JPH085679A - Method and apparatus for detecting frequency and power system stabilization system - Google Patents

Abstract

PURPOSE:To detect the frequency with higher accuracy than linear approximation during a half cycle detection period by calculating the actual detection time of half cycle using a sampling phase time difference operated according to an inverse trigonometric function and then determining the frequency based on the reciprocal. CONSTITUTION:An AC amount Vi is sampled at a constant period DELTAT(e.g. 30 deg. of electric angle) and inputted while counting Ki. A current Vi is multiplied by a previous Vi-1 and the sign is determined. If the sign is negative, a zero point of AC is present and the time difference DELTAti between the zero point and current sampling point is calculated and stored. The time Ts of detecting section (between zero points) is then calculated based on the previous phase time difference DELTAti(i) to be stored, the number of sampling times KXDELTAT, and the current DELTAti. The detection frequency is determined as one half of the reciprocal of the time Ts and then the K is cleared thus waiting next input. Since the detecting section can be extended to a multiple of half cycle while sustaining the half cycle detection period, the accuracy can be enhanced.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a frequency detecting device, and more particularly to a frequency detecting method for an AC input amount in a power system.

[0002]

2. Description of the Related Art As a frequency detection method for an alternating current amount in a power system, as is apparent from Japanese Patent Publication No. 63-46648, there is a method of obtaining a zero point by linear approximation from sampling values immediately before and after the zero point of the alternating current amount. It is well known. As a proposal for this improvement, as disclosed in Japanese Patent Laid-Open No. 57-129126, each time a zero point is detected, the time from the zero point to the zero point before the arbitrary multiple half cycle in the past is calculated to improve the frequency detection accuracy. There are ways to improve.

On the other hand, as a frequency detecting method that does not rely on linear approximation, a method is known in which a frequency is calculated by an inverse trigonometric function operation with a plurality of sampling values as inputs, as described in Japanese Patent Laid-Open No. 2-187669.

[0004]

In the method based on the linear approximation, since the zero point is directly obtained from the sampling values around the zero point, the calculation error when the frequency changes by 5% is about 0.
The detection accuracy was low at 06%, and it was difficult to detect the frequency change width.

On the other hand, the method using the above-mentioned inverse trigonometric function is
Since the input sampling value of 3 to 5 points is required for a half cycle, the value often includes the vicinity of the peak. When the measurement of the AC amount is shared, the full scale may be kept low in order to improve the detection accuracy for a specific application, and the peak vicinity may exceed the full scale of the A / D converter depending on the AC waveform. Sometimes. However, if even one point exceeds the full scale, there is a problem that the frequency cannot be measured.

An object of the present invention is to overcome the above-mentioned problems of the conventional linear approximation method and inverse trigonometric function method, and to provide a highly accurate and highly reliable frequency detection method and apparatus without causing measurement failure. It is in.

Another object of the present invention is to provide a highly reliable power system stabilizing system which monitors the system frequency at high speed and with high accuracy and enables predetermined stabilization control when the frequency is abnormal. is there.

[0008]

SUMMARY OF THE INVENTION An object of the present invention is to provide a frequency detection method for detecting the AC frequency of a sine wave in real time by sampling the instantaneous amount of AC voltage or current at a predetermined cycle and The presence of a zero point is detected from the change in polarity, and the phase time difference of sampling with respect to the zero point is calculated by the inverse trigonometric function operation (tan to ¹ ) from the sampling values immediately before and after the zero point, and from the previous zero point to the current zero point. This is achieved by calculating the time of the detection section based on the phase time difference and the number of times of sampling during this, and obtaining the AC frequency of the sine wave from the reciprocal of the time of the detection section.

Further, the detection section is set from a zero point before a predetermined number of times (n times) to the present zero point, and the time of the detection section is calculated for each zero point from the first and last phase time difference and the sampling number. , The reciprocal of the time, n / 2, to obtain the frequency of the sinusoidal alternating current.

[0010]

The principle of the present invention will be described with reference to FIG.
The peak value of the AC input waveform is vm, sampling time T1
If the phase angle of T2 from the zero point between T and T2 is represented by θ ₁ and the constant sampling period ΔT is represented by the electrical angle d, immediately before the zero point,
Immediately after the sampling values v1 and v2 are expressed by the mathematical expressions (1) and (2)
Get more.

[0011]

[Number 3] _{v1 = vm · sin (θ 1} -d) ...... (1) v2 = When you expand vm · sinθ ₁ ...... (2) equation (1), the equation (3).

[0012]

Equation 4] _{v1 = vm (sinθ 1 · cosd} -cosθ 1 · sind) ...... (3) and Equation (2) from (3), Equation (5) is obtained with Equation (4).

[0013]

[Number 5] _{v2 / sinθ 1 = v1 / (} sinθ 1 · cosd-cosθ 1 · sind) ... (4) tanθ 1 = v2 · sind / (v2 · cosd-v1) ... (5) In the formula (5), Since the signs of v1 and v2 are always opposite, the sampling angle d is 0 to 90 degrees (usually d = 30
), The denominator term on the right side does not become 0, so there is no need for so-called indefinite processing. The calculation result on the right side is always 0 or a positive value.

As the arctangent of this equation (5), the phase angle θ ₁
Can be obtained by equation (6).

[0015]

[6] ^{_{θ 1 = tan ~ 1 (v2}} · sind / (v2 · cosd-v1)) ... (6) If the sampling interval △ T is determined, the electrical angle d and sind, cosd is a constant.

The phase angle θ ₁ is an electrical angle expression of the phase time difference Δt2 at the sampling time T2 with respect to the zero point, and Δt
2 is obtained by the mathematical expression (7).

[0017]

[Formula 7] Δt2 = (θ ₁ / d) · Δt = (ΔT / d) · tan ~ ¹ (v2 · sind / (v2 · cosd-v1)) (7) Similarly, half cycle Assuming that the electrical angle immediately after the subsequent zero point is θ ₂ , Δt6 can be obtained from the equation (8).

[0018]

[Formula 8] Δt6 = (θ ₂ / d) · Δt = (ΔT / d) · tan to ¹ (v6 · sind / (v6 · cosd−v5)) (8) Therefore, from the zero point to the zero point If the half cycle up to is the detection section for frequency calculation, the time Ts of each detection section is obtained by the equation (9), and the frequency f at that time is obtained by the equation (10) with 1/2 of the reciprocal of Ts.

[0019]

Ts = Δt2 + 4ΔT−Δt6 (9) f = 1 / (2Ts) = 1 / (2 (Δt2 + 4ΔT−Δt6)) (10) Formula (9), Generalizing equation (10), equation (11)
(12) is obtained.

[0020]

[Equation 10] Ts = Δt _{(1) i} + K · ΔT−Δt _i (11) f = 1 / (2Ts) = 1 / (2 (Δt _{(1) i} + K · ΔT−Δ t _i )) ... (12) where Δt _{(1) i} : time phase difference from the previous zero point, Δ
t _i : time phase difference from the current zero point, K: number of samplings from the next sampling immediately after the previous zero point to immediately after the current zero point.

According to the frequency detection of the present invention, in the half cycle of the AC of the sine wave, the sampling phase time difference calculated by the inverse trigonometric function (tan ~ ¹ ) based on the two sampling values before and after the zero point is used. Since the real time of the detection section of the cycle is calculated in real time and the frequency is obtained from this reciprocal, the detection accuracy can be improved as compared with the conventional linear approximation.

Further, since two sampling values immediately before and after the zero point are used for the frequency calculation, if the sampling period is set to 30 degrees which is normally used, the value becomes sufficiently smaller than the peak value of the alternating current. It is possible to detect a frequency with high reliability without causing scale-over at the time of D / D conversion to make measurement impossible.

The frequency detection cycle in the above is
Although it is set as a half cycle for each zero point, it may be set as one cycle. That is, when the positive / negative determination by multiplying the sampling value is negative, only the zero point where the sampling value after the zero point becomes positive (or negative) is adopted, and the frequency calculation is performed.

Further, as an extension of the present invention, it is possible to adopt a method in which the detection interval of frequency calculation is changed from the half cycle of the equation (11) to an arbitrary multiple (n) half cycle, and the frequency detection period is half cycle. . The time Ts and the frequency f of the detection section in this case are obtained by the mathematical expressions (13) and (14).

[0025]

(11) Ts = Δt _{(n) i} + K _n · ΔT−Δt _i (13) f = n / (2Ts) = n / (2 (Δt _{(n) i} + K _n · ΔT -Δt _i )) ... (14) where Δt _{(n) i} : Phase time difference immediately after the zero point before n times half cycle, K _n : From the time immediately after the zero point before n times half cycle to the point immediately after this zero point Is the number of sampling times.

According to this, since K _n · ΔT in the denominator term of the equation (14) is about n times the K · ΔT of the equation (12), the influence of the error included in the phase time difference of the denominator term. However, the frequency detection accuracy can be further improved while maintaining high-speed detection every half cycle.

According to this, it is possible to detect the range of change in frequency, which has been difficult with conventional dedicated hardware, and it is possible to apply to a power stabilizing device.

[0028]

Embodiments of the present invention will be described with reference to the drawings.

FIG. 3 shows the configuration of a frequency detecting apparatus according to an embodiment of the present invention. The voltage (or current) of the power system 3 measured and measured by the voltage transformer 4 is taken into the frequency detection device 1 via the filter 5 for removing harmonic components and the A / D converter 6 for analog / digital conversion. Be done. The A / D converter 6 outputs a sampling value of a digital amount at every sampling period ΔT controlled by the frequency detecting device 1. The frequency detected by the frequency detection device 1 is output to the output device 7 such as display means. The filter 5 and the A / D converter 6 may be collectively referred to as a frequency detection device.

The frequency detection device 1 comprises a control unit 10 for controlling the entire processing, an input / output processing unit 11 for taking in sampling values and outputting a calculation result, and a frequency detection processing unit 12. The storage device 2 has a storage area for storing the time series of the sampling values in the range necessary for calculation, the calculation result of each unit, and the like, which are written / read by the control unit 10.

The frequency detection processing section 12 is a zero point detecting means 1 for detecting a zero point from a sign change of time-series sampling values.
21, sampling number integration means 122 that counts the number of times of sampling in the frequency detection section, phase difference calculation means 123 that calculates the phase difference of the sampling point immediately after the zero point by an inverse trigonometric function, and detection calculated using the phase difference and the number of times of sampling The frequency calculation means 124 calculates the frequency of the system from the time of the section.

The frequency detecting device 1 of this example uses a CPU and is composed of software including programs and data for executing the functions of the respective parts. In this case, the storage device 2 is the main memory and stores these programs and data.

FIG. 1 is a flow chart for explaining the operation of the frequency detecting apparatus of this embodiment, which is an example in which a half cycle of the alternating current to be measured is a detection section and a detection cycle for frequency calculation.
The sampling interval d is an electrical angle of 30 degrees which is widely adopted in protection relays for electric power systems.

This process is started each time sampling data is input. First, the current sampling value v _i is captured and stored (s101), and the sampling number K is set to +
1 (s102). The sampling number K is cleared to 0 at the previous detection of the zero point.

Next, the previous sampling value v _i-1 input in time series is multiplied by the current sampling value v _i (s
103). Whether there is a change in polarity between This is v
_i-1 and v _i are multiplied to determine whether the multiplication result is positive or negative (s10
4). If the result of the determination is positive, there is no change in the polarity (sign) during the sampling of both. That is, since there is no AC zero point of the sine wave between the previous sampling and the current sampling, the processing is terminated and the next sampling data is waited for.

On the other hand, if the positive / negative determination result is negative, there is a zero point (zero crossing) of the sine wave AC between the previous time and this time, so the phase time difference Δt _i between this zero point and the current sampling point is expressed by the formula ( It is obtained from 7) (s105).

In this example, d = 30 degrees and cosd = √3 /
Since 2, sind = 1/2, Expression (7) is replaced with Expression (15).

[0038]

[Number 12] _{△ t i = △ T / 30} · tan ~ 1 (v i / (2√3 · v i -4 · v i-1)) ... (15) phase time difference immediately after the current zero point △ t _i Are stored for use in the next frequency calculation (s106).

If the correspondence between the value of the variable part of the arctangent and Δt _i is made into a data table in advance and stored in the storage device 2, the calculation speed of the equation (15) can be increased.

Next, according to the equation (11), the time from the previous zero point to the current zero point, that is, the real time of the detection section is calculated, and the reciprocal thereof is halved to obtain the frequency in the detection section of the half cycle. Obtain (s107). After that, K is cleared to 0 (s108) and the process is terminated, and the next sampling is awaited.

When the frequency detection period is set to one cycle, the positive / negative judgment of the current input data v _i is performed after the above step s102, and if positive, step s10
Go to 3 and end if negative. It goes without saying that the frequency is calculated as the reciprocal of the real time of one cycle of the detection section.

Further, the phase time difference Δt _i-1 (= ΔT−Δt _i ) of the sampling point immediately before the zero point is calculated, and the phase time difference immediately after the previous zero point and the phase time difference immediately before the current zero point and between them are calculated. The time of the detection section may be obtained from the number of times of sampling. Furthermore, instead of immediately after the zero point, the time of the detection section can be calculated based on the phase time difference Δt _i−1 of the sampling immediately before the zero point.

According to the present embodiment, it is possible to detect a frequency with higher accuracy than a linear approximation with a short detection cycle of half a cycle. Further, since the denominator term of the arithmetic expression does not become 0 at the sampling cycle d = 30 degrees, the arithmetic processing becomes simple. Moreover, since the sampling value used for the calculation is limited to just before and after the zero point (others only monitor the sign), the measurement is not impossible due to overscale, so that the reliability is high.

Next, a second embodiment of the present invention will be described.

FIG. 4 shows a schematic configuration of a power system stabilizing system to which this embodiment is applied. The electric power system has a plurality of generators 40, as schematically shown in the dotted line.
The electric power generated from 0 is supplied to the load via the power transmission line 300. The circuit breaker 500 that disconnects / connects the generator 400 and the power transmission line 300 is controlled by the slave station 200 of the stabilizing device 100. The stabilizing device 100 receives the voltage and current of the electric power system or accident information by a protection relay or the like via the communication line 600.

The stabilizing device 100 is the frequency monitoring device 10.
0a and a setting means for the frequency settling value of the power system (not shown) are provided, and when the system frequency for the frequency settling value is too large or too small, a control command 2 is issued to the slave station 200.
00a is output. The frequency monitoring device 100a detects the frequency of the grid by using the voltage (or current) received by the stabilizing device 100 as the sampling value (vi).

In response to the control command 200a, the slave station 200 performs control 200b such as disconnection / connection of the circuit breaker 500 of the power system, load limitation, or system separation to prevent occurrence of step-out.

FIG. 5 is a flow chart for explaining the operation of the frequency monitoring device 100a. The detection interval for frequency calculation in this example is n times the half cycle, and the frequency detection period is the half cycle. Usually, n is set to 3-5.

The processing operation of the frequency monitoring device 100a is activated when there is a request for changing the frequency set value or when sampling data is input. If there is a change request (s201), a new frequency settling value f ₀ is fetched or set (s202). In addition, also when changing the multiple n of the half cycle for setting the detection section, it is set in this step.

When there is no set value change request, the current sampling value v _i is fetched and stored (s203), and the sampling number K after the previous zero point is incremented by 1 (s20).
4). Next, the previous sampling value v _i-1 is multiplied by the current sampling value v _i to determine whether the sign is positive or negative (s20
5) If the determination is positive, the process ends.

If the determination is negative, there is a zero point between the previous sampling and the current sampling, so the following processing is performed.

First, the sampling number K _n for each ΔT is counted over the detection section of n times and half cycles (s2
06). This K _n is calculated in the following formula order (arrow) as a stack of integrated values for each half cycle.

[0053]

[Formula 13] K _n = K _n-1 + K ↓ K _n-1 = K _n-2 + K ↓ ・ ・ ・ ・ ・ ・ ↓ K ₂ = K ₁ + K ↓ K ₁ = K That is, the previous (n -1) Integrated values _Kn-1 , _Kn-2 ,. ． ． ． K of the current half cycle is sequentially added to each of K ₂ and K ₁ , and the integrated value (K of the previous n half cycles to 1 cycle before is incremented by 1 step
_n ,. ． ． Calculate K ₂ ). The integrated value K ₁ = K before the current half cycle.

Next, the phase time difference Δt _i of the current sampling is calculated from the equation (15) (s207) and stored (s208). Further, the time Ts of the detection section from the zero point n times before to the current zero point is calculated by the mathematical expression (13) and the frequency f is calculated by the mathematical expression (14) (s209). After this, K is cleared to 0 (s210).

Next, the deviation Δf of the detection frequency f with respect to the frequency set value f ₀ is obtained, and it is judged whether the deviation Δf is within the allowable range (s211). If it is within the allowable range, it is considered normal and only the detected frequency is detected. Is output (s213) and displayed or recorded.

On the other hand, if the detected frequency is above the set value, the amount of power generation is greater than the load power amount, and if it is below the set value, the amount of power generation is small. In either case, if the abnormal frequency exceeds the allowable range, the system may be out of step, so it is necessary to perform stabilization control such as power generation amount restriction or load restriction or system separation of the power system.

Therefore, if the detected frequency exceeds the allowable range of the settling value, an abnormal signal of the frequency is stabilized by the stabilizing device 10.
It is output to 0 (s212).

According to the present embodiment, the frequency of the system can be detected with high accuracy and every half cycle by the detection section extended to n times the half cycle. By the way, the frequency detection error according to the present embodiment is improved by one digit from 0.06% in the conventional case to 0.009% when the frequency changes by 5%.

As a result, even when the set value of the system frequency is changed, the frequency change width can be detected at high speed and with high accuracy, so that the reliability of the power system stabilizing system can be further improved.

In the above embodiment, the frequency monitoring device 100a.
Is attached to the stabilization device 100, but it is also possible to attach it to the slave station 200 and perform similar monitoring.

[0061]

According to the frequency detecting method of the present invention, the phase difference between the zero point and the sampling period is calculated by the inverse trigonometric function based on the sampling values before and after the zero point of the alternating current, and from this, the detection section from the zero point to the zero point. Since the time is calculated, there is an effect that the detection accuracy can be improved. Moreover, it is possible to extend the detection section to half cycles of an arbitrary multiple while maintaining the detection cycle of half cycle, and there is an effect that the accuracy can be further improved.

According to the frequency detecting apparatus of the present invention, the software function enables the realization of a hardware equivalent to or more than the dedicated hardware at a low cost, and the measurement principle using the sampling value near the zero point does not cause measurement failure. There is an effect that the reliability can be improved.

According to the power system stabilizing system of the present invention, the system frequency is detected with high accuracy every half cycle, and the variation width with respect to the settling value is monitored to accurately evaluate whether the system frequency is too large or too small. Since the stabilization control of is executed,
It is possible to prevent out-of-steps in the power system and to further improve the system reliability.

[Brief description of drawings]

FIG. 1 is a flowchart illustrating a frequency detection method according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram illustrating a relationship between a waveform of an alternating current to be measured and a sampling time.

FIG. 3 is a configuration diagram illustrating a frequency detection device according to an embodiment of the present invention.

FIG. 4 is a configuration diagram illustrating an outline of a power system stabilizing system.

FIG. 5 is a flowchart illustrating a frequency monitoring method according to a second embodiment.

[Explanation of symbols]

1 ... Frequency detection device (CPU), 2 ... Storage device (main memory), 3 ... Power system, 4 ... Voltage transformer, 5 ... Filter,
6 ... A / D converter, 7 ... Output device, 10 ... Control part, 11
Input / output processing unit 12, frequency detection processing unit 100, power system stabilizing device 100a, frequency monitoring device 121
... Zero point detecting means, 122 ... Sampling number integrating means,
123 ... Phase difference calculating means, 124 ... Frequency calculating means, Δ
T ... Sampling cycle, d ... Electrical angle expression of sampling cycle, vi ... Sampling value, θi ... Phase angle from zero point to the next sampling point, Δti ... Phase time difference from zero point to the next sampling point, Ts ... Detection section Time, f ... Detection frequency.

Claims (10)

[Claims] 1. A frequency detecting method for detecting the AC frequency of a sine wave in real time, wherein the instantaneous amount of AC voltage or current is sampled at a predetermined cycle, and the presence of a zero point is detected from the polarity change of the sampled value. The sampling phase time difference with respect to the zero point is calculated from the sampling values immediately before and after the zero point by the inverse trigonometric function operation (tan ~ ¹ ), and the time of the detection section from the previous zero point to the current zero point is calculated as the phase time difference and The frequency detection method is characterized in that the frequency of the sinusoidal wave is calculated from the reciprocal of the time of the detection section. 2. The frequency detection method according to claim 1, wherein the sampling phase time difference is a sampling phase time difference Δt immediately after the zero point, and is calculated by (Equation 1). [Formula 1] Δt = (ΔT / d) · tan ~ ¹ (v2 · sind
/ (V2 · cosd-v1)) where ΔT is the sampling period, d is the electrical angle of ΔT,
v1 is a sampling value immediately before the zero point, and v2 is a sampling value immediately after the zero point. 3. The method according to claim 2, wherein the previous zero point is before a half cycle of sinusoidal wave alternating current, and the time Ts of the detection section is calculated by (Equation 2).
A frequency detecting method, characterized in that the frequency is obtained as 1/2 of the reciprocal of. ## EQU2 ## Ts = Δti ₍₁₎ + KΔT−Δti where Δti ₍₁₎ is the previous phase time difference, Δti is the current phase time difference, and K is the number of times of sampling. 4. The frequency detection method according to claim 1, 2 or 3, wherein the frequency is detected every half cycle. 5. The zero point according to claim 1 or 2,
Only when the sampling value immediately after the zero point is positive or negative is adopted, and the frequency is obtained for each cycle as the reciprocal of the time of the detection section in which the previous zero point is one cycle before the sinusoidal wave AC. Frequency detection method. 6. A frequency detecting method for detecting an AC frequency of a sine wave every half cycle, sampling an instantaneous amount of AC at a predetermined cycle, detecting the presence of a zero point from the polarity change of the sampling value, and detecting the zero point. The phase time difference of the sampling is calculated from the sampling values immediately before and immediately after by the inverse trigonometric function operation (tan to ¹ ), and the phase time difference corresponding to the zero point of the half cycle before a predetermined number of times (hereinafter, referred to as n times) and this time. Of the sine wave as the reciprocal of n / 2, which is the reciprocal of the time, based on the phase time difference corresponding to the zero point and the number of times of sampling of the detection section from the zero point before n times to the present zero point. A frequency detection method characterized by obtaining a frequency. 7. The frequency detecting method according to claim 1, wherein the predetermined cycle of the sampling is set to 30 degrees in electrical angle expression. 8. A frequency detecting device for detecting a sinusoidal AC frequency in real time, comprising: sampling means for sampling an instantaneous amount of AC voltage or current in a predetermined cycle; and polarity of a sampling value input in time series. A zero point detecting means for detecting the presence of a zero point from the change; a phase difference calculating means for calculating a sampling phase time difference with respect to the zero point from the sampling values immediately before and after the zero point by an inverse trigonometric function operation (tan to ¹ ); Sampling number integration means for counting the number of sampling times in the detection section from the zero point before (hereinafter, n times) to the present zero point, the phase time difference corresponding to the zero point before n times and the phase time difference corresponding to the current zero point, and The time of the detection section is calculated based on the number of times of sampling, and the reciprocal of the time is used to calculate the AC frequency of the sine wave. A frequency detecting device comprising: a frequency calculating means for calculating a wave number; and a storage means for storing the phase time difference from n times before to this time. 9. The sampling means according to claim 8, further comprising an A / D converter that converts the instantaneous amount of analog amount into a digital amount, and the predetermined period is set to 30 degrees in electrical angle expression. A frequency detection device characterized by the following. 10. A power system stabilizing system that controls a power supply and a system configuration when a frequency of a power system including a load, a plurality of generators, and a power transmission line that supplies power to the load is abnormal. Is sampled at a cycle of an electrical angle of 30 degrees, and the phase time difference of the sampling is calculated by an inverse trigonometric function (tan) from the sampling values immediately before and after the zero point of the AC.
~ ¹ ) to calculate the time of the detection section from the zero point of the half cycle before the predetermined number of times to the zero point of this time based on the phase time difference and the number of sampling times, and the frequency detection means for obtaining the frequency of the power system, A power system stabilization system comprising: a frequency variation width monitoring unit that obtains a deviation of the frequency detected by the frequency detection unit with respect to a frequency set value of the system and outputs an abnormal signal when the frequency exceeds a predetermined range. .