KR100823832B1 - Apparatus and method for estimating frequency of signal in power system - Google Patents

Abstract

A frequency estimating apparatus and method of a power system are provided to estimate precise frequency regardless of the change of a signal by obtaining three sample values by sampling cosine and sine signals generated through an orthogonal filter at regular intervals. A frequency estimating apparatus(100) of a power system is composed of an A/D(Analog To Digital) conversion unit(130) receiving a current or voltage value detected from the power system and converting the current or voltage value into a digital signal; a cosine filter(161) generating a cosine signal(Xr[n]) of a real-number axis from the digital signal converted in the A/D conversion unit; a sine filter(165) generating a sine signal(Xi[n]) of an imaginary axis from the digital signal converted in the A/D conversion unit; and a frequency estimating unit(170) extracting three sample values by sampling the orthogonal signals generated from the cosine and sine filters at regular intervals. The frequency estimating unit obtains a compensation coefficient of the third sample value sampled finally, calculates a sample compensation value by computing the third sample value and the compensation coefficient together, and estimates frequency of a digital signal by using the sample compensation value and the first sample value sampled first.

Description

Frequency estimating apparatus and method for power system {APPARATUS AND METHOD FOR ESTIMATING FREQUENCY OF SIGNAL IN POWER SYSTEM}

1A and 1B are diagrams for explaining a frequency estimation method using a conventional zero crossing method.

2 is a diagram illustrating a problem occurring in frequency estimation using a conventional zero crossing method.

3 is a block diagram illustrating a frequency estimating apparatus according to an exemplary embodiment of the present invention.

FIG. 4 is a diagram illustrating the magnitude of frequency response characteristics of the cosine filter and the sine filter shown in FIG. 3.

FIG. 5 is a diagram illustrating a trajectory of signals output from the cosine filter and the sine filter illustrated in FIG. 3 on a complex plane.

FIG. 6 is a diagram illustrating a signal trajectory of a plurality of sample values extracted through a cosine filter and a sine filter according to the present invention.

7 is a flowchart schematically illustrating a frequency estimation method according to the present invention.

8 is a view showing a frequency estimation simulation of the power system according to the present invention.

9 is a diagram illustrating an example in which the frequency estimating apparatus according to the present invention is applied to a relay.

10 is a view showing an example in which the frequency estimating apparatus according to the present invention is applied to a pager measuring device (PMU).

* Explanation of symbols for the main parts of the drawings

100: frequency estimator 110: anti-aliasing filter

130: A / D conversion unit 150: microprocessor

160: orthogonal filter 161: cosine filter

165: sine filter 170: frequency estimation

X1: first sample value X0: second sample value

X2: Third Sample Value X2c: Sample Compensation Value for Third Sample Value

The present invention relates to a frequency estimating apparatus and a method thereof, and more particularly, to estimate a frequency using only a sample value of a voltage or current signal of a power system, and to accurately estimate a frequency regardless of a change in signal size. The present invention relates to a frequency estimating apparatus and a method thereof.

In general, power monitors such as relays or phaser measurement units (PMUs) include a frequency estimator to estimate a frequency of a voltage or a current input from a power supply. In the protection relay, the frequency estimating apparatus is used to determine the system's failure, stability, and failure, and the pager measuring device uses the measured frequency information as basic data for monitoring the system's status and operating a stable power system.

The conventional frequency estimator estimates the frequency of a signal input by a zero-crossing method. In a conventional frequency estimation method using a zero crossing method, as illustrated in FIG. 1A, time points T1 and T2 that cross a zero point in a signal waveform of an input voltage or current are detected, and the detected zero point crossing time ( A method of calculating the frequency by calculating the interval between T1 and T2, and sampling the input signal waveform at regular intervals (Tn-1, Tn) as shown in FIG. 1B, and then sampling the two points (Tn-1). After assuming that the waveform between Tn) is linear, there is a method of calculating a zero point intersection point Tzc using an equation of a straight line.

However, the frequency estimation method using the zero crossing method has a problem that a large error occurs between the calculated frequency and the actual frequency because it is difficult to detect the exact zero crossing point. In order to reduce such an error, a quadratic or cubic interpolation method may be used, which assumes a waveform between the sampled two points Tn-1 and Tn as a quadratic function or a cubic function. However, when the second or third order interpolation method is used, the amount of calculation increases exponentially as the order is increased. This increase in computation amount puts a heavy burden on the frequency estimator.

In addition, the zero crossing method has a structure that is very vulnerable to the input noise contained in the original signal. That is, when a signal containing noise is applied, as illustrated in FIG. 2, since several zero crossing points Tzc1, Tzc2, and Tzc3 may be generated, it is difficult to detect zero crossing points.

An object of the present invention is to convert the voltage or current signal detected by the power system into a digital signal and input it to the microprocessor, and to estimate the frequency using a frequency estimation algorithm built in the microprocessor, the correct frequency only in software without additional hardware An apparatus and method for estimating a frequency of a power system capable of estimating is provided.

Another object of the present invention is to convert the voltage or current signal detected by the power system into a digital signal and input to the orthogonal filter of the microprocessor, and sample the cosine and sine signal generated by the orthogonal filter at regular intervals three samples By acquiring a value and estimating the frequency of the signal input to the quadrature filter using three sample values, the frequency estimation apparatus of the power system capable of accurately estimating the frequency regardless of the change in the signal magnitude and a method thereof are provided. There is.

Technical means of the present invention for achieving the above object, A / D conversion unit for receiving a current or voltage value detected from the power system to convert into a digital signal; A cosine filter for generating a cosine signal Xr [n] of a real axis component from the digital signal x [n] converted by the A / D converter; A sine filter for generating a sine signal Xi [n] of a imaginary axis component from the digital signal x [n] converted by the A / D converter; And extracting three sample values by sampling the quadrature signals Xr [n] and Xi [n] generated by the cosine filter and the sine filter at predetermined time intervals, and finally extracting the third sample values ( After obtaining the compensation coefficient k of X2) and calculating with the third sample value X2, the sample compensation value X2c is obtained, and the sample compensation value X2c and the first sampled first sample value X1 are obtained. And a frequency estimator for estimating a frequency f for the digital signal using the frequency estimator.

In detail, the three sample values of the frequency estimation unit include a first sample value X1 obtained by first sampling a quadrature signal generated by the cosine and a sine filter, and a time point at which the first sample value is 1/4 rotated. And a second sample value X0 sampled within the range, and a third sample value X2 sampled at the time when the first sample value is rotated n times.

In addition, the compensation coefficient k for the third sample value X2 of the frequency estimation may be obtained by using Equation 1 below.

Equation 1

이고, b는

이고, Xr[n1]은 첫번째로 샘플링된 코사인필터의 출력값이고, Xi[n1]은 첫번째로 샘플링된 사인필터의 출력값이고, Xr[n0]은 두번째로 샘플링된 코사인필터의 출력값이고, Xi[n0]은 두번째로 샘플링된 사인필터의 출력값이고, Xr[n2]는 세번째로 샘플링된 코사인필터의 출력값이고, Xi[n2]는 세번째로 샘플링된 사인필터의 출력값임.Where a is

And b is

Xr [n1] is the output value of the first sampled cosine filter, Xi [n1] is the output value of the first sampled cosine filter, Xr [n0] is the output value of the second sampled cosine filter, and Xi [n0 ] Is the output value of the second sampled sine filter, Xr [n2] is the output value of the third sampled cosine filter, and Xi [n2] is the output value of the third sampled sine filter.

The frequency of the digital signal of the frequency estimation is calculated by the following Equation 1 or 2 according to the phase difference between the first sample value X1 and the sample compensation value X2c.

Equation 1

,

,

임.Where α is

being.

Equation 2

,

,

임.Where α is

being.

In Equations 1 and 2, n1 is the first sampling time, n2 is the third sampling time, Xr [n1] is the output value of the first sampled cosine filter, and Xi [n1] is the first sampled sine filter. Xr [n2] is the output value of the third sampled cosine filter, Xi [n2] is the output value of the third sampled sine filter, Xcr [n2] is the compensation value of Xr [n2], and Xci [ n2] is the compensation value of Xi [n2], N is the number of samples per period, k is the compensation factor, and T is the phase difference between the two sample values.

Technical method of the present invention for achieving the above object, the first step of receiving a current or voltage value detected from the power system to convert into a digital signal; A second step of detecting and outputting a cosine signal (Xr [n]) and a sine signal (Xi [n]), which are orthogonal components, from the digital signal (x [n]) by using a cosine filter and a sine filter; A frequency estimating step of extracting three sample values (X1, X0, X2) by sampling the quadrature signals output from the cosine filter and the sine filter at predetermined time intervals; A fourth step of obtaining a sample compensation value (X2c) by obtaining a compensation coefficient (k) of the third sample value (X2) extracted last among the extracted three sample values and calculating a compensation coefficient on the third sample value; And a fifth step of estimating the frequency of the digital signal x [n] using the sample compensation value X2c for the first sample value X1 and the third sample value obtained above. It features.

Specifically, when extracting the three sample values in the third step, the frequency estimator is as close as possible to the time when the first sample value (X1) to the second sampling value (X0) to sample the second sampling frequency based on the sampling frequency And sampling the third sample value X2 to be sampled at a time point having a large time interval with respect to the time point at which the first sample value X1 is extracted.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

3 is a schematic block diagram illustrating an apparatus for estimating a frequency of a power system according to an exemplary embodiment of the present invention, wherein the apparatus for estimating frequency 100 includes an anti-aliasing filter 110, an A / D converter 130, and a microprocessor. 150, an orthogonal filter 160, and a frequency estimator 170.

The anti-aliasing filter 110 removes aliasing included in signals output from voltage detectors PT and current detectors CT for detecting voltage and current of a power system. Filter.

The A / D converter 130 receives an analog voltage or current signal input from the anti-aliasing filter 110, converts the signal into a digital signal, and outputs the digital signal to the microprocessor 150.

The quadrature filter 160 of the microprocessor 150 separates and outputs the digital signal input from the A / D converter 130 into two signals having orthogonal components. The quadrature filter 160 includes a cosine filter 161 and a sine filter 165.

The cosine filter 161 extracts and outputs a cosine signal that is a real axis component from the digital signal converted by the A / D converter 130.

The sine filter 165 extracts and outputs a sine signal that is an imaginary axis component from the digital signal converted by the A / D converter 130. The signals output from the cosine filter 161 and the sine filter 165 have a phase difference of 90 °.

The cosine signal Xr [n] of the real axis output from the cosine filter 161 and the sine signal Xi [n] of the imaginary axis output from the sine filter 165 are finite impulse responses as shown in Equation 1 below. A filter having a (FIR; Finite Impulse Response) characteristic.

Where x is the input signal and N is the number of samples per period of the filter.

The signals Xr [n] and Xi [n] respectively output from the cosine filter 161 and the sine filter 165 have the magnitudes of the frequency response characteristics as shown in FIG. 4. That is, when the frequency is 60 Hz, which is the fundamental frequency, the frequency response characteristics of the cosine filter 161 and the sine filter 165 are equal to 1, but the magnitude of the frequency response characteristics varies depending on the frequency. For example, when the frequency is smaller than the fundamental frequency, the magnitude of the frequency response characteristic of the sine filter 165 is larger than that of the cosine filter 161, whereas when the frequency is larger than the fundamental frequency, the frequency response of the cosine filter 161 is larger. The magnitude of the characteristic is larger than that of the sine filter 165.

Hereinafter, it is assumed that the signal x [n] input to the cosine filter 161 and the sine filter 165 is a cosine signal having a frequency f as shown in Equation 2 and a phase angle φ.

Here, A is the magnitude of the input signal and N is the number of samples per period.

When the signal shown in Equation 2 is input to the cosine filter 161 and the sine filter 165, the signals Xr [n] and Xi [n] respectively output from the cosine filter 161 and the sine filter 165. Is as shown in Equation 3 below.

Here, Hc (f) is the magnitude of the frequency response characteristic of the cosine filter, Hs (f) is the magnitude of the frequency response characteristic of the sine filter, φc is the phase angle delay occurring in the cosine filter, and φs is generated in the sine filter. Phase angle delay.

Since signals Xr [n] and Xi [n] output from the cosine filter 161 and the sine filter 165 respectively have a phase difference of 90 °, Equation 4 below is established.

By substituting Equation 4 into Equation 3, it can be represented by a cosine function and a sine function as shown in Equation 5 below.

Here, Hc (f) is the magnitude of the frequency response characteristic of the cosine filter, Hs (f) is the magnitude of the frequency response characteristic of the sine filter, A is the magnitude of the input signal, and N is the number of samples per period.

The internal variables of the cosine function and the sine function are Ω, and the locus of the signals Xr [n] and Xi [n] output from the cosine filter 161 and the sine filter 165, respectively, is shown on a complex plane. Same as 5. As shown in FIG. 5, the point where the locus meets the real axis Re and the imaginary axis Im is the magnitude of the frequency response characteristic of the cosine filter 161 (Hc (f)) and the frequency of the sine filter 165. It depends on the magnitude of the response characteristic (Hs (f)).

In addition, the magnitude (Hc (f)) of the frequency response characteristic of the cosine filter 161 and the magnitude (Hs (f)) of the frequency response characteristic of the sine filter 165 vary depending on the frequency. The trajectory also changes. For example, the signals Xr [n] and Xi [n] output from the cosine filter 161 and the sine filter 165 respectively draw a circle trajectory on a complex plane when the frequency is the fundamental frequency (60 Hz). If the frequency is not the fundamental frequency (60 Hz), draw the trajectory of the ellipse.

On the other hand, the frequency estimator 170 has three sample values (X1, X1,) at predetermined time intervals in the signals Xr [n] and Xi [n] generated by the cosine filter 161 and the sine filter 165, respectively. X0, X2) is extracted. 6 shows the first sample value X1, the second sample value X0, and the third sample value X2, which are the three sample values sampled in this manner, on the complex plane, respectively. Here, the first sample value (X1) is the first sampled value, the second sample value (X0) is the value sampled when the first sample value (X1) is a quarter of a turn, and the third sample value ( X2) represents a value sampled at the time when the first sample value X1 is rotated three times.

The time point at which the three sample values X1, X0, and X2 are extracted is only an embodiment, and the frequency estimator 170 uses the second sample value X0 as the first sample value X1 based on the sampling frequency. It is preferable to sample as close as possible to the time point at which) is extracted, and the third sample value X2 is as close as possible to the first sample value X1 based on the time point at which the first sample value X1 is sampled. It is preferable to sample so that it may become large.

The frequency estimator 170 uses the three sample values X1, X0, and X2 sampled as described above, regardless of the magnitude of the signal output from the cosine filter 161 and the sine filter 165. The frequency can be estimated.

In addition, the quadrature filter 160 and the frequency estimator 170 are frequency estimation algorithms built in the microprocessor 150 and estimate the frequency by software embedded in the microprocessor 150.

Referring to FIG. 6, when the first sample value X1 sampled by the frequency estimator 170 rotates to the third sample value X2 sampled the third time, the frequency does not change and the signal size changes. The third sample value X2 may be located at a different trajectory from the first sample value X1. Here, if the frequency is estimated using only the first sample value X1 and the third sample value X2, an error occurs because there is a change in the magnitude of the signal. Therefore, in the present invention, in order to reduce the error, the third sample value X2 is compensated to obtain the sample compensation value X2c, and then the frequency is estimated using the first sample value X1 and the sample compensation value X2c. Done.

First, an ellipse equation of the outermost ellipse is obtained using the first sample value X1 and the second sample value X0.

Where Hc (f) is the magnitude of the frequency response characteristic of the cosine filter, Hs (f) is the magnitude of the frequency response characteristic of the sine filter, A is the magnitude of the input signal, and Xr [n] is the nth sampling. Is the output value of the cosine filter, and Xi [n] is the output value of the nth sampled sine filter.

If Equation 6 is represented by a and b, it can be summarized as Equation 7 below.

Where a and b are as follows.

When the coordinates of the first sample value X1 and the second sample value X0 are input to the elliptic equation of Equation 6 and arranged into a system of simultaneous equations, Equation 8 below.

Where Xr [n1] is the output value of the first sampled cosine filter, Xi [n1] is the output value of the first sampled cosine filter, Xr [n0] is the output value of the second sampled cosine filter, and Xi [n0 ] Is the output of the second sampled sine filter.

The system of equations (8) is equal to Equation (9) below for a and b.

The ellipse with the first sample value (X1) and the ellipse with the third sample value (X2) have the same frequency, so only the size of the ellipse is different, and the ratio of the point where the real and imaginary axes meet is the same. Therefore, the third sample value X2 in the ellipse of the first sample value X1 and the sample compensation value X2c with respect to the third sample value have a proportional relationship as shown in Equation 10 below.

Where k is the compensation coefficient of the third sample value (X2).

Equation 10 is divided into a real number and an imaginary part to be expressed as Equation 11 below.

When the real axis sample compensation value X2cr and the imaginary axis sample compensation value X2ci of the third sample value X2 of Equation 11 are substituted into the elliptic equation of Equation 7, the following Equation 12 is obtained.

When the compensation coefficient k of the third sample value X2 is obtained from Equation 12, Equation 13 is obtained.

이고, b는

이고, Xr[n1]은 첫번째로 샘플링된 코사인필터의 출력값이고, Xi[n1]은 첫번째로 샘플링된 사인필터의 출력값이고, Xr[n0]은 두번째로 샘플링된 코사인필터의 출력값이고, Xi[n0]은 두번째로 샘플링된 사인필터의 출력값이고, Xr[n2]는 세번째로 샘플링된 코사인필터의 출력값이고, Xi[n2]는 세번째로 샘플링된 사인필터의 출력값임.Where a is

And b is

Xr [n1] is the output value of the first sampled cosine filter, Xi [n1] is the output value of the first sampled cosine filter, Xr [n0] is the output value of the second sampled cosine filter, and Xi [n0 ] Is the output value of the second sampled sine filter, Xr [n2] is the output value of the third sampled cosine filter, and Xi [n2] is the output value of the third sampled sine filter.

The compensation coefficient k obtained in Equation 12 is multiplied by the third sample value X2 to obtain a sample compensation value X2c, and the sample compensation value X2c and the first sample value X1 obtained as described above are converted into a frequency estimation equation. Substituting and arranging, it is shown in Equations 14 and 15 below.

,

,

이다.Where α is

to be.

,

,

이다.Where α is

to be.

In Equations 14 and 15, n1 is the first sampling time, n2 is the third sampling time, Xr [n1] is the output value of the first sampled cosine filter, and Xi [n1] is the first sampled sine. Xr [n2] is the output value of the third sampled cosine filter, Xi [n2] is the output value of the third sampled sine filter, Xcr [n2] is the compensation value of Xr [n2], and Xci. [n2] is the compensation value of Xi [n2], N is the number of samples per period, k is the compensation factor, and T is the phase difference between two sample values.

Since the cos ^-1 α value is repeated with a certain periodicity, it is necessary to find out which range the cos ^-1 α value belongs to. The range of the cos ^-1 α value can be determined by the phase difference T between the two sample values X1 and X2c. In addition, since the two sample values X1 and X2c are values taken by the frequency estimator 170, the frequency estimation unit 170 estimates the phase difference between the two sample values X1 and X2c through estimation. T) can be known. If the two phase difference (T) is determined in the sample values (X1, X2c), the value k is determined so that the ground cos ^-1 α value can be seen that for any of the range of the cosine function, the frequency (f) Can be calculated

On the other hand, in the case of increasing the interval between the two sample values (X1, X2c) it can reduce the error that occurs during the frequency estimation. Here, the error occurring during the frequency estimation is an error according to the calculation of the elements used for frequency estimation. In Equation 14 or 15, many errors occur when calculating the 'cos ^-1 α' value. The error according to the 'cos ⁻¹ α' value decreases as the interval between the two sample values X1 and X2c is increased. Therefore, the larger the interval between the two sample values X1 and X2c, the greater the accuracy in frequency estimation.

However, if the interval between the two sample values X1 and X2c is increased, the frequency estimation speed is decreased. Therefore, an appropriate sampling period should be set in consideration of the frequency estimation density and the estimation speed.

7 is a flowchart schematically illustrating a frequency estimation method according to the present invention. First, a voltage detector PT or a current detector CT detects a voltage and a current from a power system to prevent an anti-aliasing filter 110 that is an analog signal processor. The A / D converter 130 outputs the digital signal to the A / D converter 130, and converts the current or voltage value into a digital signal and outputs the digital signal to the microprocessor 150 (S1).

The orthogonal filter 160 of the microprocessor 150 receives the digital signal x [n] output from the A / D converter 130 and passes through the cosine filter 161 and the sine filter 165. The quadrature real axis cosine signal Xr [n] and the imaginary axis sine signal Xi [n] are generated as shown in Equation 5, and the generated quadrature signals Xr [n] and Xi [n ]) Is output to the frequency estimator 170 (S2).

The frequency estimator 170 extracts three sample values X1, X0, and X2 at regular time intervals from the quadrature signals Xr [n] and Xi [n] output from the quadrature filter 160. The extracted first sample value X1, the second sample value X0, and the third sample value X2 are temporarily stored (S3). 6 shows the extracted three sample values X1, X0, and X2 on the complex plane. Here, when the three sample values are extracted, the second sample value X0 to be sampled secondly is preferably extracted as close as possible to the time point at which the first sample value X1 is extracted based on the sampling frequency. Sampling the third sample value X2 at a time point at which the time interval is as large as possible based on the time point at which the first sample value X1 is extracted can reduce the frequency estimation error.

Using the three sample values X1, X0, and X2 extracted as described above, a compensation coefficient k is obtained using Equation 13 above (S4), and the frequency estimator 170 obtains the compensation coefficient k and the third sample. By multiplying the value X2, the sample compensation value X2c for the third sample value is obtained (S5).

The frequency estimator 170 substitutes the sample compensation value X2c and the first sampled value X1 sampled as described above into a frequency estimation equation such as Equation 14 or Equation 15 above to calculate the frequency (f). In operation S6, the frequency of the signal x [n] output from the A / D converter 130 is estimated.

에 속할 경우에 적용하는 주파수 추정식이고, 수학식 15의 경우에는 두 개의 샘플값(X1, X2c)의 위상차(T)가

에 속할 경우 에 적용하는 주파수 추정식이다.In the case of Equation 14, the phase difference T between two sample values X1 and X2c is

Is a frequency estimation equation to be applied to the equation, and in the case of Equation 15, the phase difference T between two sample values X1 and X2c is

It is a frequency estimation formula applied to when.

8 is a diagram illustrating a frequency estimation simulation of a power system according to the present invention, where a is a frequency of an actual voltage in a power system, and b is a first sample value (X1) and a third sample value (X2). It shows the case of estimating using only, and c shows the case of estimating the actual frequency using the frequency estimation algorithm according to the present invention.

As shown, the actual frequency a of the power system returns to the original frequency after the frequency suddenly decreases due to a sudden load change.

Herein, in the case of c to which the frequency estimation method of the present invention is applied, the actual frequency can be accurately estimated without discontinuous change even in the portion where the amount of change in the voltage magnitude is the largest, but the first sample value X1 and the third sample value ( In the case of estimating the frequency by X3) alone, it can be seen that some errors occur in the frequency estimation due to the discontinuous change in the hourly variation of the voltage magnitude, such as the signal waveform of b.

Accordingly, three sample values (X1, X0, X2) are extracted, and a sample compensation value (X2c) for the third sample value (X2) extracted last is obtained, and the sample compensation value (X2c) and the first sample value are obtained. It is desirable to estimate the frequency using (X1).

9 is a view showing an example in which the frequency estimating apparatus 100 according to the present invention is applied to the relay 200. As shown in FIG. 9, the analog signal processing unit 100 and the analog channel 110 and the A / D converter 130 are shown. The input voltage signal is input to the quadrature filter 160 and the frequency estimator 170 of the microprocessor 150 in which the frequency estimation algorithm of the present invention is incorporated, and the frequency estimated by the frequency estimator 170 is a relay element. (E.g., Underfrequency, Overfrequency, df / dt, etc.) is used as the default operation value for the operation. It can also be applied to overexcited V / F.

10 is a view showing an example in which the frequency estimating apparatus 100 according to the present invention is applied to a pager measuring apparatus 300 (PMU). As shown, the voltage or current signals input through the analog channel 110 and the A / D converter 130, which are analog signal processing units, are orthogonal to the microprocessor 150 incorporating a frequency estimation algorithm, which is a core component of the present invention. Input to the filter 160 and the frequency estimator 170, the frequency measured by the frequency estimator 170 and the phaser measured by the pager measurer 350 is transmitted to the host system through a communication line.

The estimated frequency may be used to compensate for the measured phaser.

Although the present invention has been described in detail with reference to exemplary embodiments above, those skilled in the art to which the present invention pertains can make various modifications to the above-described embodiments without departing from the scope of the present invention. I will understand. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined by the claims below and equivalents thereof.

As described above, in the present invention, the voltage or current signal detected by the power system is converted into a digital signal and input to the microprocessor, and the frequency is estimated using a frequency estimation algorithm built in the microprocessor, so that no additional hardware is required. The software alone has the advantage of allowing accurate frequency estimation.

In addition, the cosine and sine signals generated by the orthogonal filter are sampled at predetermined time intervals to obtain three sample values, and then the frequency of the signal input to the orthogonal filter is estimated using the three sample values. There is an advantage in that accurate frequency estimation is possible regardless of the change.

In addition, by estimating the frequency using the signal passing through the orthogonal filter, the frequency estimation error caused by the noise included in the signal can be reduced, and the frequency estimation when the interval of taking the sample values passed through the orthogonal filter is changed. Since the precision and the estimated speed can be changed, there is an advantage of optimizing the frequency estimation characteristics according to the design conditions.

Claims (11)

An A / D conversion unit receiving a current or voltage value detected from the power system and converting the digital signal into a digital signal; A cosine filter for generating a cosine signal Xr [n] of a real axis component from the digital signal x [n] converted by the A / D converter; A sine filter for generating a sine signal Xi [n] of a imaginary axis component from the digital signal x [n] converted by the A / D converter; And Samples of the quadrature signals Xr [n] and Xi [n] generated by the cosine filter and the sine filter are sampled at predetermined time intervals, and three sample values are extracted. After obtaining the compensation coefficient (k) of k) and computed with the third sample value (X2) to obtain a sample compensation value (X2c), using the sample compensation value (X2c) and the first sampled value (X1) first sampled And a frequency estimator for estimating the frequency (f) of the digital signal. The method according to claim 1, The three sample values extracted by the frequency estimator include a first sample value X1 obtained by first sampling a quadrature signal generated by the cosine and a sine filter, and within a time point at which the first sample value is rotated quarterly. And a third sample value (X2) sampled at the time when the sampled second sample value (X0) and the first sample value are rotated n (n is a positive integer greater than 1 in revolutions). Frequency estimation device of the power system. The method according to claim 1 or 2, Compensation coefficient k for the third sample value (X2) of the frequency estimation is a power system frequency estimation apparatus, characterized in that it is obtained using the following equation (1). Equation 1

Where a is

And b is

Xr [n1] is the output value of the first sampled cosine filter, Xi [n1] is the output value of the first sampled cosine filter, Xr [n0] is the output value of the second sampled cosine filter, and Xi [n0 ] Is the output value of the second sampled sine filter, Xr [n2] is the output value of the third sampled cosine filter, and Xi [n2] is the output value of the third sampled sine filter. The method according to claim 1 or 2, The frequency of the digital signal of the frequency estimation is calculated by the following equation (1) or (2) according to the phase difference between the first sample value (X1) and the sample compensation value (X2c). Device. Equation 1

,

Where α is

being. Equation 2

,

Where α is

being. In Equations 1 and 2, n1 is the first sampling time, n2 is the third sampling time, Xr [n1] is the output value of the first sampled cosine filter, and Xi [n1] is the first sampled sine filter. Xr [n2] is the output value of the third sampled cosine filter, Xi [n2] is the output value of the third sampled sine filter, Xcr [n2] is the compensation value of Xr [n2], and Xci [ n2] is the compensation value of Xi [n2], N is the number of samples per period, k is the compensation factor, and T is the phase difference between the two sample values. A first step of receiving a current or voltage value detected from the power system and converting the value into a digital signal; A second step of generating and outputting a cosine signal (Xr [n]) and a sine signal (Xi [n]), which are orthogonal components, from the digital signal (x [n]) by using a cosine filter and a sine filter; A frequency estimating step of extracting three sample values (X1, X0, X2) by sampling the quadrature signals output from the cosine filter and the sine filter at predetermined time intervals; A fourth step of obtaining a sample compensation value (X2c) by obtaining a compensation coefficient (k) of the third sample value (X2) extracted last among the extracted three sample values and calculating a compensation coefficient on the third sample value; And And a fifth step of estimating a frequency of the digital signal x [n] by using the first sample value X1 and the sample compensation value X2c for the third sample value obtained above. Frequency estimation method of power system. The method according to claim 5, When extracting the three sample values in the third step, the frequency estimator extracts the second sample value X0 sampled for the second time from the time point at which the first sample value X1 is extracted based on the sampling frequency. A frequency estimation method of a power system, characterized in that the extraction within the rotated time point. The method according to claim 6, When extracting the three sample values, the frequency estimator samples the third sample value X2 that is sampled last, after the point of time that the first sample value X1 has been extracted three revolutions. Frequency estimation method of the power system characterized in that. The method according to claim 5 or 6, The real axis component generated by the cosine filter in the second step is the cosine signal Xr [n] and the sine signal Xi [n] which is the imaginary axis component generated by the sine filter. Frequency estimation method of the power system, characterized in that each obtained using 2. Equation 1

Equation 2

Where Hc (f) is the magnitude of the frequency response characteristic of the cosine filter, Hs (f) is the magnitude of the frequency response characteristic of the sine filter, A is the magnitude of the input signal, N is the number of samples per period, and φc Is the phase angle delay occurring in the cosine filter, φs is the phase angle delay occurring in the sine filter,

Is π / 2. The method according to claim 5 or 6, And a sample compensation value of the fourth step is obtained by multiplying the third sample value by a compensation coefficient. The method according to claim 5 or 6, Compensation coefficient of the third sample value of the fourth step is calculated by the following equation (1) frequency estimation method of the power system. Equation 1

Where a is

And b is

Xr [n1] is the output value of the first sampled cosine filter, Xi [n1] is the output value of the first sampled cosine filter, Xr [n0] is the output value of the second sampled cosine filter, and Xi [n0 ] Is the output value of the second sampled sine filter, Xr [n2] is the output value of the third sampled cosine filter, and Xi [n2] is the output value of the third sampled sine filter. The method according to claim 5 or 6, The frequency of the digital signal of the fifth step, the frequency estimation method of the power system, characterized in that calculated according to the following equation (1) or (2) according to the phase difference between the first sample value and the sample compensation value. Equation 1

,

Equation 2

,

In Equations 1 and 2, n1 is the first sampling time, n2 is the third sampling time, Xr [n1] is the output value of the first sampled cosine filter, and Xi [n1] is the first sampled sine filter. Xr [n2] is the output value of the third sampled cosine filter, Xi [n2] is the output value of the third sampled sine filter, Xcr [n2] is the compensation value of Xr [n2], and Xci [ n2] is the compensation value of Xi [n2], N is the number of samples per period, k is the compensation factor, and T is the phase difference between the two sample values.

Images