KR20040042092A - Method for detecting fault location on transmission line using travelling waves - Google Patents

Abstract

PURPOSE: A method for detecting failure points of a transmission line by using a traveling wave is provided to rapidly and accurately detect the failure points without depending on an error factor such as a voltage phase angle, a failure resistance and a voltage difference during an accident. CONSTITUTION: A method for detecting failure points of a transmission line by using a traveling wave includes the steps of: sampling(S10) a voltage and a current values of a current period of an electric relay install position; filtering(S20) the sampled voltage and current values by using a finite impulse response(FIR) filter; extracting(S30) the superimposed component by subtracting the data stored at the previous period in the filtered data; and extracting the forward wave and the backward wave components of the electric relay install position from the superimposed components.

Description

Fault detection method of transmission line using traveling wave {METHOD FOR DETECTING FAULT LOCATION ON TRANSMISSION LINE USING TRAVELLING WAVES}

The present invention relates to a method for detecting a failure point of a transmission line using traveling waves, and more particularly, by a failure of a line using a finite impulse response (FIR) filter having a high-pass characteristic. The high frequency component is strengthened and the DC-offset component is weakened so that the peaks of the cross-correlation function of the forward and backward waves can be clearly detected. It is to provide a fast and accurate method of detecting a failure point that is not affected by the problem.

Accurately identifying the point of failure in the event of a failure of a transmission line is essential for rapid failure recovery to maintain a stable system. In recent decades, the fault-point expression algorithm has been actively studied. The method of using the traveling wave as the fault expression algorithm ([1] GB Ancell, NC Pahalawaththa, "Maximum Likelyhood Estimation of Fault Location on Transmission Lines Using Traveling Waves", IEEE Trans.on PWRD, Vol. 9, No. 9, pp 680-689, 1994), [2] T. Takagi, Y. Yamakoshi, J. Baba, K. Uemura, T. Sakaguchi, "A New Algorithm of an Accurate Fault Location for EHV / UHV Transmission." Lines: Part I-Fourier Transform Method, IEEE Trans.on PAS, Vol. 100, No. 3, pp. 1316-1323, 1988), Impedance method using fundamental wave components of voltage and current ([3] MS Sachdev , R. Agarwal, "A Technique for Estimating Transmission Line Fault Locations from Digital Impedance Relay Measurements", IEEE Trans.on PWRD, Vol. 3, No. 1, pp.121-129, 1988).

The occurrence of a fault in the line generates a traveling wave from the fault point to the relay installation point. To date, cross-correlation techniques ([4] PA Crossley, PG McLaren, "Distance Protection Based on Traveling Waves," IEEE Trans. On PAS, Vol 102, No. 9, pp.2971-2983, September (1983), but the disadvantage of this method is that it does not detect near-incidence of the relay installation point and near the voltage phase angle zero. In order to solve this problem, the effective value of the cross-correlation output is used and the correction factor is multiplied by the relay signal in the event of 0 ° voltage phase angle ([5] EH Shehab-Eldin and PG McLaren). , Traveling Wave Distance Protection-Problem Areas and Solutions, IEEE Trans., Power Delivery, Vol. 3, No. 3, 894 (1998)) have been proposed, but near and near voltage phase angle zero Basically, since it has the same shape as the power frequency [60 Hz] component, it is difficult to distinguish the distance because the peak part of the forward and backward waves necessary for cross-correlation is not prominent.

The present invention is to solve such a problem, by using a finite impulse response (FIR) filter having a high-pass characteristics (FIR) filter to enhance the high frequency component caused by the line failure And the DC-offset component is attenuated so that the peaks of the cross-correlation function of the forward and backward waves can be clearly detected so that the point of failure is not affected by sources of error such as voltage phase angle, fault resistance and power supply phase difference at both ends. It is for providing a detection method.

Since the FIR filter of the present invention has the effect of amplifying high frequency as well as the DC-offset elimination effect, the FIR filter is applied to both the voltage and current signals without judging the voltage phase angle, and as a result, the peak of the cross-correlation function is more easily distinguished. We propose an algorithm. Therefore, accurate fault distance calculation is possible even in a near-incidence near a relay installation point having only a power frequency component and a voltage phase angle of 0 °.

1 is a schematic diagram showing traveling waves such as forward waves and backward waves generated during an accident on a transmission line.

FIG. 2 is a diagram illustrating a frequency response of a finite impulse response (FIR) filter in the case of 12 samplings per period.

Figure 3 shows the result of applying the FIR filter at the voltage phase angle 0 ° accident.

Fig. 4 shows the calculation result of the cross-correlation function when (a) the prior art correction factor is used and (b) the FIR filter according to the present invention is applied.

5 shows a lattice diagram for the calculation of the fault distance.

6 shows a schematic flowchart in accordance with an embodiment of the present invention.

7 shows an example for verifying the effectiveness of an embodiment of the present invention.

8 is a graph comparing the error in calculating the fault distance in the case of voltage phase angles 0 ° and 90 ° when the above embodiment is applied.

9 is a graph comparing errors in calculating a failure distance when the failure resistance is changed when the above embodiment is applied.

10 is a graph comparing errors in calculating a failure distance according to a change in power supply phases at both ends when the above embodiment is applied.

According to an aspect of the present invention, a method for determining a failure distance when a failure occurs in a transmission line includes: sampling voltage and current values of a current cycle of a relay installation point; Filtering the sampled voltage and current values using a finite impulse response (FIR) filter; Extracting a superimposed component by subtracting a data value stored in a previous period from the filtered data; And extracting forward and backward wave components of the relay installation point from the overlap component.

According to another aspect of the present invention, there is provided a method for determining a fault distance, the method comprising: storing the extracted forward wave sample; Comparing the magnitude of the forward wave with a predetermined threshold to determine an accident when the threshold is greater than or equal to the threshold; Obtaining a cross-correlation function value of the forward wave component and the backward wave component when it is determined that the accident is occurring; Obtaining a time at which the cross-correlation function value has a maximum value; And determining a distance to the failure point based on the time having the maximum value.

The extracting of the forward and backward wave components may include: a modal conversion step of obtaining an actual mode of a traveling wave from the overlapping component; And extracting the forward and backward wave components from the aerial mode signal.

Here, the finite impulse response filter, Is expressed by a relational expression of k, where k is a data index; M: It is preferable that it is the size of a filter window.

The storing of the forward wave is preferably to include at least one sample before the accident judgment.

Determining the distance to the failure point is a time τ at which the correlation function has a maximum value Satisfies the relationship of, where L is the distance from the viewpoint to the failure point; v: propagation velocity in the line ( Is preferred.

Hereinafter, with reference to the drawings will be described in detail a preferred embodiment of the present invention.

1 is a schematic diagram showing traveling waves such as forward waves and backward waves generated during an accident on a transmission line.

As shown in FIG. 1, when a transmission line TL is provided between one equivalent power supply terminal A and another equivalent power supply terminal B, an accident occurs at a certain point F of the transmission line TL. In this case, a traveling wave occurs at the accident point. The traveling wave has two components, a forward wave f ₁ (t) and a backward wave f ₂ (t). Hereinafter, the traveling wave will be described in more detail.

In this example, the power system is treated as a lossless distribution constant circuit that does not depend on frequency, and the displayed wave equation used as the second-order partial differential equation is used. The solution of the wave equation, widely known as the D'Alembert solution, is shown in equation (1).

(One)

Where u is the propagation velocity Z ₀ is the characteristic impedance of the line.

At this time, the forward and backward waves of the relay installation point can be expressed as shown in equation (2).

(2)

6 shows a schematic flowchart in accordance with an embodiment of the present invention.

As shown in FIG. 6, the method for detecting a failure point according to an embodiment of the present invention includes sampling a voltage and current value of a current period of a relay installation point (S10); Filtering the sampled voltage and current values using a finite impulse response (FIR) filter (S20); Extracting a superimposed component by subtracting a data value stored in a previous period from the filtered data (S30); Extracting forward and backward wave components of the relay installation point from the overlap component (S40); Storing the extracted forward wave sample (S50); Comparing the magnitude of the forward wave with a predetermined threshold to determine an accident when the threshold is greater than the threshold (S60); Obtaining a cross-correlation function value of the forward wave component and the reverse wave component when it is determined to be an accident (S70); Obtaining a time at which the cross-correlation function value has a maximum value (S80); And determining a distance to a failure point based on the time having the maximum value (S90).

If necessary, as shown in FIG. 6 to prevent signal aliasing, a step S05 of performing low pass filter processing on the voltage and current data before sampling may be added. In addition, a step (S22) of storing one period data may be provided for the data subjected to the FIR filter processing (S20), but it is not necessary to store only one period here but may be stored for several periods. In addition, if necessary, if more than one period of data is not input step (S24) to return to the first step can be provided. Further, after the superposition component is extracted (S30), the modal transformation process (S32) can be performed on the data.

Hereinafter, the extraction (S30) of the overlapping components will be described in detail. The occurrence of an accident on a line is equivalent to the application of a voltage of the same magnitude and a different sign at the point of failure. Therefore, the voltage and current components after the accident can be viewed as the sum of the steady state components before the accident and the transient components after the accident.

As a result, the overlapping component s used as the relaying signal is expressed by Equation (3).

(3)

Here, each subscript represents po: post accident value and pr: pre accident value.

Hereinafter, the modal transformation S32 will be described. Voltage and current of a line including traveling wave information in a three-phase transmission line are composed of several modes. Since each mode has different propagation speed, attenuation constant, and characteristic impedance, three-phase voltage and current signals can be divided into three independent modes. In the present embodiment, a Wedepohl matrix such as Equation (4) is used to easily interpret a three-phase soft loop line as a single phase line.

(4)

Among the three independent modes, the ground mode has the characteristics of high attenuation and slowness, so in this embodiment, the aerial mode (see [5] EH Shehab-Eldin and PG McLaren, Traveling Wave Distance) Protection-Problem Areas and Solutions, IEEE Trans., Power Delivery, Vol. 3, No. 3, 894 (1998)).

Hereinafter, the cross correlation function will be described in detail. The cross-correlation function tells how correlated the incoming wave and the reflected wave, that is, the degree of similarity. The cross correlation function represents the degree of similarity between the sampled signal X and the appearing signal Y after the time delay τ. The equation is as shown in (5).

(5)

Δt: time of the sample value interval, M: number of samples per window.

If the signal X appears in the incoming signal Y sequentially, the cross-correlation function has a maximum value when τ = T in equation (5). The time difference T is used to find the distance of failure when a failure occurs. In practice, forward and backward waves have different mean values, which distort the cross-correlation function of the two signals. In order to solve this problem, the present invention uses discrete cross-correlation, which subtracts two different mean values. The equation is as shown in (6).

(6)

here, , Indicates.

In the embodiment, the forward wave and the backward wave are compared, and the forward wave is already stored and the comparison is continuously performed with the backward wave sequentially received after the sampling time interval. When the waveform most similar to the stored forward wave comes in, the cross correlation function has a peak value P as shown in FIG.

Hereinafter, the process of calculating the failure distance (S90) will be described in detail. When a fault occurs in a line, a backward wave is generated due to the changed components of voltage and current, and the reverse wave is reflected at the discontinuity point (the main bus) behind the viewpoint, turns into a forward wave, leaves the viewpoint and proceeds along the line, and then reflects at the fault point. You are returned to the observation point. That is, to calculate the distance to the failure point, it is used by measuring how much time delay the forward wave leaving the observation point reaches as the backward wave as the backward wave.

In this embodiment, the data window size used to obtain the cross-correlation function so that the traveling wave reciprocates at least once even in the case of far-end failure is set to more than twice the propagation time of the protection section. In addition, in this embodiment, two samples before reaching a threshold are stored together so that forward waves leaving the relay point in the window more reliably and accurately.

5 shows a lattice diagram for the calculation of the fault distance. As shown in FIG. 5, the time at which the maximum value of the discrete cross-correlation function appears is the time 2t ₂ at which the forward wave leaves the observation point and reflects back from the failure point. Therefore, the distance from the observation point to the failure point can be expressed as Equation (7).

(7)

Where L is the distance from the observation point to the failure point, and v is the propagation velocity on the track. ).

Hereinafter, the operation of the finite impulse response (FIR) filter, which is a major feature of the present invention, will be described in detail.

In general, the FIR filter is represented by Equation (8).

(8)

Here, M: represents the filter window size, and the data window of this digital filter can be arbitrarily determined. Assuming a dc-offset component in a window composed of an arbitrary number of sample data as a constant, the input waveform in a certain window is expressed as shown in equation (9).

(9)

Where c is a constant, N is a number of samples per period, and n is a harmonic order.

When the waveform of equation (9) passes through the filter represented by equation (8), its output is represented by equation (10).

10

here, ,

, Indicates.

FIG. 2 is a diagram illustrating a frequency response of a finite impulse response (FIR) filter at 12 samplings per cycle, as an example.

Among the superposition components obtained, the high frequency component is maximized at the voltage phase angle of 90 ° and minimum at the voltage phase angle of 0 °. The current signal includes a DC-offset component, which is minimized at a voltage phase angle of 90 degrees and maximum at a voltage phase angle of 0 degrees. In case of an accident at or near voltage 0 °, the traveling wave is low in level and slow to change.

On the other hand, the peak appears in the discrete cross-correlation function due to the high frequency component. However, the peak portion becomes inconspicuous in the voltage phase angle 0 ° accident and the proximity accident. In case of voltage phase angle 0 ° accident, the method of the related art ([5] above) uses a correction factor ( ) Is multiplied by the relay signal. However, in this method, when the voltage phase angle is close to 0 °, only the power frequency [60Hz] appears in the forward wave and the backward wave, and thus the distance calculation is not accurate.

In the present invention, the FIR filter is applied so that the peak of the discrete cross correlation is prominent even in the case of 0 ° and close proximity accidents, so that the accurate failure distance calculation is possible.

The sampling frequency of this embodiment is 60 [kHz]. The applied FIR filter has a linear phase delay and no phase distortion at each frequency.

Let's say

ego

(11)

Equation (11) shows the frequency response of the FIR filter. When the fundamental wave of maximum magnitude X _m is passed through the FIR filter proposed in this paper, the magnitude of the fundamental wave is 0.00628 times and the phase is 89.8 ° ahead. This reduced fundamental wave size compensates to be equal to the fundamental wave size of the original signal.

In this embodiment, the FIR filter is applied to both voltage and current signals to calculate the accident distance without determining the phase angle of the accident, thereby reducing the computational burden on the algorithm.

Figure 3 shows the result of applying the FIR filter at the voltage phase angle 0 ° accident.

Hereinafter, in order to verify the method proposed in the present invention, a result of simulating a one-line ground fault using EMTP targeting 100km of a one-line 345 [KV] three-phase balanced transmission line of Sinpo Port in Sinokcheon, which is an actual system, will be described. do.

Figure 7 schematically shows the subject schematic in the above case for validating the effectiveness of an embodiment of the present invention.

For each simulation system, this study examined the change of fault distance, fault resistance, and power supply phase difference at both ends. The fault distances were 0.1 [pu] ~ 0.9 [pu], and the fault resistance was changed to 0 [Ω], 10 [Ω], 20 [Ω], 30 [Ω]. Sampling frequency for data acquisition was 60 [kHz], and a second-order Butterworth low pass filter with 30 [kHz] cutoff frequency was used to prevent aliasing errors. Fault distance calculation errors were calculated using equation (12).

(12)

4 illustrates the calculation result of the cross-correlation function when (a) the prior art correction factor and (b) the FIR filter according to the present invention are applied in the case of simulating a voltage phase angle of 0 ° 50 km.

In Figure 4 (a) when using the correction factor the peak (peak) is unclear and the fault distance calculation is not accurate, (b) when the FIR filter is applied, the peak (peak) is clearly visible and the fault distance calculation was accurate.

8 is a graph comparing the error in calculating the fault distance in the case of voltage phase angles 0 ° and 90 ° when the above embodiment is applied. FIG. 8 shows a fault distance calculation error at a voltage phase angle of 0 ° and a voltage phase angle of 90 ° when the power supply phase difference between both ends is 10 ° and the fault resistance is 10Ω. In case of 5km accident, it showed accurate fault distance calculation result with error 0.06 [%] and maximum error was 1.07 [%] at 0.9 [pu]. In this paper, only the magnetic stage information is used and the error increases with the failure of the fabric.

9 is a graph comparing errors in calculating a failure distance when the failure resistance is changed when the above embodiment is applied. In Figure 9 it can be seen that the fault distance calculation error is not related to the fault resistance.

10 is a graph comparing errors in calculating a failure distance according to a change in power supply phase difference between both ends when the above embodiment is applied. As shown in FIG. 10, since the present embodiment extracts and uses overlapping components, it may be understood that the present invention is also irrelevant to the influence of the load current.

As described above, in the detailed description of the present invention, specific embodiments have been described, but it should be taken as exemplary, and various modifications may be made without departing from the technical spirit of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined not only by the claims below, but also by the equivalents of the claims.

By using the transmission line failure point detection method using the traveling wave of the present invention, the peak of the cross-correlation function of the forward wave and the backward wave is clearly detected so that it is not affected by error sources such as voltage phase angle, fault resistance, and power supply phase difference at both ends in case of accident. It is possible to provide a fast and accurate method of detecting a failure point.

Since the FIR filter of the present invention has the effect of amplifying high frequency as well as the DC-offset elimination effect, it is easier to distinguish the peak of the cross-correlation function by applying the FIR filter to both voltage and current signals without judging the voltage phase angle. It is possible. Therefore, accurate fault distance calculation is possible even in a near-incidence near a relay installation point having only a power frequency component and a voltage phase angle of 0 °.

In addition, through the case application, the algorithm of one embodiment of the present invention verified that the fault distance calculation result was not affected by the voltage phase angle, fault resistance, and power supply phase difference at both ends. In this case, only the information of the magnetic terminal of the relay is used without any communication means, and an accurate result within 1.1% of the maximum error is shown.

Claims (6)

In the method for determining the failure distance in the event of a failure of the transmission line, Sampling the voltage and current values of the current period of the relay installation point; Filtering the sampled voltage and current values using a finite impulse response (FIR) filter; Extracting a superimposed component by subtracting a data value stored in a previous period from the filtered data; And And extracting forward and backward wave components of the relay installation point from the overlapping components. The method of claim 1 Storing the extracted forward wave sample; Comparing the magnitude of the forward wave with a predetermined threshold to determine an accident when the threshold is greater than or equal to the threshold; Obtaining a cross-correlation function value of the forward wave component and the backward wave component when it is determined that the accident is occurring; Obtaining a time at which the cross-correlation function value has a maximum value; And And determining a distance to a failure point based on the time having the maximum value. The method according to claim 1 or 2 The extracting of the forward and backward wave components may include: a modal conversion step of obtaining an actual mode of a traveling wave from the overlapping component; And And extracting the forward and backward wave components from the at least one of the actual mode signals. The method of claim 3 The finite impulse response filter, Expressed by the relation of Where k: data index; M: A method of determining the failure distance, which is the size of the filter window. The method of claim 3 The storing of the forward wave includes a failure distance determining method including at least one sample before the accident judgment. The method of claim 3 Determining the distance to the failure point is a time τ at which the correlation function has a maximum value To satisfy the relationship of Where L is the distance from the observation point to the failure point; v: propagation velocity in the line ( How to determine the fault distance.