, after first frequency band is rejected, calculate it and simplify energyWith simplified gravity frequencyWherein, it is simple
Change energyIt is calculated as follows:
<mrow>
<msub>
<mi>E</mi>
<mrow>
<mi>s</mi>
<mi>n</mi>
</mrow>
</msub>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>r</mi>
<mo>=</mo>
<mn>2</mn>
</mrow>
<mn>90</mn>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>k</mi>
</munderover>
<msubsup>
<mi>f</mi>
<mrow>
<mi>r</mi>
<mi>n</mi>
</mrow>
<mn>2</mn>
</msubsup>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msubsup>
<mi>E</mi>
<mrow>
<mi>s</mi>
<mi>n</mi>
</mrow>
<mo>*</mo>
</msubsup>
<mo>=</mo>
<msub>
<mi>E</mi>
<mrow>
<mi>s</mi>
<mi>n</mi>
</mrow>
</msub>
<mo>/</mo>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>E</mi>
<mrow>
<mi>s</mi>
<mi>n</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
</mrow>
In formula, frn(t) it is circuit ln'sThe wavelet package reconstruction coefficient of r-th of frequency range after WAVELET PACKET DECOMPOSITION, r frequency band numbers,
R=2 ..., 128;K is reconstruction coefficients sample length;EsnTo be eachThe gross energy of each frequency band after WAVELET PACKET DECOMPOSITION;
Simplify gravity frequencyIt is as follows:
<mrow>
<msub>
<mi>f</mi>
<mrow>
<mi>g</mi>
<mi>n</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>r</mi>
<mo>=</mo>
<mn>2</mn>
</mrow>
<mn>90</mn>
</munderover>
<mrow>
<mo>(</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>k</mi>
</munderover>
<msubsup>
<mi>f</mi>
<mrow>
<mi>r</mi>
<mi>n</mi>
</mrow>
<mn>2</mn>
</msubsup>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
<mo>)</mo>
</mrow>
<mo>*</mo>
<mi>r</mi>
</mrow>
<msub>
<mi>E</mi>
<mrow>
<mi>s</mi>
<mi>n</mi>
</mrow>
</msub>
</mfrac>
<mo>*</mo>
<mn>39</mn>
<mi>H</mi>
<mi>z</mi>
</mrow>
<mrow>
<msubsup>
<mi>f</mi>
<mrow>
<mi>g</mi>
<mi>n</mi>
</mrow>
<mo>*</mo>
</msubsup>
<mo>=</mo>
<msub>
<mi>f</mi>
<mrow>
<mi>g</mi>
<mi>n</mi>
</mrow>
</msub>
<mo>/</mo>
<mn>1000</mn>
</mrow>
In formula, fgnFor transient state gravity frequency;R is frequency band number.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510661962.6A CN105259471B (en) | 2015-10-14 | 2015-10-14 | A kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510661962.6A CN105259471B (en) | 2015-10-14 | 2015-10-14 | A kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105259471A CN105259471A (en) | 2016-01-20 |
CN105259471B true CN105259471B (en) | 2017-12-08 |
Family
ID=55099236
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510661962.6A Active CN105259471B (en) | 2015-10-14 | 2015-10-14 | A kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105259471B (en) |
Families Citing this family (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105738764B (en) * | 2016-02-22 | 2018-09-07 | 国家电网公司 | Fault Section Location of Distribution Network based on transient information Whole frequency band |
CN105974264B (en) * | 2016-04-28 | 2019-10-11 | 长沙理工大学 | Fault line selection method based on phase current transient characteristics |
CN106501683A (en) * | 2016-12-22 | 2017-03-15 | 南京因泰莱电器股份有限公司 | One kind is without main website power distribution network one-way earth fault localization method |
CN107132444A (en) * | 2017-03-13 | 2017-09-05 | 国网山东省电力公司淄博供电公司 | A kind of small current grounding failure wire selection system |
CN107167702A (en) * | 2017-05-04 | 2017-09-15 | 国网福建省电力有限公司 | A kind of distribution feeder fault type recognition method and device |
CN109507531B (en) * | 2018-11-21 | 2021-01-22 | 杭州电力设备制造有限公司 | Power distribution network single-phase grounding line selection method, system, device and readable storage medium |
CN109782122A (en) * | 2018-12-17 | 2019-05-21 | 济南大学 | A kind of detection method of S Injection Signal |
CN109752630B (en) * | 2019-03-06 | 2021-05-25 | 云南电网有限责任公司保山供电局 | Fault line selection method for double-circuit direct current lines on same tower |
CN110535151B (en) * | 2019-09-21 | 2022-11-01 | 沈阳工程学院 | Method for improving transient stability of tail-end power grid containing high-proportion wind power |
CN110954743B (en) * | 2019-12-18 | 2020-09-25 | 山东山大电力技术股份有限公司 | Distributed wave recording device and low-current grounding line selection method |
CN112083280B (en) * | 2020-08-27 | 2022-07-08 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | Method for identifying fault interval of hybrid multi-terminal direct-current power transmission system |
CN112505474B (en) * | 2020-10-30 | 2022-09-23 | 国网山东省电力公司聊城供电公司 | Fault positioning method and device based on phase current increment |
CN116457673A (en) * | 2020-11-09 | 2023-07-18 | 麦格纳国际公司 | Determination and classification of insulation degradation of motor windings |
CN113049913A (en) * | 2021-03-15 | 2021-06-29 | 广东电网有限责任公司梅州供电局 | Distribution line online monitoring method, device and system |
CN117471366B (en) * | 2023-12-27 | 2024-04-09 | 国网福建省电力有限公司 | Metering device neutral line contact failure studying and judging method based on least square method |
CN118378167A (en) * | 2024-06-24 | 2024-07-23 | 湖南科技大学 | Transmission chain fault diagnosis method based on zero sequence current correlation entropy reinforcement sparse learning |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103901323A (en) * | 2014-04-01 | 2014-07-02 | 河南理工大学 | Fault line selection method with improved oscillator system |
WO2014154164A1 (en) * | 2013-03-29 | 2014-10-02 | 北京映翰通网络技术股份有限公司 | Method and system for detecting and locating single-phase ground fault on low current grounded power-distribution network |
CN104950230A (en) * | 2015-07-06 | 2015-09-30 | 河南理工大学 | Power distribution network fault line selection method based on variable-scale bi-stable system |
-
2015
- 2015-10-14 CN CN201510661962.6A patent/CN105259471B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2014154164A1 (en) * | 2013-03-29 | 2014-10-02 | 北京映翰通网络技术股份有限公司 | Method and system for detecting and locating single-phase ground fault on low current grounded power-distribution network |
CN103901323A (en) * | 2014-04-01 | 2014-07-02 | 河南理工大学 | Fault line selection method with improved oscillator system |
CN104950230A (en) * | 2015-07-06 | 2015-09-30 | 河南理工大学 | Power distribution network fault line selection method based on variable-scale bi-stable system |
Non-Patent Citations (1)
Title |
---|
基于稳态零序电流的故障选线方法的研究;刘扬 等;《湖州师范学院学报》;20111130;第33卷;第4-7页 * |
Also Published As
Publication number | Publication date |
---|---|
CN105259471A (en) | 2016-01-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105259471B (en) | A kind of three-dimensional fault-line selecting method based on accidental resonance and transient current signal | |
CN104950230B (en) | A kind of distribution network fault line selection method based on mutative scale bistable system | |
CN103529332B (en) | A kind of extra high voltage direct current transmission line based on voltage degree of association and wavelet transformation transient state energy distribution character thunderbolt interference identification method | |
US10128650B2 (en) | Relay protection method and apparatus against LC parallel circuit detuning faults | |
CN101545943B (en) | Method for fault line selection of cable-wire mixed line of electric distribution network by using wavelet energy relative entropy | |
CN101546906B (en) | Method for fault line selection of electric distribution network by using S transformation energy relative entropy | |
CN102253283B (en) | A kind of distributed micro-grid grid-connected island detection method based on Wavelet Packet Energy Spectrum | |
Biswas et al. | A dual-time transform assisted intelligent relaying scheme for the STATCOM-compensated transmission line connecting wind farm | |
CN103197124B (en) | Based on the overvoltage identification method of time-frequency matrix singular value | |
CN101404408A (en) | Distribution network cable-wire mixed line failure route selection method by utilizing relative energy | |
Song et al. | Natural frequency based protection and fault location for VSC-HVDC transmission lines | |
CN102005740A (en) | Extra-high voltage direct current line boundary element method adopting polar wave wavelet energy ratio | |
CN108896869A (en) | A kind of HVDC transmission line lightning stroke interference identification method based on S-transformation | |
Moravej et al. | A new approach for fault classification and section detection in compensated transmission line with TCSC | |
CN106324406A (en) | Transformer direct-current magnetic bias fault diagnosis method and device | |
CN110007198A (en) | A kind of novel singlephase earth fault starting method | |
Ma et al. | Voltage transient disturbance detection based on the rms values of segmented differential waveforms | |
CN106546882A (en) | A kind of method of detection power transformer internal discharge failure | |
CN108647469A (en) | A kind of mathematic(al) representation approximating method of TEV mass datas | |
Shu et al. | A detection method of high impedance arcing fault for distribution network with distributed generation based on CEEMDAN and TEO algorithm | |
CN108092234A (en) | Suitable for the differential protecting method and device of adjustable frequency motor | |
CN105044557A (en) | Fault feature extraction method of weak transient zero-sequence current on background of strong noise | |
Gao et al. | Internal overvoltage identification of distribution network via time-frequency atomic decomposition | |
Abdullah | Towards a new paradigm for ultrafast transmission line relaying | |
Jiandong et al. | Non-unit Boundary Protection using Synchro-squeezing Wavelet for HVDC lines |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |