CN104810796B - Excitation surge current discrimination method based on variable data window normalized area index - Google Patents
Excitation surge current discrimination method based on variable data window normalized area index Download PDFInfo
- Publication number
- CN104810796B CN104810796B CN201510218433.9A CN201510218433A CN104810796B CN 104810796 B CN104810796 B CN 104810796B CN 201510218433 A CN201510218433 A CN 201510218433A CN 104810796 B CN104810796 B CN 104810796B
- Authority
- CN
- China
- Prior art keywords
- data window
- current
- point
- sampling
- value
- 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
- 230000005284 excitation Effects 0.000 title claims abstract description 14
- 238000012850 discrimination method Methods 0.000 title abstract 2
- 238000005070 sampling Methods 0.000 claims abstract description 101
- 238000000034 method Methods 0.000 claims description 32
- 238000010606 normalization Methods 0.000 claims description 12
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 7
- 230000007704 transition Effects 0.000 claims description 2
- 230000000903 blocking effect Effects 0.000 abstract 2
- 238000010586 diagram Methods 0.000 description 10
- 229920006395 saturated elastomer Polymers 0.000 description 9
- 101100511184 Dictyostelium discoideum limB gene Proteins 0.000 description 2
- FFBHFFJDDLITSX-UHFFFAOYSA-N benzyl N-[2-hydroxy-4-(3-oxomorpholin-4-yl)phenyl]carbamate Chemical compound OC1=C(NC(=O)OCC2=CC=CC=C2)C=CC(=C1)N1CCOCC1=O FFBHFFJDDLITSX-UHFFFAOYSA-N 0.000 description 2
- 230000001747 exhibiting effect Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Landscapes
- Emergency Protection Circuit Devices (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a kind of excitation surge current discrimination method based on variable data window normalized area index, using variable data window distinguished number：Threshold value ± ε is set respectively under on a timeline, using threshold value and the extreme value of a certain section of sample rate current, determines the starting point and terminating point of variable data window；And in the variable data window selected; according to sampling interval and data window length; fictionalize a standard sine curve backward forward with this extreme value, the normalized area index of difference area between standard of comparison sine curve and sample rate current, and then realize the locking and opening of protection.Inrush and fault current waveform can be distinguished to greatest extent, can be with fast and reliable Blocking Differential Protection when excitation surge current occurs for transformer；Blocking Differential Protection is not missed during failure internally；Differential protection can also be quickly opened in transformer belt failure idle-loaded switching-on；And there is stronger anti-transformer saturability.
Description
Technical Field
The invention relates to an excitation inrush current identification method based on a variable data window normalized area index, and belongs to the technical field of power systems.
Background
At present, the existing power transformer is widely used in power plants and substations as an important component for converting and transmitting electric energy. The differential protection is the main protection of the transformer, and ideally, when the transformer differential protection is in normal operation and is in external short circuit, the current flowing into the differential relay is zero, and the protection device is reliable and does not act. In practice, when the transformer is switched on in no-load state to generate excitation inrush current, the magnitude of the excitation inrush current reaches several times or even dozens of times of rated current of the transformer; on the other hand, the magnetizing inrush current flows only on the power supply side of the transformer, but no current flows on the load side due to an open circuit, and the magnetizing inrush current completely flows into the differential circuit of the longitudinal differential protection, so that a very large unbalanced current is generated in the differential relay. Therefore, discrimination between magnetizing inrush current and internal fault current becomes a key issue to enhance protection reliability.
Disclosure of Invention
The purpose is as follows: in order to overcome the defects in the prior art, the invention provides an excitation inrush current identification method based on a variable data window normalized area index.
The technical scheme is as follows: in order to solve the technical problems, the technical scheme adopted by the invention is as follows:
a variable data window normalization area index-based excitation surge current identification method is characterized in that a variable data window identification algorithm is adopted: respectively setting threshold values +/-epsilon on the upper and lower parts of a time axis, and determining the starting point and the ending point of a variable data window by using the threshold values and the extreme values of a certain section of sampling current, wherein the extreme values are maximum values or minimum values; virtualizing a standard sinusoidal curve forwards and backwards according to the extreme value in the selected variable data window according to the sampling interval and the length of the data window, and calculating a normalization index Z of the difference area between the standard sinusoidal curve and the sampling current; along with the passing of sampling points, the algorithm forms a plurality of variable data windows, each data window outputs a normalization index Z, and the algorithm real-timely combines the normalization index Z and a setting value Z set And comparing to realize locking and opening of protection.
The excitation inrush current identification method based on the variable data window normalization area index is characterized by comprising the following steps:
firstly, a standard sinusoidal curve is virtualized according to an extreme point in a certain section, and the virtualized standard sinusoidal curve can be divided into two sections:
the first segment is a curve segment Y between an extreme point in the data window and the starting point of the data window 1 Is formulated as:
Y 1 (k 1 )＝I lim *sin[π/2-2π(N 1 -k 1 )/N]，k 1 ＝1,2,…N 1
wherein k is 1 Is a first section of a standard sine curve Y which is virtualized out 1 The discrete point sequence number of (2); i is lim Is the maximum value (maximum value or minimum value) in the data window, N is the total number of sampling points in one period 1 Is a curve segment Y 1 Total number of data points;
the second section is a curve section Y from an extreme point in the data window to a termination point of the data window 2 Is formulated as:
Y 2 (k 2 )＝I lim *sin[π/2-2πk 2 /N]，k 2 ＝1,2,…N 2
wherein k is 2 Is a second section of a standard sine curve Y which is virtualized 2 Discrete point number of (2), N 2 Is a curve segment Y 2 Total number of data points;
the total standard sinusoid can be defined as:
y＝[Y 1 Y 2 ]
order S 1 Is the area of the sampled current curve with respect to the time axis, S 2 Is the area of the virtual standard sine curve and the time axis, then S 1 And S 2 Can be expressed as:
wherein N is 3 Is Y 1 And Y 2 Total length of data window j is N 3 The sampling point serial number of the actual current in the data window, k is Y 1 And Y 2 The serial number of discrete points of the total standard sinusoid y taken together, i (j) is the instantaneous value of the sampled current at the jth sampling point in the total data window, and y (k) is the instantaneous value of the total standard sinusoid (the length of which is the length of the total data window) at the kth discrete point.
And enabling the serial numbers of the sampling points of the actual current in the data window to correspond to the serial numbers of the discrete points of the standard sinusoidal curve y, namely j = k, and calculating the difference area S between the sampling current and the standard sinusoidal curve as follows:
normalizing the discrimination index to obtain S min ＝min(S 1 ，S 2 )
Z＝S/S min
S min Is S 1 ，S 2 The smaller the value, Z is the index for judging the magnetizing inrush current or the fault current by the variable data window; z is a linear or branched member set Setting value of Z;
along with the passing of sampling points, the algorithm forms a plurality of variable data windows, each data window outputs a normalized area index Z, and the algorithm real-timely compares the normalized area index Z with a certain value Z set Making a comparison when Z&When Zset, judging the current waveform in the data window as inrush current; when Z is ≤ At Zset, the current waveform in the data window is discriminated as a fault current.
Wherein, the selection of the variable data window specifically refers to: firstly, respectively setting a threshold value +/-epsilon above and below a time axis, setting a protection sampling rate f, setting a sampling interval T =1/f, and setting the number of sampling points per period as N, and comprising the following steps:
1) Determination of maxima or minima: before the system is switched on normally and in an idle load state, the differential flow is within a threshold interval; when the differential current exceeds a threshold interval, starting a variable data window discrimination algorithm, selecting a sampling point when the sampling current penetrates out of the threshold interval as a starting point of first extremum value search, and selecting a sampling point when the sampling current penetrates out of the threshold interval again as a termination point of the first extremum value search; finding the first positive maximum I in the data window from the start point to the end point of the first extreme search max Or reversed minimum value I min Recording the sampling value I corresponding to the extreme point lim And sampling point number n, searching in the same waySecond extreme point I lim2 ；
2) Determination of the variable data window: at a first extreme value I lim For the forward and backward search of the base point, the first point of the current passing through the threshold region in the forward search process is defined as the starting point of the variable data window, and the instantaneous value of the starting point is I 1 The sampling point serial number is a, and the searching condition is as shown in a formula (1); if no sampling point is left in the area, the first point after the current crosses the threshold area is taken as the starting point I of the variable data window in the forward search process 1， The search condition is given by equation (2), and the algorithm for finding the starting point of the data window is: when in use
i(j+1)<＝|ε|&i(j)>|ε| j＝n-1,n-2,n-3… (1)
Or
When the temperature of the water is higher than the set temperature,
let a = j, I 1 ＝i(a)
Wherein I is sampling current instantaneous value, j is current sampling point serial number, the determined sampling point serial number of the data window starting point is a, and the instantaneous value of the data window starting point is I 1 N is the number of extreme sampling points, I (n) = I lim ；
In the backward search process, the first point of the current after passing out of the threshold region is taken as the termination point of the variable data window, and the instantaneous value is I 2 The sampling point serial number is b, and the searching condition is as shown in a formula (3); if no sampling point is left in the area, the first sampling point after the current crosses the threshold area is also used as the termination point I of the variable data window in the backward search process 2 The search condition is as in formula (4), i.e.
i(j-1)<＝|ε|&i(j)>|ε| j＝n+1,n+2,n+3… (3)
Or
When the temperature of the water is higher than the set temperature,
let b = j, I 2 ＝i(b)；
Wherein I is the sampling current instantaneous value, j is the current sampling point serial number, the sampling point serial number of the determined data window termination point is b, and the instantaneous value of the data window termination point is I 2 N is the number of extreme sampling points, I (n) = I lim ；
3) Calculating a first extreme value I lim And I 1 Number of sampling points in between N 1 Comprising I lim And I 1 Two points, virtualize the first segment of the standard sinusoid, let
Y 1 (k 1 )＝I lim *sin[π/2-2π(N 1 -k 1 )/N]，k 1 ＝1,2,…N 1
Calculate I lim And I 2 Number of sampling points in between N 2 Does not contain I lim But comprises I 2 Point, virtualize the second segment of the standard sine curve, order
Y 2 (k 2 )＝I lim *sin[π/2-2πk 2 /N]，k 2 ＝1,2,…N 2
Will sequence Y 1 And Y 2 Splicing to make
y＝[Y 1 Y 2 ]
4) Calculate I 1 ，I 2 Number of points between N 3 Comprising I 1 And I 2 And enabling the serial numbers of the sampling points of the actual current in the data window to correspond to the serial numbers of the discrete points of the standard sinusoidal curve y, namely j = k, and calculating the difference area S between the sampling current and the standard sinusoidal curve as follows:
along with the transition of the data window, the algorithm continuously outputs the Z value, and the judgment rule of Z value output under different conditions is shown in the following table 1:
TABLE 1 decision rules for Z-value output under different conditions
Setting a threshold value +/-epsilon, and setting epsilon =0.18I in consideration of matching with a transformer longitudinal differential protection starting current N ，I N The rated current of the transformer.
Considering the reliability of protection, set Z set ＝0.55。
Has the advantages that: the invention provides a variable data window normalization area index-based excitation inrush current identification method. The method has the main innovation points that: 1. the method skillfully utilizes the method of setting the threshold value near the zero point, solves the problem of selecting the variable data window when some inrush current waveforms of the super-large transformer are all positioned at one side of a time axis, namely the problem of not exceeding the zero point, and can also unify the selection rules of the variable data window selection algorithm under different conditions by setting the threshold value interval. 2. By analyzing the current waveform characteristics under different conditions, a variable data window selection algorithm is designed, and the algorithm can enable the variable data window to comprise an interruption angle part of inrush current during inrush current excitation, so that the included area between the inrush current waveform and a virtual standard sine waveform is maximized as much as possible; under the condition of a fault, the area between the fault current waveform and the virtual standard sine waveform is minimized as much as possible. 3. Through the difference of Z values output by the variable data window in sequence and the combination of the split-phase locking principle, the transformer can be unlocked or even unlocked quickly under the condition of no-load switching-on with a fault. 4. The algorithm has strong transformer saturation resistance.
The effectiveness and superiority of the technical scheme of the invention are verified through dynamic simulation tests, and when the transformer is subjected to no-load closing with a 1.7% turn-to-turn short circuit fault, the differential protection is opened 20ms after the closing, and the speed of the open protection is 23 times faster than that of the second harmonic braking. And when the no-load closing mutual inductor is saturated, the algorithm still can ensure that the protection is reliably locked.
Drawings
FIG. 1 is a diagram of a search for maxima or minima of a variable data window;
FIG. 2 is a diagram of an asymmetric inrush variable data window;
FIG. 3 is a diagram of a symmetrical inrush current variable data window;
FIG. 4 is a selection of a fault current variable data window;
FIG. 5 is a diagram of a variable data window with fault no-load closing;
FIG. 6 (a) is a diagram of normal no-load closing three-phase difference flow and (b) Z value output;
FIG. 7 (a) is a three-phase difference current diagram of a 1.7% turn-to-turn short circuit fault with no-load closing and (b) a Z value output diagram;
fig. 8 (a) is a three-phase differential current diagram of no-load closing in an internal earth fault and (b) a Z value output diagram;
FIG. 9 is a graph of internal ground fault three-phase differential current and (b) Z-value output during transformer operation;
fig. 10 (a) is a three-phase difference current and (b) Z value output diagram when the no-load closing transformer is saturated.
Detailed Description
The present invention will be further described with reference to the following examples.
As shown in fig. 1 to 10, a magnetizing inrush current identification method based on a variable data window normalized area index includes:
respectively setting threshold values +/-epsilon on the upper and lower parts of a time axis, and determining the starting point and the ending point of a variable data window by using the threshold values and the extreme values of a certain section of sampling current, wherein the extreme values are maximum values or minimum values; and virtualizing a standard sinusoidal curve forwards and backwards according to the extreme value in the selected variable data window according to the sampling interval and the length of the data window, and comparing the normalized area indexes of the difference areas between the standard sinusoidal curve and the sampling current so as to realize the locking and opening of protection. The method can distinguish the inrush current waveform and the fault current waveform to the maximum extent, and can quickly and reliably lock differential protection when the transformer generates excitation inrush current; differential protection is not locked by mistake when internal faults occur; the differential protection can be quickly opened when the transformer is switched on with a fault and no load; and has stronger mutual inductor saturation resistance.
Therefore, as the data window is changed, the algorithm continuously outputs the Z value, and can classify and discuss various situations, and the decision rule of Z value output in different situations is given in table 1:
TABLE 1 decision rule for Z value output under different conditions
The invention not only can quickly and reliably lock the differential protection when the transformer is normally closed in a no-load way, but also can not lock the differential protection by mistake when the transformer has internal faults, and can quickly open the differential protection when the transformer is closed in a no-load way with faults.
Example 1:
a) Selecting a variable data window: introducing a variable data window selection process by combining an inrush current identification algorithm, firstly setting small threshold values +/-epsilon above and below a time axis respectively, setting a protection sampling rate f, setting a sampling interval T =1/f, and setting the number of sampling points per period as N; the method comprises the following steps:
1. determination of maxima or minima: before the system is normally closed and is unloaded, the differential flow is within a threshold interval. When the differential current exceeds the threshold interval, starting a variable data window discrimination algorithm, as shown in fig. 1, selecting a sampling point when the sampling current penetrates out of the threshold interval as a starting point of first extremum search, and selecting a sampling point when the sampling current penetrates out of the threshold interval again as a terminating point of the first extremum search; finding out the first positive maximum or negative minimum in the data window, and recording the extreme valueSampling value I corresponding to point lim (I in FIG. 1) max ) And the sampling point number n, the second extreme point I can be found according to the above rule lim2 (I in FIG. 1) min )；
2. Determination of variable data window (here the maximum I found in the first variable data window of fig. 1) max For illustration purposes): at a maximum value of I max For the forward and backward search of the base point, it is provided that the first point of the forward search process after the current has passed out of the threshold region is used as the starting point of the variable data window, and the instantaneous value is I 1 The sampling point serial number is a, and the searching condition is as formula (1); if no sampling point is left in the area, the first point after the current crosses the threshold area is taken as the starting point I of the variable data window in the forward search process 1， The search condition is as in equation (2), and the algorithm for finding the starting point of the data window is as follows:
when in use
i(j+1)<＝|ε|&i(j)>|ε| j＝n-1,n-2,n-3… (1)
Or
When the temperature of the water is higher than the set temperature,
let a = j, I 1 ＝i(a)
Wherein I is sampling current instantaneous value, j is current sampling point serial number, the determined sampling point serial number of the data window starting point is a, and the instantaneous value of the data window starting point is I 1 N is the number of extreme sampling points, I (n) = I max ；
The first point of the backward search process after the current passes through the threshold region is taken as the end point of the variable data window, and the instantaneous value is I 2 The sampling point serial number is b, and the searching condition is as shown in formula (3); if no sampling point is left in the area, the first sampling point after the current crosses the threshold area is also used as the termination point I of the variable data window in the backward search process 2 The search condition is as in formula (4), i.e.
i(j-1)<＝|ε|&i(j)>|ε| j＝n+1,n+2,n+3… (3)
Or
When the temperature of the water is higher than the set temperature,
let b = j, I 2 ＝i(b)；
Wherein I is the sampling current instantaneous value, j is the current sampling point serial number, the sampling point serial number of the determined data window termination point is b, and the instantaneous value of the data window termination point is I 2 N is the number of extreme sampling points, I (n) = I max ；
3. Calculating a first extreme value I max And I 1 Number of sampling points in between N 1 Comprising I max And I 1 Two points, virtualize the first segment of the standard sine curve, let
Y 1 (k 1 )＝I max *sin[π/2-2π(N 1 -k 1 )/N]，k 1 ＝1,2,…N 1
Calculate I max And I 2 Number of sampling points N in between 2 Does not contain I max But comprises I 2 Point, virtualize the second segment of the standard sine curve, order
Y 2 (k 2 )＝I max *sin[π/2-2πk 2 /N]，k 2 ＝1,2,…N 2
Will sequence Y 1 And Y 2 Splicing is carried out to make
y＝[Y 1 Y 2 ]
4. Calculate I 1 ，I 2 Number of points between N 3 Comprising I 1 And I 2 As shown in fig. 2, if the sampling point numbers of the actual currents in the data window correspond to the discrete point numbers of the standard sinusoidal curve y, i.e. j = k, the difference area S between the sampling current and the standard sinusoidal curve is calculated as:
as shown in fig. 2 and 3, in the case of asymmetric inrush current and symmetric inrush current, the data windows are selected according to the above algorithm, and since the data windows include the discontinuous corner portions, the area S of the shaded portion of each variable data window becomes very large, and the algorithm continuously outputs a large normalized area index Z. For internal faults, as shown in fig. 4, the area S of the shaded portion of each variable data window is small, and the algorithm continuously outputs a small normalized area index Z.
For the no-load closing with the fault, from the local waveform of the current, although the magnetizing inrush current is larger in the saturation section of the transformer and the characteristic of the fault current is covered, when the transformer is out of saturation and enters a non-saturation region, the current will show the characteristic of the fault current. The variable data window selection algorithm has the capability of automatically selecting the unsaturated segment current. As shown in fig. 5, the variable data window selection algorithm can alternately select the current in the saturated section and the current in the unsaturated section in the first period after the no-load switching with the fault, and then continuously and alternately output the larger and smaller normalized area indexes Z, so that the algorithm is completely capable of identifying the switching-on condition with the fault in a shorter time, and the protection is opened as soon as possible.
B) Setting of normalized area index and criterion
In practical situations, the protection sampling rates may be inconsistent, the data windows may be different in size according to the data window selection algorithm, and the magnitudes of currents in different situations are different, so that the indexes for measuring the area magnitudes of the shadow parts in different situations may be inconsistent. To solve this problem, the criterion needs to be normalized and set
Wherein N is 3 Is Y 1 And Y 2 Total length of data window j is N 3 The sampling point serial number of the actual current in the data window, k is Y 1 And Y 2 The serial number of discrete points of the total standard sinusoid y taken together, i (j) is the instantaneous value of the sampled current at the jth sampling point in the total data window, and y (k) is the instantaneous value of the total standard sinusoid (the length of which is the length of the total data window) at the kth discrete point.
And enabling the serial numbers of the sampling points of the actual current in the data window to correspond to the serial numbers of the discrete points of the standard sinusoidal curve y, namely j = k, and calculating the difference area S between the sampling current and the standard sinusoidal curve as follows:
normalizing the discrimination index to S min ＝min(S 1 ，S 2 )
Z＝S/S min
S min Is S 1 ，S 2 The smaller Z is the index of judging the magnetizing inrush current or the fault current by the variable data window; z is a linear or branched member set Setting value of Z;
along with the passing of sampling points, the algorithm forms a plurality of variable data windows, each data window outputs a normalized area index Z, and the algorithm real-timely compares the normalized area index Z with a certain value Z set Making a comparison when Z&When Zset, judging the current waveform in the data window as inrush current; when Z is ≤ At Zset, the current waveform in the data window is discriminated as a fault current. The criterion for identifying the fault current and the magnetizing inrush current is as follows:
preferably, the threshold value +/-epsilon is set in consideration of matching with the transformer longitudinal differential protection starting current, wherein epsilon =0.18I N ，I N For the transformerAnd (4) fixing the current. Considering the reliability of protection, set Z set ＝0.55。
Example 2:
the algorithm is verified in a moving die test mode, the effectiveness and the superiority of the scheme are verified in the moving die test, and 20ms after the switching-on is protected to open differential protection when the transformer has no-load switching-on with 1.7% turn-to-turn short circuit fault. The following is the analysis of the behavior characteristics of the invention under various working conditions:
1. normal no-load closing of transformer
FIG. 6 (a) is the three-phase differential current I, wherein the three phases are A, B and C, and the three-phase differential current I is shown under the normal no-load closing condition of the transformer A 、I B And I C Indicated by solid, dashed and dotted lines, respectively. As can be seen from FIG. 6 (a), the transformer is switched on at 0.02s time-space load, I in three-phase difference current A And I B The water is single-side surge flow, and has obvious discontinuous angles which are larger; I.C. A C Symmetrical inrush current is adopted, and the break angle is small; i is B Since the influence of the direct current component is not zero, the selection of the variable data window is realized by a method of setting a threshold value.
Three-phase Z values are respectively Z A 、Z B And Z C (ii) a As can be seen from FIG. 6 (b), due to I A And I B For asymmetric inrush currents, I C For symmetric inrush current, the algorithm, Z, is selected according to a variable data window in 5 cycles A And Z B The number of output points is Z C Half of the number of output points. The first Z value output by the C phase 6ms after no-load closing is larger than Z set And the A phase and the B phase respectively lock differential protection 24ms after no-load closing according to algorithm criteria.
2. No-load switch-on in transformer slight turn-to-turn short circuit fault
Fig. 7 (a) is a three-phase difference current of a 1.7% turn-to-turn short circuit fault of the star-shaped side A phase of the no-load closing transformer. Because the A phase has a short circuit fault with a small turn ratio, fault current and inrush current exist simultaneously, and the fault current is slight, so that the differential current is mainly characterized by the inrush current in a saturated section of the transformer, and the differential current is completely characterized by the fault current in a non-saturated section of the transformer. When variable data windows with inrush current characteristics and fault characteristics are extracted by using a variable data window discrimination algorithm, the three-phase Z value output is as shown in fig. 7 (b).
As can be seen from fig. 7 (b), the a-phase Z value outputs each exhibit a form of fluctuation up and down, because the a-phase differential current is continuously divided alternately into a data window exhibiting the inrush current characteristic and a data window exhibiting the failure characteristic. Within the data window showing the inrush current characteristic, the Z value output is larger; in the data window showing fault characteristics, the difference flow waveform is very similar to a standard sine wave, so that Z value output is small. The first Z value point output by the phase A13 ms after no-load closing is judged as inrush current, and the phase A is quickly locked; but at 20ms, a second Z value point is output, which is less than Z set If the current is judged to be the fault current, the protection of the phase A is rapidly opened according to the judgment rule of the table 1, and the differential protection can rapidly act due to the adoption of the split-phase locking principle. Under the condition of the same test current, the second harmonic braking scheme opens protection about 460ms after no-load closing.
3. No-load switch-on in transformer internal grounding fault
Fig. 8 (a) shows an internal earth fault of the star-side a-phase of the no-load closing transformer. Due to a relatively severe failure of phase A, then I A And I C Differential current is primarily characterized by fault current, and I B There is still a large magnetizing inrush current characteristic.
As can be seen from FIG. 8 (b), Z is due to the influence of the fault current component A And Z C Output less than Z set . And I B Z is due to still showing the inrush current characteristic B The output is still greater than Z set . And because of adopting the split-phase locking logic, the A phase and the C phase can not be locked after no-load closing, and the differential protection can rapidly act.
4. Internal ground fault of transformer in operation
Fig. 9 (a) shows that in operation, the star-shaped side A phase of the transformer is grounded, and three-phase difference flows all present sinusoidal waveform characteristics.
As can be seen from fig. 9 (b), the Z values of the three phases are all one due to the occurrence of the internal ground faultOutput twice in the period, and the output values are far less than the set value Z set The protection is not locked.
5. Current transformer saturation during no-load closing
Fig. 10 (a) shows three-phase differential current when the transformer is saturated during no-load closing, since the symmetrical inrush current of phase B is small, the transformer is not saturated, while phase a and phase C are both saturated, the inrush current rapidly drops when the transformer is saturated, a reverse charging current is generated, and the discontinuity angle disappears. FIG. 10 (b) shows the three-phase Z value output in this case, and it can be seen that the Z values of the A phase and the C phase are smaller than the output of the transformer in the unsaturated case of FIG. 10 (b) due to saturation, but the three-phase Z value output is in Z set Thus, the protection can be reliably locked.
The above description is only of the preferred embodiments of the present invention, and it should be noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the invention and these are intended to be within the scope of the invention.
Claims (5)
1. A variable data window normalization area index-based excitation surge current identification method is characterized in that a variable data window identification algorithm is adopted: respectively setting thresholds +/-epsilon above and below a time axis, and determining a starting point and an ending point of a variable data window by using the thresholds and extreme values of a certain section of sampling current, wherein the extreme values are maximum values or minimum values; virtualizing a standard sinusoidal curve forwards and backwards according to the extreme value in the selected variable data window according to the sampling interval and the length of the data window, and calculating a normalization index Z of the difference area between the standard sinusoidal curve and the sampling current; along with the passing of sampling points, the algorithm forms a plurality of variable data windows, each data window outputs a normalization index Z, and the algorithm real-timely combines the normalization index Z and a setting value Z set And comparing to realize locking and opening of protection, specifically comprising the following steps:
firstly, a standard sinusoidal curve is virtualized according to an extreme point in a certain section, and the virtualized standard sinusoidal curve can be divided into two sections:
the first segment is a curve segment Y between an extreme point in the data window and the starting point of the data window 1 Is formulated as:
Y 1 (k 1 )＝I lim *sin[π/2-2π(N 1 -k 1 )/N]，k 1 ＝1,2,…N 1
wherein k is 1 Is a first section of a standard sine curve Y which is virtualized out 1 The discrete point number of (2); I.C. A lim Is an extreme value in the data window; n is the total number of sampling points in one cycle, N 1 Is a curve segment Y 1 Total number of data points;
the second section is a curve section Y between an extreme point in the data window and a termination point of the data window 2 Is formulated as:
Y 2 (k 2 )＝I lim *sin[π/2-2πk 2 /N]，k 2 ＝1,2,…N 2
wherein k is 2 Is a second section of a virtual standard sine curve Y 2 Discrete point number of (2), N 2 Is a curve segment Y 2 Total number of data points;
the total standard sinusoid can be defined as:
y＝[Y 1 Y 2 ]
order S 1 Is the area of the sampled current curve with respect to the time axis, S 2 Is the area of the virtual standard sine curve and the time axis, then S 1 And S 2 Can be expressed as:
wherein N is 3 Is Y 1 And Y 2 The total length of the data window is combined, j is the sampling point serial number of the actual current, and k is Y 1 And Y 2 The discrete point numbers of the total standard sinusoid y taken together, i (j) are the dataSampling the instantaneous value of the current at the jth sampling point in the total length of the window, wherein y (k) is the instantaneous value of the total standard sine curve at the kth discrete point;
let N 3 And (3) the serial numbers of sampling points of the actual current in the data window correspond to the serial numbers of discrete points of the standard sinusoidal curve y, namely j = k, and then the difference area S between the sampling current and the standard sinusoidal curve is calculated as follows:
normalizing the discrimination index to S min ＝min(S 1 ，S 2 )
Z＝S/S min
S min Is S 1 ，S 2 The smaller of the two, Z set Setting value of Z;
along with the transition of the sampling point, the algorithm forms a plurality of variable data windows, each data window outputs a normalization index Z, and the algorithm combines the normalization index Z and Z in real time set Making a comparison when Z>Z set Judging the current waveform in the data window as the inrush current; when Z is less than or equal to Z set The current waveform in the data window is discriminated as a fault current.
2. The method for identifying magnetizing inrush current based on variable data window normalized area index as claimed in claim 1, wherein: the selection of the variable data window specifically refers to: firstly, respectively setting a threshold value +/-epsilon on a time axis, setting a protection sampling rate f, setting a sampling interval T =1/f, and setting the number of sampling points per period as N, and the method comprises the following steps:
1) Determination of maxima or minima: before the system is normal and unloaded, the differential flow is within a threshold interval; when the differential current exceeds the threshold interval, starting a main variable data window discrimination algorithm, selecting a sampling point when the sampled current penetrates out of the threshold interval as a starting point of first extremum value search, and selecting a sampling point when the sampled current penetrates out of the threshold interval again as a terminating point of the first extremum value search; searching at the first extremeSearching the first positive maximum I in the data window from the starting point to the ending point of the cable max Or reversed minimum value I min Recording the sampling value I corresponding to the first extreme point lim And the number n of the extreme value sampling point, and searching a second extreme value point in the same way;
2) Determination of the variable data window: searching forwards and backwards by taking the first extreme point as a base point, and defining the first point after the current penetrates out of the threshold region as a starting point of the variable data window in the process of searching forwards, wherein the instantaneous value of the starting point is I 1 The sampling point serial number is a, and the searching condition is as formula (1); if no sampling point is left in the area, the first point after the current crosses the threshold area is taken as the starting point of the variable data window in the forward search process, the search condition is as shown in formula (2), and the algorithm for seeking the starting point of the data window is as follows:
when in use
i(j+1)≤|ε|&i(j)>|ε|j＝n-1,n-2,n-3… (1)
Or
When the temperature of the water is higher than the set temperature,
let a = j, I 1 ＝i(a)
Wherein I is the instantaneous value of the sampled current, j is the serial number of the sampling point of the actual current, n is the serial number of the sampling point of the extreme value, I (n) = I lim ；
In the backward search process, the first point of the current after passing out of the threshold region is taken as the termination point of the variable data window, and the instantaneous value is I 2 The sampling point serial number is b, and the searching condition is as shown in formula (3); if no sampling point is left in the area, the first sampling point after the current crosses the threshold area is also used as the ending point of the variable data window in the backward search process, and the search condition is as the formula (4), namely
i(j-1)≤|ε|&i(j)>|ε|j＝n+1,n+2,n+3… (3)
Or
When the utility model is used, the water is discharged,
let b = j, I 2 ＝i(b)；
Wherein I is the instantaneous value of the sampled current, j is the serial number of the sampling point of the actual current, n is the serial number of the extreme value sampling point, I (n) = I lim ；
3) Calculating the first extreme point and I 1 Number of sampling points in between N 1 Comprising I lim And I 1 Two points, virtualize the first segment of the standard sinusoid, let
Y 1 (k 1 )＝I lim *sin[π/2-2π(N 1 -k 1 )/N]，k 1 ＝1,2,…N 1
Calculating the first extreme point and I 2 Number of sampling points in between N 2 Does not contain I lim But comprises I 2 Point, virtualize the second segment of the standard sine curve, order
Y 2 (k 2 )＝I lim *sin[π/2-2πk 2 /N]，k 2 ＝1,2,…N 2
Will sequence Y 1 And Y 2 Splicing is carried out to make
y＝[Y 1 Y 2 ]
4) Calculate I 1 ，I 2 Number of points between N 3 Comprising I 1 And I 2 Let N stand for 3 And (3) the serial numbers of sampling points of the actual current in the data window correspond to the serial numbers of discrete points of the standard sinusoidal curve y, namely j = k, and then the difference area S between the sampling current and the standard sinusoidal curve is calculated as follows:
3. the variable data window normalized area index-based magnetizing inrush current identification method according to claim 1, characterized in that: along with the passage of the data window, the algorithm continuously outputs the Z value, and the judgment rule of the Z value output under different conditions is as follows: in one period, the Z value is output once or twice under different conditions;
(1) For the case that the Z value is output once in one period, only Z is output at the moment>Z set If the situation is determined to be inrush current, locking is carried out, the locking time is shorter than a period, no opening time exists, and the waveform is asymmetric inrush current;
(2) Judging whether protection is locked or not or whether opening is guaranteed or not by using the first output of the Z value under the condition that the Z value is output twice in a period, and then, not locking; carrying out re-judgment by using Z value for second output, determining whether the protection is re-opened after locking when the Z value is output for the first time, and when Z is not more than Z set The protection is reopened; when Z is>Z set Or the Z value is not output for the second time, and the locking protection is continued; the method is specifically divided into the following 4 categories:
first, Z is output once>Z set To lock and output Z once again>Z set Continuously keeping the front locking state, wherein the current is an inrush waveform without opening time, and the inrush waveform is a symmetrical inrush current;
the second is to output Z less than or equal to Z once set Judging that the opening is guaranteed, outputting Z less than or equal to Z once again set If the fault is determined to be an internal fault, no locking and opening time exists;
the third is to output Z less than or equal to Z once set Judging as guaranteed open, outputting Z once again>Z set The method comprises the following steps of (1) re-judging whether the fault is carried out no-load switching-on, and then, not locking and opening time;
fourth, outputting Z once>Z set Then locking is performed and Z is output once again set And re-judging to be zero-load switching-on with fault, and opening a protection outlet, wherein the locking time is a half period and the opening time is a period.
4. The method for identifying magnetizing inrush current based on variable data window normalized area index as claimed in claim 1, wherein: considering the matching with the transformer longitudinal differential protection starting current, a threshold value is set to be +/-epsilon, and epsilon =0.18I N ，I N The rated current of the transformer.
5. The variable data window normalized area index-based magnetizing inrush current identification method according to claim 1, characterized in that: considering the reliability of protection, set Z set ＝0.55。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510218433.9A CN104810796B (en) | 2015-04-30 | 2015-04-30 | Excitation surge current discrimination method based on variable data window normalized area index |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510218433.9A CN104810796B (en) | 2015-04-30 | 2015-04-30 | Excitation surge current discrimination method based on variable data window normalized area index |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104810796A CN104810796A (en) | 2015-07-29 |
| CN104810796B true CN104810796B (en) | 2018-02-13 |
Family
ID=53695406
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510218433.9A Active CN104810796B (en) | 2015-04-30 | 2015-04-30 | Excitation surge current discrimination method based on variable data window normalized area index |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104810796B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106129966B (en) * | 2016-07-15 | 2018-07-31 | 华北电力大学 | The method and device for preventing transformer differential protection malfunction based on coefficient of kurtosis |
| CN106655097B (en) * | 2017-01-09 | 2018-08-17 | 湖南大学 | Power transformer excitation surge current recognition methods |
| CN112526338B (en) * | 2020-11-24 | 2024-07-02 | 重庆理工大学 | Novel synchronous time window excitation power unit online fault diagnosis method |
| CN112905958B (en) * | 2021-01-27 | 2024-04-19 | 南京国电南自电网自动化有限公司 | Short-time data window telemetry data state identification method and system based on measurement and control device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101257208A (en) * | 2007-12-21 | 2008-09-03 | 清华大学 | Method for identifying transformer excitation surge current |
| CN103633622A (en) * | 2013-12-11 | 2014-03-12 | 国家电网公司 | Method and system for ultra-high voltage regulating transformer excitation inrush current identification |
| CN105137220A (en) * | 2015-07-20 | 2015-12-09 | 贵州电网有限责任公司电力调度控制中心 | Current transformer saturation detection method based on waveform area ratio method |
-
2015
- 2015-04-30 CN CN201510218433.9A patent/CN104810796B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101257208A (en) * | 2007-12-21 | 2008-09-03 | 清华大学 | Method for identifying transformer excitation surge current |
| CN103633622A (en) * | 2013-12-11 | 2014-03-12 | 国家电网公司 | Method and system for ultra-high voltage regulating transformer excitation inrush current identification |
| CN105137220A (en) * | 2015-07-20 | 2015-12-09 | 贵州电网有限责任公司电力调度控制中心 | Current transformer saturation detection method based on waveform area ratio method |
Non-Patent Citations (2)
| Title |
|---|
| 基于最值差别的归一化面积识别励磁涌流方法研究;张凯等;《电力科学与工程》;20130531;第29卷(第5期);第15-19页 * |
| 采用自适应数据窗电流的励磁涌流鉴别方法;王业等;《中国电机工程学报》;20140205;第34卷(第4期);第702-711页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104810796A (en) | 2015-07-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104810796B (en) | Excitation surge current discrimination method based on variable data window normalized area index | |
| CN104810795B (en) | A kind of modified excitation surge current discrimination method | |
| CN105552838B (en) | Transformer zero-seguence differential protection algorithm based on the identification of standardized graphics similarity mode | |
| WO2015081826A1 (en) | Electrical line protection method | |
| CN104466903A (en) | Current transformer saturation recognition method for differential protection | |
| Sahebi et al. | Efficient method for discrimination between inrush current and internal faults in power transformers based on the non‐saturation zone | |
| CN105653882A (en) | Method for identifying magnetizing inrush current by utilizing current waveform characteristics | |
| CN103048567B (en) | Method for judging parallel reactor-containing transmission line malfunction properties based on wave form estimation | |
| CN104034997B (en) | Power converter and short circuit detection device and method | |
| CN106129976A (en) | Differential protecting method based on Hausdorff distance algorithm and device | |
| US20170192045A1 (en) | Method for Quickly Identifying Disconnection of CT in Protection of 3/2 Connection Mode Based Bus | |
| CN103439624B (en) | Supertension line fault phase selection method based on voltage fault component | |
| CN106569163A (en) | Method for detecting trailing amount of transformer | |
| Bastos et al. | Identifying switched capacitor relative locations and energizing operations | |
| CN109494684B (en) | Transformer excitation inrush current and fault differential current identification method based on babbitt distance algorithm | |
| CN113109671B (en) | Method for accelerating detection of small half-wave stage of fault current | |
| CN102156216B (en) | Method for discriminating magneting inrush current of direct current transmission converter transformer | |
| Vahidi et al. | A novel approach to adaptive single phase autoreclosure scheme for EHV power transmission lines based on learning error function of ADALINE | |
| Zhou et al. | Method of cable incipient faults detection and identification based on wavelet transform and gray correlation analysis | |
| CN102035178A (en) | Method for transformer excitation surge current latching | |
| CN110531195A (en) | A method of identification transformer excitation flow and internal fault | |
| CN113013854B (en) | Method for pre-judging faults inside and outside area based on differential current change rate in dynamic process | |
| CN114844000A (en) | Inter-turn protection method and system for converter transformer | |
| CN108429229B (en) | Method for judging excitation inrush current of no-load transformer in phase selection reclosing process of circuit breaker | |
| CN111585300B (en) | Method for automatically identifying type of power grid and inverter device thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |