JP2856750B2 - AC electricity detection device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Purpose of the Invention] (Industrial application field) The present invention relates to a device for detecting an AC electric quantity, which detects a magnitude of a fundamental wave of the electric quantity from a sample value of the AC electric quantity.

(Prior Art) Conventionally, a voltage, a current, and the like detected for monitoring or controlling the state of an electric power system are an effective value, that is, an average value of the sum of squares of a fundamental wave component and each harmonic component, and is practically any value. There was no problem. However, in recent power system voltage stabilization problems, sophisticated VQC (Voltage Reactive Power Control) or S
There is a demand for VC (Static Reactive Power Compensation) and the like, and it is required to improve the accuracy of detecting the amount of electricity as the amount of attention. For this purpose, it is necessary to remove the harmonic component and correctly detect the fundamental wave, that is, the system frequency component. This is because harmonic components cannot always be effective energy. Also, from the viewpoint of cooperative control between electric stations only, it is necessary to exclude harmonic components, which are unstable elements, and pay attention to only system frequency components.

(Problems to be Solved by the Invention) In order to remove harmonic components, it goes without saying that an analog or digital filter is inserted. However
A filter that removes harmonics close to the fundamental wave, such as 2,3,5 times, causes a gain gradient even near the fundamental wave. Since the system frequency fluctuates within a certain range, simply inserting a filter causes a frequency error due to the gain gradient and does not achieve its purpose.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide an AC electric quantity detection device for detecting a magnitude of a fundamental wave of an AC electric quantity which does not cause an error due to a frequency near a fundamental wave frequency. And

[Constitution of the Invention] (Means for Solving the Problems) In the present invention, the slope of the gain near the frequency of the fundamental wave generated by the filter is calculated from the sampled value of the AC electric quantity or the digital data output from the filter. The correction is performed by an amount corresponding to the frequency of the fundamental wave.

(Operation) By inserting a filter necessary for removing harmonics and correcting the resulting gain slope near the frequency of the fundamental wave later, the magnitude of the fundamental wave of AC electric quantity is independent of frequency. Is to be detected.

(Example) Hereinafter, an example is described with reference to drawings.

FIG. 1 is a block diagram showing a configuration of an embodiment for explaining an AC electricity amount detecting device according to the present invention. In FIG. 1, 1 is an AD converter, 2 is a digital filter, 3 is amplitude calculating means, 4 is frequency corresponding amount calculating means, 5
Is a correction calculating means.

The AD converter 1 includes a sampling circuit, an analog-to-digital converter, a well-known pre-analog filter (not shown), and the like. Here, the sampling interval is typically an electrical angle of 30 ° at a commercial frequency of 50 Hz (or 60 Hz, hereinafter represented as 50 Hz). The subscript m is a number representing the sampling time according to a conventional method.

Digital filter 2, as illustrated as an example, F (Z) = K ( 1-Z -6) (1 + Z -2) (1 + 1.6Z -1 + Z -2) (1 + 1.8Z -1 + Z -2) (1) The sample value vm is filtered to be digital data um. Here, K is a normalization coefficient, for example, a constant such that the gain becomes 1 at a frequency of 50 Hz. The gain characteristics of this digital filter will be shown in the following figures.

Amplitude calculation means 3, a known amplitude squares calculation method, for example, as shown, _{^{i.e., U 2 = um -3 2 -um}} um -6 ... (2) calculation carried out in the amplitude squared U ² digital data um Get. Since this value includes the gain characteristic of the digital filter 2 with respect to the frequency and the gain characteristic due to the frequency shift of the equation (2) itself, it is referred to as an apparent amplitude squared. The gain characteristic of the equation (2) itself is as follows by a well-known simple calculation, assuming that the true amplitude square with respect to the digital data um is Uid ² and the rate of the frequency shift is δf / fo.

U ² = Uid ² cos ² (πδf / (2fo)) (3) In the frequency-corresponding-quantity calculating means 4, as shown in the drawing, fd = (um _-3 um _-12 -um um _{-15) ^{) / (um -3 2 -um um}} -6) ... (4) by determining the frequency corresponding amount fd corresponding to the frequency. This frequency corresponding amount fd is disclosed in JP-A-51-12166 or
As disclosed in -12167, fd = sin (2πδf / fo) / cos (πδf / (2fo)) ≒ 2πδf / fo (5), which is an amount substantially proportional to the frequency shift. However, in the present invention, equation (4) is used, but equation (5) itself is not used. The expression (5) is merely provided to explain the meaning of the frequency corresponding amount obtained by the expression (4). In the actual calculation, the denominator of the equation (4) can use the result of the equation (2) as it is, but is not particularly related to the gist of the present invention.

In the correction calculating means 5, the polynomial h of the frequency correspondence amount fd
(Fd) was used to correct the amplitude square U ² of apparent obtain amplitude square V ² of the input AC electric quantity v. This polynomial h (fd) is
For example, it is a quadratic or quaternary equation of the frequency correspondence amount fd. As will be described later, the gain characteristic with respect to the frequency shift is sufficiently corrected to obtain a practically error-free amplitude square value. .

FIG. 2 is a characteristic diagram for explaining the operation of FIG. Curve (a) shows the frequency versus gain characteristics of the digital filter 2. As is well known, this characteristic is expressed as follows, where n is the harmonic number.
In equation (1), z = e ^{j30 ゜ n} and | F
(E ^{j30 ゜ n} ) | As can be seen from the figure, this digital filter is capable of
It has a sharp characteristic that removes all the components up to the harmonic, so that a steep slope inevitably occurs near the fundamental wave.

The curve (b1) is an enlarged view of the vicinity of the harmonic order 1 of the curve (a), that is, the vicinity of the frequency of 50 Hz. Curve (b2) is overall characteristics of the amplitude calculation characteristics of the equation (3) curve (b1), the ratio between the amplitude squared U ² and the true amplitude square V ² of an apparent U ² / V ²
Represents the frequency characteristic of However, the curve (b2) takes the square root of the ratio U ² / V ² and is expressed in the dimension of the ratio U / V for easy understanding.

The curves (c1) and (c2) are obtained by converting the curve (b2) to the frequency corresponding amount fd.
Indicate the remaining errors resulting from the correction by the quadratic and quaternary equations, respectively. That is, V ² = U ² (1 + h (fd)) (6) h (fd) = A · fd + B · fd ² (in the case of quadratic expression) (7) h (fd) = A · fd + B · fd ² + C · fd ³ + D · fd ⁴ (in the case of the fourth order) The amplitude square V ² is obtained from (8). Here, the coefficients A, B, C or D are constants that allow the relationship between the apparent amplitude square U ² and the true amplitude square V ² to be approximated by the equation (6), and are determined by various known methods. Can be. That is, the polynomial h (fd) is set as in the equation (7) or (8), and a value satisfying these equations at a specific frequency is obtained. For example, in the case of the second order, the coefficients A and B are determined using the polynomial h (fd) based on the equation (7) so that the equation (6) is exactly satisfied at two frequency points. Alternatively, the coefficient that best approximates the equation (6) at three or more frequencies may be determined by the well-known least square method. This applies even to a quartic or other order.

Note that the apparent amplitude square U ² itself is an unknown quantity calculated from the input, but the ratio U / V is a known relationship determined from the equations (2) and (3) as shown by the curve (b2). Yes, the coefficient can be determined in advance regardless of the actual input.

As can be seen from the curve (c1) or (c2), the residual error after correction by the quadratic equation is about 0.05%,
It is sufficiently smaller than 0.001%.

As described above, according to the present embodiment, a harmonic filter is sufficiently removed by inserting a sharp filter without being restricted by a gain gradient occurring near the fundamental wave. The magnitude of the AC electric quantity can be detected with high accuracy by performing the correction using the polynomial of the corresponding quantity.

FIG. 3 is a block diagram showing the configuration of another embodiment of the present invention.

This embodiment differs from FIG. 1 only in that the frequency correspondence amount calculation means 6 is used in place of the frequency correspondence amount calculation means 4 in FIG. 1, and the input is a sample value vm. That is, the calculation of the frequency correspondence amount is based on the difference between the value before the digital filter and the data after the filter, and it can be easily understood that the same principle as that in FIG. 1 holds. I do. This embodiment is suitable for a case where the distortion is relatively small and the influence on the final result is secondary even if the frequency correspondence amount is calculated based on the value before the filter. In this case, the delay in the calculation of the frequency correspondence amount can be reduced by the amount of the digital filter.
Response can be made faster.

FIG. 4 is a block diagram showing the configuration of another embodiment of the present invention.

This embodiment is different from FIG. 1 in that the digital filter 2 of FIG. 1 is divided into digital filters 7 and 8, and the input of the frequency correspondence amount calculating means 9 is the output wm of the digital filter 7. Functional form of the digital filter 7 and 8 corresponds to the second half and the first half of the functional form of the digital filter 2 of the respective first view, also K ₁ and K ₂ is a normalization factor similar to the respective foregoing. Due to the cascade of the digital filters 7 and 8, the output of the digital filter 8 becomes digital data um similar to the case of FIG. The algorithm of the frequency correspondence amount calculating means 9 is the same as that of FIG. 1 except that the input is the output wm of the digital filter 7, and it is clear that the same principle as that of FIG. 1 holds. This embodiment has intermediate features between the case of FIG. 1 and the case of FIG. That is, the function form of the digital filter 7 removes the components near the fifth harmonic and the seventh harmonic. The components around this are relatively large, and only the frequency corresponding amount is removed before calculation. It is suitable for the case.

FIG. 5 is a block diagram showing the configuration of another embodiment of the present invention.

This embodiment is the same as FIG. 1 except that the correction calculating means 10 is located before the amplitude calculating means 11. That is, the data xm obtained by correcting the digital data um by the frequency corresponding amount fd in the correction calculating means 10 is applied to the amplitude calculating means 11 to calculate the amplitude. Since the correction calculation means 10 calculates not the amplitude square dimension but the amplitude dimension, the correction polynomial h ₁ (fd)
Uses the expression converted to the amplitude dimension. In this case as well, it can be easily understood that the same principle as in FIG. This embodiment has no particular advantage over the other embodiments, but shows that such variations are possible.

FIG. 6 is a block diagram showing the configuration of another embodiment of the present invention.

This embodiment in the correction calculation unit 15, upon to correct the amplitude square U ² of the apparent point of selecting a correction term H by the state of the frequency corresponding amount fd differs from the embodiment of Figure 1. That is, the change of the frequency corresponding amount fd is determined by the change determining means 12, and if the change is within a predetermined range, the correction term H is equivalent to the polynomial h (fd) as described above by the correction term updating means 13. The value is applied to the correction calculation means 15. If the change is out of the predetermined range, the correction term H is applied to the correction calculation means 15 by the correction term keeping means 14.
The correction term installation means 14 does not need to be provided as a single block, but is inserted in order to make the explanation of the function easy to understand. This embodiment is effective for avoiding the influence of an error in the frequency correspondence amount due to a sudden change in phase due to the occurrence of a system fault, and the embodiment of FIGS. 3 to 5 can be further modified with the same gist. .

FIG. 7 is a block diagram showing the configuration of another embodiment of the present invention. Reference numeral 16 in the figure denotes frequency corresponding amount calculation means. In the embodiments described above, the frequency corresponding amount is calculated from the sample value of the AC electric quantity whose magnitude is to be detected or the digitally filtered data. In this embodiment, the frequency corresponding amount fd is calculated from a sample value ym of another electric quantity having the same fundamental frequency as the AC electric quantity whose magnitude is to be detected. Alternatively, the same applies to data obtained by digitally filtering a sample value of another electric quantity as ym.
It is easy to understand that each of the above embodiments can be modified in accordance with the gist, and thus details are omitted. This embodiment is effective in a case where the frequency corresponding amounts are not individually calculated for a plurality of input amounts belonging to the same electric system, but are calculated using a representative input amount.

Although not particularly shown in the above embodiments, various modifications as listed below are conceivable.

As the function form of the digital filter, various well-known function forms other than those illustrated can be used. Also, for purposes of simplicity of explanation, the normalization factor K, but using K ₁ or K _2, etc., if not normalized, or even when using other coefficients by circumstances such as the scale of the calculation, the same gist Can of course be applied.

Although the apparent amplitude square is calculated by the equation (2), various well-known calculation methods such as a root-mean-square method can be used. Further, other than the amplitude square, the amplitude, the square of the effective value, the effective value, or the like may be obtained. This is because the present invention aims at eliminating harmonics as one of the main objectives, and they are simply a matter of conversion.

As for the calculation of the frequency corresponding amount, various methods such as those disclosed in JP-A-51-12166 or JP-A-51-12167 can be used.

Usually, a pre-analog filter is arranged before the sampling, but it can be corrected including its frequency gain characteristic. That is, the characteristics of the pre-analog filter are known, and this is exactly the same as the total frequency gain characteristics in addition to the digital filter and the amplitude calculation characteristics.

[Effects of the Invention] As described above, according to the present invention, a filter necessary for removing harmonics is inserted without being restricted by the gain gradient near the fundamental frequency, and the harmonics are sufficiently removed. Since the slope of the gain near the fundamental frequency of the filter is corrected, the magnitude of the fundamental wave of the AC electric quantity can be detected with high accuracy without being affected by distortion or frequency.

[Brief description of the drawings]

FIG. 1 is a block diagram showing the configuration of one embodiment of the present invention,
FIG. 2 is a characteristic diagram for explaining the operation of FIG. 1, and FIGS.
FIG. 13 is a block diagram showing the configuration of another embodiment of the present invention. 1 AD converter 2,7,8 Digital filter 3,11 Amplitude calculating means 4,6,9,16 Frequency corresponding amount calculating means 5,10,15 Correction calculating means 12 Change determination means, 13: Correction term updating means 14: Correction term keeping means

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Mitsuru Yamaura 1 Toshiba-cho, Fuchu-shi, Tokyo Inside the Toshiba Fuchu factory (56) References JP-A-63-224617 (JP, A) JP-A-61-221515 (JP, A) JP-A-62-285513 (JP, A) JP-A-61-157223 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) G01R 19/00-19 / 32 H02H 3/00-3/02

Claims (1)

(57) [Claims] An A / D converter for inputting an input amount of AC electricity and converting it into a sample value, and filtering the A / D-converted sample value to form a low-order signal having even and odd orders close to a fundamental wave. A digital filter for removing the next harmonic, and an amplitude for calculating an amplitude value including a gain characteristic of the digital filter and a gain characteristic due to a frequency shift included in the amplitude square method by an amplitude square method using data from the digital filter. Calculating means; frequency-corresponding-quantity calculating means for calculating a frequency-corresponding amount which is also substantially proportional to the frequency shift using data from the digital filter; And a correction calculating means for correcting using a polynomial of a frequency correspondence amount by the means.