JPH10319083A - Partial discharge measuring method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a partial discharge measuring method which accurately determine noise signals generated in the upper right quadrant and the lower left quadrant. SOLUTION: A ϕ-g-t pattern (where, ϕ is phase, q is electric charge, and t is time) is created from measured signals, the ϕ-q-t pattern is divided into two blocks with phase angle. Following statistical values are determined: (1) number of signals (n), (2) average generated phase angle Avg (ϕ), (3) gradient A of a minimum square line derived from recursion with equation ϕ=A×t+B, (4) standard deviation (S) of displacement from the minimum square line, (5) standard deviation of discharging electric charge/average electric charge s(q)/ Avg (q) (where (1) to (4) are corresponding to data divided into two blocks, and (5) is corresponding to data integrated from two blocks). When all or part of these statistical values are within the specified statistical value range, it is determined that partial discharge occurs.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a power cable line including a power cable and its accessories, a gas insulated line (GIL), and the like, and a partial discharge measuring method for the accessories and its power devices.

[0002]

2. Description of the Related Art In partial discharge measurement of power cables,
When it is determined whether the measured signal is a partial discharge signal or a noise signal, φ-q− is determined from the amount of discharge charge (q), potential application phase (φ), and occurrence frequency per unit time (n) of the measurement signal.
A method has been proposed in which n patterns are created, and the φ-qn pattern is input to a neural network in which the φ-qn pattern of the partial discharge pulse has been learned in advance to determine the presence or absence of a partial discharge. (For example, see JP-A-5-25
No. 6895).

[0003]

According to past research, the partial discharge signal has a phase angle of 0 ° of the applied voltage phase signal (sine wave).
It is known that it is easy to occur in the range of 90 to 90 degrees (first quadrant) and 180 to 270 degrees (third quadrant). FIG. 10 is a diagram showing an example of a φ-qn pattern of a partial discharge signal. The horizontal axis represents the potential application phase (φ), the vertical axis represents the partial discharge charge amount (q), and the frequency of occurrence per unit time. (N) is indicated by shading. In general, most of the noise signal has no correlation with the applied potential phase. Therefore, it is possible to discriminate the partial discharge from the partial discharge by the determination by the neural network using the φ-qn pattern of the above-described conventional method. However, the noise signal generated in the first quadrant and the third quadrant has the same phase as that of the partial discharge, so that there is a problem that the above-described conventional determination method erroneously determines the noise signal.

FIG. 11 shows an example of a φ-qn pattern of a noise signal generated in the first quadrant and the third quadrant, which is likely to be mistaken by the conventional determination method. In the example of FIG. 11, the noise signal is generated in the first quadrant and the third quadrant, and it is difficult to discriminate from the partial discharge signal φ-qn shown in FIG. The present invention has been made in consideration of the above-mentioned problems of the related art, and an object of the present invention is that it is difficult to discriminate by a conventional determination method using a neural network using a φ-qn pattern. An object of the present invention is to provide a partial discharge measurement method capable of accurately determining a noise signal generated in a first quadrant and a third quadrant.

[0005]

FIG. 12 shows a φ-qt pattern of the partial discharge signal shown in FIG. In the figure, the horizontal axis represents the potential application phase φ, the vertical axis represents time t, and the charge amount q
Is indicated by shading (or color). As shown in the figure, the phase angle at which the partial discharge occurs occurs randomly within a certain phase range. FIG. 13 shows a φ-qt pattern of the noise signal shown in FIG. As shown in the figure, the phase angle of the noise tends to be continuously generated in the same phase. When an attempt is made to make a determination using a conventional φ-qn pattern, measurement data is converted into a φ-qn pattern, so that information on temporal changes is collected as an occurrence frequency n. That is, the time-series information of the measurement signal is lost, and it is difficult to distinguish between partial discharge and noise.

Accordingly, in the present invention, statistical processing is performed from a graph of φ-qt using a φ-qt pattern in which time-series information of a measurement signal is not lost, and at least one The above statistical values are obtained, and when all or some of these statistical values are within a predetermined statistical value range, it is determined that partial discharge has occurred. For example, as shown in FIG.
The graph of -t is divided into two blocks by the phase angle, the number of signals per unit time: n, and the average generated phase angle: Avg
(Φ), slope of the least-square line regressed by equation φ = A × t + B: A, standard deviation of deviation from the least-square line: s, standard deviation of discharged electric charge / average electric charge: s (q) / Avg
The statistical values of (q) are obtained, and when all or some of these statistical values are within a predetermined statistical value range, it is determined that partial discharge has occurred. In the present invention, since the partial discharge is determined as described above, a noise signal generated in the first quadrant and the third quadrant can be accurately determined, and the partial discharge signal and the noise signal can be accurately determined. Can be well identified.

[0007]

BEST MODE FOR CARRYING OUT THE INVENTION 275 kVC laid on site
Noise was determined for the data of the noise signal measured on the V cable line by the method of the present invention described above. FIG. 2 shows a configuration example of the partial discharge determination device according to the embodiment of the present invention. In the figure, 1 is a power cable, 2a, 2b
Is an aerial terminal unit, and an AC voltage is applied to the power cable 1 from the power application transformer 12 via the aerial terminal unit 2a. In the cable line having the above configuration, the power transformer 1
1 and a partial discharge detector 4a between the aerial terminal 2a via a coupling capacitor 3 and a power cable 1
Is connected to the partial discharge detector 4b, and the partial discharge detector 4a,
The output of 4b was input to the partial discharge measuring device 6 via the balancer 5.

Then, the measurement signal obtained by the partial discharge measuring device 6 was input to the A / D converter 7 together with the application potential phase signal 8 of the AC voltage applied to the power cable 1. A
The / D converter 7 converts the measurement signal and the potential application phase signal 8 into digital signals for a preset time (t), and outputs the digital signals to the personal computer 9. The personal computer 9 converts the measurement signal into a discharge charge amount (q), and determines the relationship with the potential application phase (φ) by φ-q
It is obtained as a −t characteristic, and five types of statistical numerical values are calculated from the φ-qt pattern and output to the monitor 10.

FIG. 3 is a flowchart showing the outline of the partial discharge determination process in the present embodiment. In the present embodiment, it is determined whether the discharge is a partial discharge or a noise as follows. The personal computer 9 uses the t-axis in units of
The axis divides 360 ° into 100 to obtain a φ-qt pattern, and divides the φ-qt pattern into two blocks in the phase domain (360 °). The first of the two blocks is between -36 ° and 144 °, the second block is between 144 ° and
324 °. Fig. 4 shows φ- divided into two blocks.
The outline of qt pattern is shown. The straight line shown in the figure is a straight line obtained by regressing the φ-qt pattern by the least square method.

After dividing into two blocks as described above, as shown in FIG. 3, the following five kinds of statistical numerical values were obtained. Number of data: n Number of data in each block generated within time t Average generated phase angle: Avg (φ) Average value of generated phase angle of partial discharge signal in each block Minimum slope of linear line: A φ in each block -A slope A on a straight line φ = A × t + B obtained by regressing the qt pattern by the method of least squares (see FIG. 4). Standard deviation of deviation of φ from least square line: s Standard deviation of deviation of phase angle φ from least square line φ = A × t + B found in each block Standard deviation of discharge charge / average charge: s (q) / A
vg (q) The ratio between the standard deviation s (q) of the discharge charge amount and the average charge amount Avg (q) obtained for the data obtained by combining the two blocks.

After obtaining the five types of statistical values as described above, it was determined whether the discharge was a partial discharge or a noise under the following identification conditions as shown in FIG. 10 ≦ n ≦ 80 15 ≦ Avg (φ) ≦ 40 −0.5 ≦ A ≦ 0.5 2.0 ≦ s ≦ 15.0 0.3 ≦ s (q) / Avg (q) determining when the previously set whether to use the advance which statistical data, a case where the set statistical data meets the condition and the partial discharge, the case does not meet the noise.

The conditions for identifying the partial discharge are set using data of the actually generated partial discharge. In the present embodiment, data on a defect in which a treeing needle was inserted into the insulator from above the outer semiconductive layer of the 22 kVCV cable and a data on a trauma defect in which the outer semiconductive layer was damaged with a cutter knife were used. Five kinds of statistical values were calculated from the 49 partial discharge data as described above, and the existence range was obtained for each statistical value. FIG. 5 to FIG. 9 are diagrams showing statistical values for the above 49 data, and as shown in FIG.
Five types of statistical numerical values were obtained for the nine data, and from the obtained statistical numerical values, the partial discharge was determined when the value was in the range of the above, and the noise was determined in the other range.

Using the above-described partial discharge determination device, 2
On-site noise data of 75kVCV cable line 2745
A judgment was made on the matter. Table 1 shows the case where only the statistical values are used and the case where only the statistical values are used.
It is a table | surface which shows the result of having determined about 745 data. Table 1 also shows the result of determining the same noise data by the conventional partial discharge determination method using a neural network using the φ-qn pattern.

[0014]

[Table 1]

From Table 1, it can be seen that the method using statistical numerical values performs more accurate classification than the conventional method using a neural network. That is, 97.6% when only two types of statistical values are used,
A judgment rate of 98.1% was obtained when all the statistical values of the above were used, but it was 87.5% with the method using the conventional neural network. Therefore, the conventional φ-q-
The determination based on the statistical values according to the present invention makes it possible to distinguish between partial discharge and noise with higher accuracy than the determination based on the neural network using n patterns. In addition, this determination was made based on whether or not the statistical value is within a predetermined statistical value range.However, a fuzzy rule is set by applying a neuro-fuzzy theory or the like, and the determination is made by fuzzy inference. High judgment can be expected.

[0016]

As described above, in the present invention, it is difficult to discriminate by the conventional determination method using the neural network using the φ-qn pattern.
It is possible to accurately determine a noise signal generated in the quadrant and the third quadrant, so that it is possible to more accurately distinguish partial discharge from noise.

[Brief description of the drawings]

FIG. 1 is a diagram showing an outline of the present invention.

FIG. 2 is a diagram illustrating a configuration example of a partial discharge determination device according to an embodiment of the present invention.

FIG. 3 is a flowchart illustrating an outline of a partial discharge determination process according to the embodiment of the present invention.

FIG. 4 is a diagram showing an outline of a φ-qt pattern divided into two blocks.

FIG. 5 is a diagram showing the number n of signals of partial discharge data.

FIG. 6 shows an average phase angle Avg of partial discharge data.
FIG.

FIG. 7 is a diagram showing a slope A of a least square line of partial discharge data.

FIG. 8 is a diagram showing a standard deviation s of deviation of partial discharge data from a least square line.

FIG. 9 is a view showing standard deviation / average charge amount s (q) / Avg (q) of discharge charge amount of partial discharge data.

FIG. 10 is a diagram showing an example of a φ-qn pattern of a partial discharge signal.

FIG. 11 is a diagram illustrating an example of a φ-qn pattern of a noise signal generated in a first quadrant and a third quadrant;

FIG. 12 is a diagram illustrating an example of a φ-qt pattern of the partial discharge signal of FIG. 10;

FIG. 13 is a diagram illustrating an example of a φ-qt pattern of the noise signal in FIG. 11;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Power cable 2a, 2b Air terminal part 3 Coupling capacitor 4a, 4b Partial discharge detector 5 Balancer 6 Partial discharge measuring instrument 7 A / D converter 8 Potential application phase signal 9 Personal computer 10 Monitor 11 Power supply transformer

Claims (1)

[Claims] 1. A partial discharge measuring method for measuring a partial discharge of a power cable and accessories, a power cable line, a power device, etc., and performing an insulation diagnosis, wherein the partial discharge is detected by a phase signal of an applied voltage and a partial discharge detector. From the obtained measurement signal, a discharge charge amount (q) and a generation phase (φ) of the measurement signal during a predetermined time (t) are obtained to create a φ-qt pattern, and a φ-qt pattern Statistical processing from at least 1
A partial discharge measurement method, wherein one or more statistical values are obtained, and when all or a part of these statistical values are within a predetermined numerical range, a partial discharge is determined.