JPS61221515A - Digital type protective relay - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a digital protective relay, and particularly to a digital protective relay that has little error over a wide frequency range and has improved frequency characteristics.

[Technical background of the invention]

Protective relays are installed to protect the power system, but
In this case, it may be necessary to calculate the amplitude value of the input AC amount of the power system to be protected. In this case, the digital protective relay samples the sinusoidal current and/or voltage of the power system at predetermined time intervals, converts the sampled values into digital data, encodes it, and transmits it. The amplitude value of the input AC amount is calculated by performing calculations on this digital data.

When calculating an amplitude value from this digital data input, generally, as shown in FIG. 9, digital data input processing 1 and digital filter processing 2 for attenuating the DC component and the second harmonic acid component are performed. After that, amplitude value calculation operation 3 is performed.

Here, the sampling frequency is 12 times the input AC frequency, that is, for a large input of 60 Hz, the sampling frequency is 7
Taking the case of 20Hz as an example, the sampling data is '
m l 'm-1m 'm-2"(" is a time series), using the filter, the result is the data after digital filter processing! . is obtained. And the amplitude value! .

For example, as an amplitude value calculation operation to obtain
IHl+l in I,= as shown in No. 41513
lm-5l)+Kzll■m1-llm- Rectifying addition method expressed by a river function, I, = ν& and JP-A-1983-
r, = z-! as shown in No. 88514. , -t-
-t--, h-functions such as the square method have been used conventionally.

[Problems with background technology]

The conventional method described above is designed on the assumption that the input AC is near the commercial frequency, so when the input AC is near the commercial frequency, for example around 60Hz, sufficient performance accuracy can be obtained, but if the input AC is far from the commercial frequency, The error will be large for inputs that are

Therefore, like a generator protection relay, 20 Hz to 1
This method could not be used in applications requiring a sufficiently small error for inputs of 00 Hz (when the commercial frequency is 60 Hz).

°This will be explained below. Here, the angular frequency at the commercial frequency is ω0. The angular frequency of the input alternating current is ω.

When the amplitude is expressed as I2 time by t and the sampling period is 42 of the commercial period, the input AC 'ms'm-6 is expressed by the following equation (1).

If you do the trick, the crab will come out. becomes as follows.

Im”T”m-hair□-6) =T (Icosωt-Icos(cc+t-resistant)
) The numerical characteristics are as shown in Figure 1O.

I'm going to growl.

Next, as an amplitude value calculation operation, ■. -I,! 3+I, *
I.

Taking as an example the case where the square method is used, the following equation (4) is obtained.

I, =I, 3- Irn・--6 −■t2cos2 ett-, jq, arrogance-1t2ωg(
ωt')cos(oa'-2g)-I'2dn2 According to the force, the gain of the square method described above, that is, the ratio of the input AC amplitude I' to the obtained amplitude value I6 is as follows. become.

The overall gain A that combines the total processing and the square method of x lm-4-IH"ζ: is expressed by the following equation (5), and its gain-frequency characteristics are shown in Figure 12. It becomes like this.

The amplitude value obtained by the above method represents the amplitude of the input AC when the frequency of the input AC is near the commercial frequency, but as the frequency moves away from the commercial frequency, the fluctuation width at each frequency becomes smaller than when the frequency is at the commercial frequency. Although it is small, the amplitude is estimated to be small.

Next, as another example, consider the case where the amplitude value is determined by the square method expressed as I, -iζ ears, as follows.

Engineering. =Work □+lm-5 =■/2ωS2(ωttsutenI'2cos2(artt-
No. 4) Io changes depending on t', that is, the timing of sampling, but takes a value between the following two values.

Therefore, Io will change over time unless it is between the two values shown in (6) below.

Therefore, the dyne-frequency characteristic becomes as shown in Figure 13, and depending on the sampling timing, the overall gain obtained by combining the dyne and the dyne is the gain A1 expressed by the following equation (7) and the gain A1 expressed by the equation (8). The gain frequency characteristics are as shown in FIG. 14 between the values of the gain A2 expressed by .

In other words, the amplitude value obtained by the above method represents twice the input AC frequency when the frequency of the input AC is near the commercial frequency, but as the frequency moves away from the commercial frequency, the fluctuation range becomes larger than when the frequency is at the commercial frequency. Become. Furthermore, even when the rectification and addition method described above is used for amplitude value calculation, as the frequency of the input alternating current moves away from the commercial frequency, the fluctuation range of the amplitude value increases compared to when the frequency is the commercial frequency.

When a digital protective relay that uses digital filter processing and amplitude value calculation operations as described above is used to protect power transmission lines or transformers, AC at the commercial frequency is always input, so it is necessary to use the relay away from the commercial frequency. There is no need to consider the increase in error caused by this, and this is not a problem. However, in generator protection, the frequency of the output AC changes depending on the rotation speed, so for example, in a 60Hz system, the frequency of the output AC changes depending on the rotation speed.
There is a need for a relay judgment method with as small an error as possible in a wide frequency band such as Hz to 100Hz, and a method for amplitude value calculation. .

This is a problem because the error is large for inputs that are far from the numbers.

[Purpose of the invention]

The present invention has been made to solve the above problems, and an object of the present invention is to provide a digital protective relay with small errors over a wide frequency band including commercial frequencies.

[Summary of the invention]

In the present invention, when the sampling period is /n of the commercial period (where n is a natural number), the input AC signal has a frequency characteristic in which the gain increases as it moves away from the commercial period, and by 2 conversion, − (However, lkl < 1 with 1- to -2n constants) or a digital filter of a function expressed by an approximate expression, and r the data after digital filter processing. When (m is the sampling time series),
The amplitude value is expressed as I, -1/i, n-'m''m-8, and the amplitude value is The error is reduced by complementing the .

[Embodiments of the invention]

Examples will be described below with reference to the drawings. FIG. 1 is a flow chart of an embodiment showing the procedure of digital filter processing and amplitude value calculation operation applied to a digital protective relay according to the present invention. Since the other components of the digital protective relay are well-known circuits and procedures, their explanations will be omitted. In Figure 1, the digital data input processing IA is based on sampling from the input AC signal. The obtained sample value is converted into a digital signal, and the resulting digital data im, i□-1゜'m-
2... (m is time series) is read and processed. And the sampling period is /4 u (the reciprocal of the frequency) of the commercial period (the reciprocal of the frequency).
n is a natural number). Next Digital Filter Processing Two people convert the filtered data Irn (m is a time series) using the following equation (9), which is expressed as 1+to-2n<where -klkl<1. It is then passed to the next amplitude value calculation calculation 3A.

1...(9)11nW(
1-10 kim-2n) In the amplitude value calculation calculation 3A described above, the amplitude value I0 is calculated by the square method expressed by the following equation 61.

! . ” ”m! n-'m"m-π -<1
The zero-order action will be explained. Here, if the amplitude of the input AC signal is I, the angular frequency is ω, the time is t, the commercial angular frequency is ω0, and the sampling period (the reciprocal of the frequency) is the commercial period'An, then the above digital data lr
n, 1rn-2n are represented by the following equation (α). Substitute this equation (α) into equation (9).

= to F; □-0,1m-2n is
It can be expressed as the following equation (b).

The data after these filter processing Im#'m-n
* "m-2n" is input to the amplitude value calculation calculation 3A, and the amplitude value is determined by the above-mentioned equation (G).

The formula (7) and formula (b) are as follows. Therefore, the amplitude value is calculated for the amplitude I of the input AC signal. The ratio of , that is, the gain A is expressed by the following formula 0, and its gain-frequency characteristics are as follows depending on the value of k:
For example, as shown in FIGS. 2 to 4, as explained above, in this embodiment, two conversions are performed.
Amplitude value calculation process using multiplication. =, rζs'n +η,
Since the amplitude value is determined by -I-, =, for example, -0
, 2, as shown in Figure 2, 0.41ω0≦
The range of ω≦1.59ω・, that is, the commercial frequency is 50H
E: 20.5 Hz to 79.5 Hz, 60 Hz
In the range of 24.6Hz to 95.4Hz, 0.8≦
A≦1.0), that is, the gain error is within +O1, -20%.

For example, in the case of k -0.25, as shown in Fig. 3, the range of 0.5ω0≦ω≦1.5ω0, that is, when the commercial frequency is 50 Hz, the range is 25Hz to 75Hz,
60Hz is 0.97 in the range of 30Hz to 90Hz
≦A≦1.04°, that is, the gain error is within ±4%. Furthermore, in the case of k -0.3, as shown in FIG. 4, the range of 0.37ωG≦ω≦1.63ω0, that is,
When the commercial frequency is 50 Hz, it is 18.5 Hz ~ 81.5
Hz, 22.2Hz to 97.8H at 60H+c
In the range of z, 0.9≦A≦1.1, that is, the gain error is within ±10 inches.

Therefore, the error is smaller in a wider frequency band including commercial frequencies than in the conventional method shown in FIGS. 9 to 14.

FIG. 5 is a flowchart showing the procedure of digital filter processing and amplitude value calculation operation according to another embodiment of the present invention. In addition,
Similar to FIG. 1, the other components of the digital protective relay are well-known circuits and procedures, so their explanation will be omitted.

In this embodiment, the data I! after filter processing is calculated using the following equation (b), which is expressed in digital filter processing 2B. Find n (I, , is a time series) and perform amplitude value calculation processing 3A
It is intended to be passed on to. Other processing is the same as in FIG.

Next, the effect will be explained. Similar to the embodiment described above,
Digital data 'm, 1m-2n, 1m-4n are αth
→Represented by the formula.

Substitute this α-th peak equation into the equation (b).

X plane (ωt+β-÷) ・-('-1-A8) ×al1g(cl)t+β-) Here, if ωt'=ωt+β-1, then the data after filtering is the following CLη equation It can be expressed as

×house ωt'...The gain of the digital filter to be tuned (b) is expressed by the following formula (to) formula % The gain of the amplitude value calculation operation 3A is determined by jgtn as shown in formula (4). - Since H>l, the amplitude value obtained in this example. The amplitude I of the input AC signal
The gain A with respect to is expressed by the following fourth equation.

. . . The gain-frequency characteristics are as shown in FIGS. 6 to 8, for example, depending on the value of k.

As mentioned above, in this embodiment, after processing of the digital filter using 2 transformation,
Calculate the amplitude value using -n "m" m-2n, and calculate the amplitude value! . For example, in the case of =0.37, as shown in Figure 6, 0.425ω・≦ω≦1.57
5ω. range, i.e., when the commercial frequency is 50 Hz, 21
．． 25Hz ~ 8525Hz, 25 at 60Hz
．． In the range of 5 Hz to 94.5 Hz, 0.85≦A≦
1, that is, the gain error is within +0% and -15'j. Also, in the case of = 0.4, as shown in Figure 7,
0.395ω.

≦ω≦1.605ω. range, that is, the commercial frequency is 50
In Hz, 19.75Hz ~ 8525Hz, “6
At 0 Hz, in the range of 23.7 Hz to 96.3 Hz, 0.85≦A≦1, that is, the dyne error is +0%, -
It falls within 15%.

Furthermore, in the case of k=0.46, as shown in FIG.
．． 295ω. ≦ω≦1.705ω. range, i.e., when the commercial frequency is 50 Hz, 14.75 Hz to 852
5Hz, 60Hz is 17.7Hz ~ 102.
In the range of 3 Hz, 0.8≦A≦1, that is, the gain error is within +O4, -20%.

Therefore, even in the first embodiment described above, the error is reduced over a wider frequency band.

In each of the embodiments described above, the sampling period is set to 1 of the commercial period of the input AC signal. < n is a natural number) 4n, but if the sampling period is such that a close interval of the commercial period can be obtained, for example, '/AI%# of the commercial period.
This indicates that any number such as m#'A is acceptable. Also, the commercial frequency is not limited to 59Hz or 60Hz, but any number of Hz.
It is also applicable even if it is z.

Furthermore, as a function of digital filter processing, the equation is not limited to 2, but it is also possible to use an approximation of 2 transforms, such as It is also clear that the error of Furthermore, even if each function of these digital filter processes is multiplied by a constant K,
The gain is only doubled, and the value of k is not limited to the values shown in each of the above embodiments; other values may also be used.
It is clear that only the gain-frequency characteristics will be different.

Further, in each of the above embodiments, the amplitude value is adjusted. ,■e
""m"-n - "m""m-2n" is calculated in the form of a square root, but it is not limited to this, and at least the calculation "@ "m-n-Im" m-2n is used. Of course, it would be better if

〔Effect of the invention〕

As explained above, according to the present invention, when the sampling period of input AC is 17 (n is a natural number) and the transfer function of digital filter processing is 2 conversions,
Amplitude value! . -1-Ue□・lm-2n Since it is configured to perform arithmetic operations, it is possible to provide a digital protective relay that has small errors in a wide frequency band including commercial frequencies and is capable of determining the amplitude value of an AC signal. .

[Brief explanation of drawings]

Fig. 1 is a flowchart of an embodiment showing the procedure of digital filter processing and amplitude value calculation operation applied to a digital protective relay according to the present invention, Figs. 2 to 4 are dyne-frequency characteristic diagrams, and Fig. 5 9 is a flowchart showing the procedure of digital filter processing and amplitude value calculation calculation according to another embodiment of the present invention, FIGS. 6 to 8 are gain-frequency characteristic diagrams, and FIG.
The figure is a flowchart showing the procedure of conventional digital filter processing and amplitude value calculation calculation, and FIGS. 10 to 14 are gain-frequency characteristic diagrams. 1.1 person...Digital data input processing 2.2A, 2
B...Digital filter processing 3.3A...Amplitude value calculation calculation Patent applicant Toshiba Corporation Patent attorney Norio Ishii Figure 4 Figure 5 Section 7 Figure 8 Figure 14

Claims (3)

[Claims] (1) By inputting an alternating current signal such as a current or voltage, sampling the alternating current signal at a fixed period, and performing analog/digital conversion on the digital data obtained, perform digital filter processing and amplitude value calculation operation. In a digital protective relay that performs a protective operation by calculating the amplitude of an AC signal, the sampling period is 1/4n of the commercial cycle (reciprocal of the frequency) of the AC signal,
The transfer function of the digital filter processing is converted to K by Z transformation.
・(1+k)/(1-kZ^-^2^n) (However, K, k
is a constant function expressed as |K|<1) or an approximate function thereof, and the amplitude value calculation operation is performed when the data after digital processing is I_m (m is the sampling time series) and the amplitude value is I_e. , at least I_e^2=I_m
A digital protective relay characterized by using the calculation _-_n^2-I_m・I_m_-_2_n. (2) The transfer function of digital filter processing is K(1+kZ^-^2^n)/(1-k) (where K,
2. The digital protective relay according to claim 1, wherein k is a constant and is a function expressed as |k|<1. (3) The transfer function of digital filter processing is K (1+kZ^-^2^n+k^2Z^-^4^n) in Z transformation.
/(1-k+k^2) (where K and k are constants |k|<1
) The digital protective relay according to claim 1, wherein the digital protective relay is a function expressed as: