CN102998593A - Method for accurately positioning electrical cable faults through interpolation convolution calculating - Google Patents

Abstract

A method for accurately locating a power cable fault location through interpolation and convolution calculations belongs to the technical field of electrical communication. Select waveform data x such as current wave, voltage wave, electromagnetic wave, etc. during the actual operation fault; use Matlab to draw a two-dimensional graph of the data; perform convolution calculation on the waveform data, and initially locate the fault location; interpolate the waveform data, and then perform Convolution calculation, precise fault location; waveform correlation formula:

, where R _xy is the value of the correlation between X and Y sequences, X _n and Y _n are each corresponding value in the sequence, m and n represent the number in the sequence, and Y ^* represents the number of values in the sequence Conjugate, E represents the expected value, and the point with the greatest correlation after interpolation is the precisely located fault point. In this way, through this method, the location of the fault point can be located more precisely without improving the hardware condition. The present invention can improve the accuracy of actual fault location without increasing the hardware sampling rate.

Description

A Method of Accurately Locating Power Cable Fault Locations Through Interpolation and Convolution Calculations

technical field

The invention relates to a method for accurately locating fault locations of power equipment through interpolation and convolution calculation, in particular to a method for locating discharge faults of distributed equipment such as GIS, power cables, and overhead lines, and belongs to the technical field of electrical communication.

Background technique

Power equipment, especially distributed equipment such as GIS, power cables, overhead lines, etc., will have a certain probability of discharge faults, such as power cables and power equipment, etc. will produce discharge faults.

Power cable fault location can be divided into two categories in principle: traveling wave method and impedance method. Due to the limitation of traveling wave sampling rate in the existing technology, the fault location is in a larger range. The present invention is mainly based on the traveling wave method, and proposes a method for improving detection precision by using convolution interpolation. However, this technology can greatly reduce the scope of faults and provide great help for quickly finding fault points.

Traveling wave distance measurement uses the time for traveling waves to go back and forth between the measurement point and the fault point, and the distance can be obtained after simple calculation. The traveling wave method is specifically divided into two categories: the first category is a method using a voltage traveling wave signal. The method is to make the cable fault point break down under the action of DC high voltage or pulse high voltage signal, and then measure the distance by observing the time when the discharge pulse goes back and forth between the observation point and the fault point. The second type is the ranging method using current traveling wave signal. The difference between this method and the former method is that the former uses a linear current coupler to measure the current signal generated when the cable fault breaks down, and the current waveform coupled by the sensor is also easier to distinguish. The method we adopt is current traveling wave for fault location measurement. Its ranging accuracy is basically not affected by factors such as line fault location, fault type, line length, grounding resistance, etc. The formula it is based on is mainly:

$l_1=\frac{L-(t_2-t_1)v}{2}$ and $l_2=\frac{L-(t_1-t_2)v}{2}$ ,

In the formula: l ₁ and l ₂ are the distances from the fault point to the two ends respectively; t is the time for the traveling wave to reach the two ends of the line. We can see from the formula that the key to finding the fault point is to find the time difference between the arrival of the current wave (ie t ₁ -t ₂ ). The waveforms arriving at both ends of the cable are actually current waves from the same fault point. Finding the maximum correlation point of the two waveforms reaching both ends is to find the maximum autocorrelation point of the same waveform. The formula for determining waveform correlation is . In the formula: R _xy is the value of the correlation between the two sequences of X and Y, X _n and Y _n are each corresponding value in the sequence, m and n represent the number in the sequence, Y ^* represents the value of the value in the sequence Conjugate, E stands for expected value.

It can be seen from the formula that the maximum autocorrelation point found is actually the number of points where the waveforms at both ends differ (if the sampling rate is 1Mhz, and the current wave propagation speed is 3×108m/s, then the distance between each two points L=V×T =V/f=300m). According to the number of difference points and the distance represented by the actual two points, the location of the fault point can be obtained.

Contents of the invention

The invention performs interpolation and convolution calculation after the signal is sampled, which can improve the accuracy of fault location and provide great help for accurate fault location.

A method for accurately locating a fault location of a power cable through interpolation convolution calculation, the steps of the method are as follows:

1. Select the waveform (current wave, voltage wave, electromagnetic wave, etc.) data x of the actual operation failure; use Matlab to draw a two-dimensional graph of the data;

2. Carry out convolution calculation on the waveform data, and initially locate the fault location;

3. Interpolate the waveform data, perform convolution calculation again, and accurately locate the fault location.

，其中：R_xy是X、Y两个序列相关性的值，X_n与Y_n是序列中对应的每一个数值，m与n代表序列中的第几个数，Y^*代表数列中数值的共轭，E代表期望值。从公式中可知，找到的自相关最大点其实就是两端波形到达相差的点数（若采样率为1Mhz，电流波传播速度为3×108m/s，则每两点间的距离L=V×T=V/f=300m）。根据相差点数以及实际两点间代表的距离，既可求出故障点位置。The waveforms arriving at both ends of the cable are actually current waves from the same fault point. Finding the maximum correlation point of the two waveforms reaching both ends is to find the maximum autocorrelation point of the same waveform. The formula for determining waveform correlation is

, where: R _xy is the value of the correlation between the two sequences of X and Y, X _n and Y _n are each corresponding value in the sequence, m and n represent the number in the sequence, Y ^* represents the number of values in the sequence Conjugate, E stands for expected value. It can be seen from the formula that the maximum autocorrelation point found is actually the number of points where the waveforms at both ends differ (if the sampling rate is 1Mhz, and the current wave propagation speed is 3×108m/s, then the distance between each two points L=V×T =V/f=300m). According to the number of difference points and the distance represented by the actual two points, the location of the fault point can be obtained.

The point with the greatest correlation after interpolation is the precisely located fault point. In this way, the sampling rate can be increased through this method, and the fault point can be located more accurately without improving the hardware condition.

The present invention can improve the accuracy of actual fault location without increasing the hardware sampling rate.

Description of drawings

Fig. 1 is a flow chart of the present invention.

Figure 2 is a time domain diagram of the current waveform data.

Fig. 3 is a diagram of the calculation results of waveform data convolution.

Fig. 4 is a data diagram after interpolation of the first group of data.

Figure 5. The calculation results of the correlation between the two groups of data after interpolation.

Detailed ways

The present invention will be further described below in combination with specific embodiments.

Fig. 1 is a flow chart of the present invention. A method for accurately locating the position of electric equipment through interpolation and convolution calculation, the steps of the method are as follows:

Select the current wave data x during actual operation (the sampling rate can be selected according to the actual situation of the equipment. If the sampling rate of the equipment is higher, the measurement accuracy will be higher. The data sampling rate listed below in this embodiment is 100Ms/s, and a set of data 40 points). Use Matlab to draw the first two-dimensional graph of the data, as shown in Figure 2. Figure 2 is a time domain diagram of the current waveform data. Figure 2 shows our sampling of continuous waveforms at the fault site, with a total of 40 points of sampling data. The abscissa is the number of points, and the ordinate is the current amplitude of each point.

(2) Carry out convolution calculations on two sets of data to find the maximum correlation point, as shown in Figure 3 (Figure 3 is the calculation result after using the xcorr function in matlab to calculate the correlation between the first set of data and the second set of data Fig. The abscissa is the relative position of the two sets of data, 0 points indicate that the two sets of data are completely corresponding, 1 point indicates that the two sets of data are staggered by one point; the ordinate is the correlation calculation value), the maximum convolution value is at the 40th point, That is, the two sets of data have the greatest correlation at the 40th point.

(3) Interpolate two sets of data (insert 4 points between each two points, this embodiment only inserts 4 points for convenience, the number of insertion points is selected according to the actual accuracy requirements in actual use. The more the number of insertion points, the more the measurement The higher the accuracy.), the first set of data after interpolation is shown in Figure 4, and Figure 4 is the data map after the first set of data interpolation (the abscissa is the number of points, and the ordinate is the current amplitude of each point),

First group

Second Group

(4) Carry out convolution calculation again on the two sets of data after interpolation, and the result is shown in Figure 5. Figure 5 shows the correlation calculation results of the two sets of data after interpolation, and the convolution value is the largest at the 201st point.

It can be seen from the calculation results that the correlation between waveforms can be accurately expressed through a reasonable sampling rate, and the sampling rate is greatly reduced at the same time.

In this way, increasing the sampling rate through software can make our positioning more accurate without improving hardware conditions.

In the second step, our sampling rate is 100Ms/s, if the location of the fault point is between the 40th point and the 41st point, the fault range is 3m. After the re-interpolation convolution calculation, the fault point is at the 200th point and the 201st point, and the fault range is 0.6m. This increases the sampling rate by a factor of 5

The present invention improves the accuracy of actual fault location through a calculation method without increasing the hardware sampling rate.

Claims (2)

1. A method for accurately locating a power cable fault position by interpolation convolution calculation, is characterized in that, the method steps are as follows: (1) Select the waveform data x during the actual operation failure; use Matlab to draw a two-dimensional graph of the data; (2) Carry out convolution calculation on the waveform data, and initially locate the fault location; (3) Interpolate the waveform data, perform convolution calculation again, and accurately locate the fault location; The waveforms arriving at both ends of the cable are current waves from the same fault point. Finding the maximum correlation point of the two waveforms reaching both ends is to find the maximum autocorrelation point of the same waveform. The formula for determining the waveform correlation is $R_{\mathrm{xy}}\left(m\right)=\mathrm{E.}\left\{x_{\mathrm{no}+m}{\mathrm{the\; y}}_{\mathrm{no}}^{*}\right\}=\mathrm{E.}\left\{x_{\mathrm{no}}{\mathrm{the\; y}}_{\mathrm{no}-m}^{*}\right\}$ , Among them, R _xy is the value of the correlation between the two sequences of X and Y, X _n and Y _n are each corresponding value in the sequence, m and n represent the number in the sequence, Y ^* represents the total number of values in the sequence Yoke, E represents the expected value; it can be seen from the formula that the maximum autocorrelation point found is actually the number of points where the waveforms at both ends reach the difference; according to the number of points of difference and the distance represented by the actual two points, the location of the fault point can be obtained; The point with the greatest correlation after interpolation is the precisely located fault point. 2. A method for accurately locating a power cable fault location through interpolation and convolution calculation according to claim 1, wherein the waveform is a current wave, a voltage wave or an electromagnetic wave.

Images