CN101257208A - Method for identifying transformer excitation surge current - Google Patents

Abstract

The invention provides an identification method for transformer inrush current which belongs to power system main equipment relay protection technology field. The method has characteristics in that the method identifies inrush current and fault current base on waveform correlation degree. By comparison with each sampling point algebraic sum in dynamic differential current short data window, non-saturation region of transformer under inrush current and fault current is searched, two form standard sine waves are constituted of max value and position of sampling data in non-saturation region, then differential current sampling waveform in non-saturation region and two form standard sine waves correlation degrees are calculated respectively, inrush current and fault current are differentiated according with correlation coefficient average. The method can open longitudinal differential protection rapidly when internal fault and no-load closing with fault occur in transformer operation; and can lock longitudinal differential protection reliably when on-load closing occurs inrush current.

Description

A kind of discrimination method of transformer excitation flow

Technical field

The invention belongs to the power system main equipment technical field of relay protection, relate in particular to a kind of discrimination method of transformer excitation flow.

Background technology

Although there is certain defective in transformer differential protection on principle, performance factor is lower, and its application time is remote, in extensive range, and the main protection as transformer at present also is in irreplaceable status.The serious problems that transformer differential protection faces are magnetizing inrush currents; because transformer iron core when idle-loaded switching-on or the voltage recovery of external fault excision back is saturated; very big magnetizing inrush current can appear in transient process; magnetizing inrush current is as differential current; if there are not the corresponding precautionary measures, longitudinal difference protection is easy to malfunction.Round how preventing that magnetizing inrush current from causing the problem of malfunction, proposed the discrimination method of multiple magnetizing inrush current.

The on-the-spot at present main secondary harmonic brake principle that adopts prevents that magnetizing inrush current from causing the longitudinal difference protection malfunction, but along with the raising of transformer manufacturing technology and the improvement of manufactured materials, the saturation magnetic induction of modern transformer is low, and the second harmonic composition significantly reduces when saturated.Under such condition, when giving the transformer charging of low saturation magnetic induction, longitudinal difference protection may malfunction.If but reduce the second harmonic ratio; because the high-power transformer electric pressure is high and the end of being everlasting connects long power transmission line; the capacity effect of power transmission line is fairly obvious; the large value capacitor of static reactive is extensive use of in addition; when the inner generation of high-power transformer catastrophe failure; resonance between inductance and the electric capacity can make the harmonic content in the short circuit current obviously increase, and might cause longitudinal difference protection deferred action.In addition, because the secondary harmonic brake method is difficult for adopting the phase-splitting closedown mode, cause when the band minor failure closes a floodgate, the longitudinal difference protection action is very slow.Though there are not the problems referred to above in the magnetizing inrush current discrimination method of another interval angle principle, when but current transformer is saturated, because the progress of disease produces reverse current to the magnetizing inrush current of instrument transformer secondary side, the waveform distortion, cause the interval angle that shoves to disappear, recover original interval angle not a duck soup exactly.Secondary harmonic brake principle and interval angle principle have just been utilized between magnetizing inrush current and the fault current difference characteristic in a certain respect, thereby have limitation.

For solving transformer excitation flow discrimination method accuracy and the not high problem of reliability, the present invention identifies the unsaturation working region of transformer by the algebraical sum of short data window interpolation stream sampled point, further construct the standard sine wave of equal length, analyze the degree of correlation of the standard sine wave of interior actual samples waveform of non-saturated region and structure, differentiate transformer excitation flow.

Summary of the invention

The object of the present invention is to provide a kind of discrimination method of transformer excitation flow.

A kind of discrimination method of transformer excitation flow is characterised in that specifically this method is that computer is realized successively according to the following steps:

(1) comprises determining of unsaturation zone crude sampling ordered series of numbers

1. the selection of observation data window.The phase sampling number is n weekly, the transformer spill current i (k) that to get two cycle length be 2n is as the observation data window, and the length of getting the short data window is n/2-1 (odd number), when for example the phase is 24 point samplings weekly, short data window length is 11, and the length of short data window is odd number.

2. judge the biased direction of difference stream.Seek the maximum i of observation data window interpolation stream _MaxWith minimum value i _MinIf, i _Max-i _Min〉=0, be forward bias difference stream; If i _Max-i _Min＜0, be negative sense bias difference stream.

3. seek the difference stream crude sampling ordered series of numbers that comprises non-saturated region.Starting point by the observation data window begins, and moves a sampled point successively backward, the difference of calculating each sampled point in the short data window flow algebraical sum S (k), k=1 ~ n/2+1, the relatively size of n/2+1 the interior sampled point algebraical sum of short data window.If, get the short data window of corresponding algebraical sum minimum for forward bias difference stream; If for negative sense bias difference stream, get the short data window of corresponding algebraical sum maximum, this short data window is formed crude sampling ordered series of numbers X.

(2) building method of standard sine wave

Two kinds of standard sine waves of original difference stream structure according to selected short data window.

1. a kind of is the peak value i that flows with original difference in the short data window _pAs the peak value of structure standard sine wave, be the center with this peak point again, about it, serve as to construct at interval n/4-1 point with 2 π/n, put formation standard sine wave ordered series of numbers Y for this n/2-1 ₁, Y _1k=i _pSin (pi/2-2 π k/n), k=1 ~ (n-1)/2.

2. another kind also is the peak value i with original difference stream in the short data window _pAs the peak value of structure standard sine wave, but the position of this peak value is identical with the crude sampling ordered series of numbers, and establishing the position is m, again according to Y _2k=i _pSin[pi/2-2 π (k-m)/n], k=1 ~ (n-1)/2, structure standard sine wave ordered series of numbers Y ₂

(3) recognizer of magnetizing inrush current

1. the original waveform ordered series of numbers calculates with the coefficient correlation of the standard sine wave ordered series of numbers that constructs

On the basis of original waveform ordered series of numbers X, construct standard sine wave ordered series of numbers Y ₁, Y ₂After, calculate X and Y by following formula respectively ₁, X and Y ₂Correlation coefficient r _XY1And r _XY2

${\mathrm{<figure-callout\; id="1"\; label="r\; XY"\; filenames="A20071030376000101.png,A20071030376000111.png"\; state="\{\{state\}\}">r</figure-callout>}}_{\mathrm{XY}}=\frac{\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}(X_k-\stackrel{\&OverBar;}{X})(Y_{1k}-{\stackrel{\&OverBar;}{Y}}_1)}{\sqrt{\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}{(X_k-\stackrel{\&OverBar;}{X})}^2\&CenterDot;\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}{(Y_{1k}-{\stackrel{\&OverBar;}{Y}}_1)}^2}}---\left(1\right)$

$r_{\mathrm{XY}2}=\frac{\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}(X_k-\stackrel{\&OverBar;}{X})(Y_{2k}-{\stackrel{\&OverBar;}{Y}}_2)}{\sqrt{\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}{(X_k-\stackrel{\&OverBar;}{X})}^2\&CenterDot;\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}{(Y_{2k}-{\stackrel{\&OverBar;}{Y}}_2)}^2}}---\left(2\right)$

Wherein, X, Y ₁, Y ₂Be respectively ordered series of numbers X, Y ₁, Y ₂Mean value.Coefficient correlation has reflected the similarity degree of two waveforms.Two waveforms are similar more, and its coefficient correlation is big more; Otherwise coefficient correlation is more little.

2. the criterion of differentiating magnetizing inrush current and fault current is:

Criterion form 1:r _XY=min (| r _XY1|, | r _XY2|) (3)

Or criterion form 2: $r_{\mathrm{XY}}=\frac{\left|{\mathrm{<figure-callout\; id="1"\; label="r\; XY"\; filenames="A20071030376000101.png,A20071030376000111.png"\; state="\{\{state\}\}">r</figure-callout>}}_{\mathrm{XY}}\right|+\left|r_{\mathrm{XY}2}\right|}{2}---\left(4\right)$

Work as r _XY＞r _SetThe time, be judged as fault current; Work as r _XY＜r _SetThe time, be judged as magnetizing inrush current.Wherein, r _SetBe r _XYSetting value, get 0.8.

This method can be opened longitudinal difference protection fast when transformer generation internal fault in service and tape jam idle-loaded switching-on; When idle-loaded switching-on generation magnetizing inrush current, reliable locking longitudinal difference protection.

Description of drawings

Fig. 1 is a Y/ Δ wiring transformer schematic diagram in the embodiment of the invention.

Fig. 2 is the result of calculation of no-load transformer combined floodgate time difference stream in the embodiment of the invention.

Fig. 3 is the structure of unsaturation zone crude sampling ordered series of numbers and two kinds of standard sine waves when no-load transformer closes a floodgate in the embodiment of the invention, wherein,

Be original difference stream,

Be standard sine wave ordered series of numbers Y ₁, Be standard sine wave ordered series of numbers Y ₂

Fig. 4 is the result of calculation of coefficient correlation when no-load transformer closes a floodgate in the embodiment of the invention, also is the identification result of magnetizing inrush current.

Fig. 5 is the result of calculation of transformer generation turn-to-turn short circuit time difference stream in the embodiment of the invention

Fig. 6 is the structure of unsaturation zone crude sampling ordered series of numbers and two kinds of standard sine waves during transformer generation turn-to-turn short circuit in the embodiment of the invention, wherein, Be original difference stream,

Be standard sine wave ordered series of numbers Y ₁, Be standard sine wave ordered series of numbers Y ₂

Fig. 7 is the result of calculation of coefficient correlation during transformer generation turn-to-turn short circuit in the embodiment of the invention, also is the identification result of magnetizing inrush current.

Fig. 8 is the flow chart of program in the embodiment of the invention.

Embodiment

Below in conjunction with accompanying drawing principle of the present invention and concrete execution mode are described once.

Fig. 1 is a Y/ Δ wiring transformer schematic diagram, Y side three-phase current i _A, i _B, i _CCan measure Δ side three-phase current i by current transformer _a, i _b, i _c

Fig. 2 is the result of calculation of no-load transformer combined floodgate time difference stream in the embodiment of the invention.Consider the connection group and the no-load voltage ratio relation of transformer, Δ side electric current is converted Y side electric current after, respectively differed stream with Y side current summation again.From three waveforms that differ stream as can be seen, they all differ bigger with sine wave.

Fig. 3 is the structure of unsaturation zone crude sampling ordered series of numbers and two kinds of standard sine waves when no-load transformer closes a floodgate in the embodiment of the invention.As can be seen, for magnetizing inrush current, two standard sine wave ordered series of numbers Y that constructed ₁, Y ₂Differ bigger, Y with the crude sampling waveform ₁, Y ₂Itself also differ bigger.

Fig. 4 is the result of calculation of coefficient correlation when no-load transformer closes a floodgate in the embodiment of the invention, also is the identification result of magnetizing inrush current.By the coefficient correlation of calculating among the figure as can be seen, the coefficient correlation of three-phase is all less than setting value 0.8, so identify transformer excitation flow reliably, reliable locking longitudinal difference protection.

Fig. 5 is the result of calculation of transformer generation turn-to-turn short circuit time difference stream in the embodiment of the invention.Consider the connection group and the no-load voltage ratio relation of transformer, Δ side electric current is converted Y side electric current after, respectively differed stream with Y side current summation again.From three waveforms that differ stream as can be seen, they are all similar to sine wave.

Fig. 6 is the structure of unsaturation zone crude sampling ordered series of numbers and two kinds of standard sine waves during transformer generation turn-to-turn short circuit in the embodiment of the invention.As can be seen, for internal fault, two standard sine wave ordered series of numbers Y that constructed ₁, Y ₂Very similar to the crude sampling waveform, Y ₁, Y ₂Itself is also similar.

Fig. 7 is the result of calculation of coefficient correlation during transformer generation turn-to-turn short circuit in the embodiment of the invention, also is the identification result of magnetizing inrush current.By the coefficient correlation of calculating among the figure as can be seen, the coefficient correlation of three-phase is very fast to surpass setting value 0.8, so differentiate to fault current reliably, can open longitudinal difference protection fast.

Fig. 8 is the flow chart of program in the embodiment of the invention, at first obtain transformer both sides electric current, the difference of calculating transformer both sides flows again, seek the non-saturated region in the transformer difference stream, according to two sine waves of unsaturation data configuration, the coefficient correlation of the original difference stream and two sine waves of constructing in the calculating non-saturated region, last size according to coefficient correlation is differentiated magnetizing inrush current.

Claims (2)

1, a kind of discrimination method of transformer excitation flow is characterized in that, described method realizes in computer successively according to the following steps:

Step (1). determine to comprise the crude sampling ordered series of numbers of the difference stream of unsaturation Operational Zone, described poor stream is meant former limit, secondary phase current poor of the transformer that is moving, and is measured by current transformer respectively by described computer:

Step (1.1). select the observation data window

Set: the phase sampling number is n weekly, and n is an even number, and the transformer spill current i (k) that to get two cycle length be 2n is as the observation data window, and k is the sampled point sequence number, surveys the data window from taking on a new look and gets the short data window that a length is n/2-1 (odd number);

Step (1.2). judge the biased direction of difference stream

In the observation data window described in the step (1.1), seek the maximum i of difference stream _MaxWith minimum value i _Min:

If: i _Max-i _Min〉=0, then be forward bias difference stream,

If: i _Max-i _Min＜0, then be negative sense bias difference stream;

Step (1.3). seek the crude sampling ordered series of numbers of the difference stream that comprises the unsaturation Operational Zone

Starting point by the described observation data window of step (1.1) begins, the short data window is moved a sampled point successively backward, calculate the algebraical sum S (k) of each difference stream in the described short data window, k=1 ~ n/2+1, the relatively algebraical sum of n/2+1 interior each sampled point of short data window, (1.2) described method judgement set by step again:

If: be forward bias difference stream, then get that short data window of the algebraical sum minimum of the poor stream on each sampled point, form the crude sampling ordered series of numbers that difference flows by the difference stream in this short data window,

If: be negative sense bias difference stream, then get that short data window of the algebraical sum maximum of the difference stream on each sampled point, form the crude sampling ordered series of numbers of difference stream by the difference stream in this short data window;

Step (2). the structure standard sine wave

Peak value i with original difference stream sampling ordered series of numbers in the described short data window in the step (1.3) _pAs the peak value of the standard sine wave that will construct, be the center with this peak point again, about it, serve as to construct at interval n/4-1 point with 2 π/n, put by whole n/2-1 and construct standard sine wave ordered series of numbers Y ₁, Y _1k=i _pSin (pi/2-2 π k/n), k=1 ~ (n-1)/2;

Flow the peak value i of sampling ordered series of numbers again with original difference in the described short data window in the step (1.3) _pAs the peak value of the standard sine wave that will construct, but the position of this peak value is identical with the peaked position of difference stream among the described poor initial data ordered series of numbers X that flows, and establishing the position is m, again according to Y _2k=i _pSin[pi/2-2 π (k-m)/n], k=1 ~ (n-1)/2, structure standard sine wave ordered series of numbers Y ₂

Step (3). discern magnetizing inrush current according to the following steps

Step (3.1). be calculated as follows the crude sampling ordered series of numbers X of the described difference stream of step (1.3) and the correlation coefficient r of the described standard sine wave ordered series of numbers of step (2), wherein:

The crude sampling ordered series of numbers X and the standard sine wave ordered series of numbers Y of difference stream ₁Correlation coefficient r _XY1For:

$r_{\mathrm{XY}1}=\frac{\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}(X_k-\stackrel{\&OverBar;}{X})(Y_{1k}-{\stackrel{\&OverBar;}{Y}}_1)}{\sqrt{\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}{(X_k-\stackrel{\&OverBar;}{X})}^2\&CenterDot;\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}{(Y_{1k}-{\stackrel{\&OverBar;}{Y}}_1)}^2}},$

The crude sampling ordered series of numbers X and the standard sine wave ordered series of numbers Y of difference stream ₂Correlation coefficient r _XY2For:

$r_{\mathrm{XY}2}=\frac{\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}(X_k-\stackrel{\&OverBar;}{X})(Y_{2k}-{\stackrel{\&OverBar;}{Y}}_2)}{\sqrt{\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}{(X_k-\stackrel{\&OverBar;}{X})}^2\&CenterDot;\underset{k=1}{\overset{n/2-1}{\mathrm{\&Sigma;}}}{(Y_{2k}-{\stackrel{\&OverBar;}{Y}}_2)}^2}},$

Wherein, X, Y ₁, Y ₂Be respectively ordered series of numbers X, Y ₁, Y ₂Mean value,

Step (3.2). differentiate magnetizing inrush current by following criterion:

r _XY＝min(|r _XY1|，|r _XY2|)，

Work as r _XY＞r _SetThe time, be fault current,

r _XY＜r _SetThe time, be magnetizing inrush current,

r _SetBe correlation coefficient r _XYSetting value, get r _XY=0.8.

2, the discrimination method of a kind of transformer excitation flow according to claim 1 is characterized in that, in the criterion described in the step (3.2) is: $r_{\mathrm{XY}}=\frac{\left|r_{\mathrm{XY}1}\right|+\left|r_{\mathrm{XY}2}\right|}{2}.$

Images