JPH05157796A - Partial discharge measuring apparatus - Google Patents

Abstract

PURPOSE:To enable discrimination of a discharge pulse from a noise for each of detection outputs containing the discharge pulse and the noise in the measurement of a partial discharge. CONSTITUTION:This apparatus is provided with a plurality of bandpass filter circuits 10 different in the passage area into which a detection output P is inputted containing a discharge pulse and a noise in the measurement of a partial discharge and a trigger signal generation circuit 11 to detect the presence of the detection output P. A discrimination circuit 12 is provided and compares outputs of the bandpass filter circuits 10 with an output of the trigger signal generation circuit 11 to judge whether the detection output P is a discharge pulse or a noise from frequency component outputs of the filter circuits compared. Each time the detection output P is applied to both the circuits 10 and 11, the judgment of whether the detection output P is the discharge pulse or noise is performed for each of the detection outputs P.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a partial discharge measuring device used for an insulation test of power equipment such as a power cable and a transformer and for measuring the insulation test.

[0002]

2. Description of the Related Art Conventionally, partial discharge measurement has been carried out as a conformance test or a soundness confirmation test of electric power equipment such as electric power cables and transformers, and has become a standard test item for electric power equipment.

For example, for the purpose of preventing insulation breakdown during power transmission, it is also used for an insulation survey for inspecting the insulation state of a line regularly or irregularly, and in that case, the measurement is shown in FIG. It is performed by the measuring circuit.

That is, a high voltage is applied to the test cable C via the blocking coil L, and the applied high voltage causes a slight voltage change due to partial discharge (corona) generated at the defective portion of the cable C. A pulsed voltage change is extracted by the coupling capacitor CR and the detection element S connected to the coupling capacitor CR, and the magnitude of the pulse voltage of the discharge pulse, the frequency of occurrence, etc. are extracted from the extracted discharge pulse, and the attenuator 1 is used. , An amplifier 2, a count rate meter 3, and an oscilloscope 5 are used for measurement.

At this time, generally, a discharge pulse (partial discharge)
Is represented by the apparent discharge charge Qa defined by the product of the discharge start voltage Ve of the defective portion and the capacitance Ca in series with the defective portion, and the discharge charge amount Qa = Ca × Ve and The number of occurrences depends on the size and shape of the defective portion of the cable C, the applied voltage, and the content rate of the defective portion of the cable C. Therefore, the insulation property is judged by the discharge charge amount at a certain occurrence frequency.

By the way, the detection output of the partial discharge measurement includes noise induced in the line other than the discharge pulse, for example, pulsed noise.

Therefore, a weak discharge (1PC to several tens of P
In the partial discharge measurement for measuring C), it is necessary to prevent the noise voltage (pulse noise) from entering the partial discharge measuring instrument as much as possible, and for example, narrow the amplification band of the amplifier 2 of the measuring instrument or external noise. To select a band with less
As another method, there has been tried a method in which the distribution data of the magnitude or the distribution of the generated phase is taken for all the pulses at a certain time, and the discharge pulse and the noise are separated by the pattern of the distribution or the change of the pattern. There is.

[0008]

However, in the case of the above-mentioned measurement method in which the noise processing is performed, in the former case, the noise component in the output output from the amplifier is completely zero even if the band of the amplifier is limited. Therefore, the discharge pulse and noise are mixed in the detection output output from the amplifier. Therefore, for example, at the time of measurement, a skilled observer must observing the oscilloscope while adjusting the trigger level of a counting rate meter, a pulse counter, etc. to empirically discriminate the discharge pulse from the noise.

Further, in the latter case, since the statistical processing is performed, it is difficult to discriminate the discharge pulse when the pulse noise is much larger than the discharge pulse.
There is a problem in that the surrounding measurement environment must be stable because the two cannot be distinguished unless the pulse noise during measurement has a periodical generation pattern.

Therefore, an object of the present invention is to make it possible to discriminate the discharge pulse from the pulse noise for each detection output without any trouble in experience or statistical processing.

[0011]

In order to solve the above problems, according to the present invention, a plurality of frequencies each having a different pass band are input to which the detection output of the partial discharge measurement including the discharge pulse and the noise is input. A component analysis filter circuit, a trigger signal generation circuit that detects the presence or absence of the detection output, and a trigger output that detects the detection output is output from the trigger signal generation circuit. The outputs of the respective frequency component analysis filter circuits output within the period are compared, and the frequency spectrum component ratio contained in the detection output of the partial discharge measurement is analyzed from the compared respective filter circuit outputs, and the analysis is performed. Based on the result, the detection output discriminates whether it is a discharge pulse or noise, and a discrimination circuit for outputting the discrimination result is provided.

The number and types of the frequency component analyzing filter circuits and the setting of the pass band of each filter circuit are as follows.
It is appropriately determined by experience and experiments based on the frequency spectrum of the discharge pulse and noise generated by the measurement target.

The trigger signal generating circuit may also be used as a frequency component analyzing filter circuit.

[0014]

In the partial discharge measuring device configured as described above,
Focusing on the difference in the spectrum components of the frequency spectrum of the discharge pulse and the frequency spectrum of the noise, the two are distinguished by the difference in the spectrum components.

That is, for example, when the partial discharge of a CV (power) cable is measured, the frequency spectrum characteristic of the pulse noise of the detection output output at that time is several MHz.
With the above, it attenuates rapidly. On the other hand, since the frequency spectrum characteristic of the discharge pulse becomes flat up to about 10 MHz, the output of each frequency component analysis filter circuit when the detection output is detected by the trigger signal generation circuit is set to the propagation time of each output. Since it depends on the frequency, after sampling for a certain period of time, the maximum values of the sampled output values are compared, and from the comparison result, all the filters up to about 10 MHz in the pass band in the frequency component analysis filter circuit are compared. When the analysis output is output from the circuit, the detection output is a discharge pulse. On the other hand, the output of the filter circuit having a pass band within several MHz is large, and the output is several M.
When the output of the ones above Hz is small, the detected output is discriminated from noise. In this way, the detection output of the partial discharge measurement is individually discriminated to determine whether the output is a discharge pulse or noise.

[0016]

Embodiments of the present invention will be described below with reference to the drawings. The partial discharge measuring apparatus shown in FIG. 1 includes four frequency component analyzing filter circuits 10 to which a partial discharge measurement detection output P including a discharge pulse and noise is input, and a trigger for detecting the presence or absence of the detection output P. It comprises a signal generation circuit 11 and a discrimination circuit 12 for discriminating whether the detection output P is a discharge pulse or pulse noise.

A band pass filter circuit is used as the frequency component analysis filter circuit 10, and the pass band and the cutoff (center) frequency of each filter circuit are different. The values of the pass band and the cutoff frequency are, for example, CV (power) even if the discharge charge amount is the same.
Since the frequency spectrum of the discharge pulse and the pulse noise may be different depending on the length and type of the cable or the like, it is appropriately set for each measurement object prior to the measurement. Also, the selectivity of each filter circuit 10 is appropriately determined at that time.

The trigger signal generating circuit 11 is a bandpass filter circuit having a pass band that passes all the frequency components of the detection output P and blocks noise components other than the detection output P, and the filter circuit 11 thereof is used. For example, even if a commercial frequency component (hum noise) or the like is input, the signal is not output, and when the detection output P is input, the input detection output P is output to the determination circuit 12 as a trigger pulse. Note that the trigger signal generation circuit 11 can be omitted depending on conditions, and in that case, the detection output P is directly input to the determination circuit 12.

The discrimination circuit 12 comprises a maximum value detection circuit 13, a gate circuit 14, a maximum value holding circuit 15 and a comparison circuit 16.

The maximum value detection circuit 13 is formed of, for example, a rectifier circuit or a low-pass filter circuit, and is provided for each frequency component analysis filter circuit 10. The maximum value detection circuit 13 converts the output of the frequency analysis filter circuit 10 into a DC level signal.

The gate circuit 14 is, for example, a digital IC.
When the trigger output T of the detection output P that has passed through the trigger signal generation circuit 11 is input in the pulse width control circuit that is a combination of the above gate circuits, a sampling signal SG having a constant width is generated at the rising of the trigger output T. ..

As the maximum value holding circuit 15, a peak detection circuit using an OP amplifier or the like is used, and it is provided for each maximum value detection circuit 13. This maximum value holding circuit 15
Holds the maximum value of the output of the maximum value detection circuit 13 when the sampling signal SG is output.

The comparison circuit 16 compares the maximum values of the maximum value holding circuits 13 and determines whether the detection output P is a discharge pulse or noise. For this determination, for example, the comparison circuit 1
In FIG. 6, digital processing may be performed using an A / D converter and a microprocessor.

This embodiment is configured as described above. Now, when the detection output P of the partial discharge measurement is inputted to this device, the inputted detection output P is the trigger signal generating circuit 1
1 and the frequency component analysis filter circuit 10, and each frequency component analysis filter circuit 10 outputs a frequency component corresponding to the pass band of each filter circuit. This output is successively converted to a DC level by the maximum value detection circuit 13. On the other hand, the detection output P input to the trigger signal generation circuit 11 is output from the trigger signal generation circuit 11 to the gate circuit 14 as the trigger output T. The gate circuit 14 generates a sampling signal SG having a constant width from the output, and the generated sampling signal SG has a maximum value holding circuit 15
Output to. The maximum value holding circuit 15 samples the output of the maximum value detection circuit 13 with the sampling signal SG,
The peak value is retained. Here, the sampling signal SG
Is a constant width because the frequency component analysis filter circuit 1
This is because each output of 0 causes a phase delay due to a frequency component, and thus it becomes difficult to detect a peak value when the sampling signal SG width is narrowed.

Each maximum value holding circuit 1 held in this way
The maximum value of the frequency component of the detection output P of 3 is the comparison circuit 16
Is input to and compared, and it is determined whether the detection output P is a discharge pulse or noise.

At this time, the determination is made by the output of each component analysis filter circuit 10, that is, the magnitude of the frequency spectrum output by each component analysis filter circuit 10. The pattern of the magnitude relation is measured. It is necessary to accumulate a certain amount of measured data.

For example, when the measurement target is a CV cable, as described above, the spectral characteristic of the discharge pulse measured in the connection box of the cable has a flat characteristic up to about 10 MHz, while mainly from the terminal. The spectral characteristics of intruding external noise are often sharply attenuated at several MHz. Therefore, based on the difference in the spectral characteristics of each pulse, the outputs of the component analysis filter circuits 10 are relatively compared with each other to determine each pulse.

Further, when the detected output P determined is a discharge pulse, the comparison circuit 16 outputs a pulse output corresponding to the discharge pulse.
The discharge frequency of the discharge pulse can be measured by arranging the pulse output by the pulse counter 4 after the amplifier 2 of the conventional partial discharge measuring instrument shown in FIG. When the amplitude of the detection pulse when this pulse output is output is measured with the oscilloscope 5 or the like, the voltage value of the discharge pulse can also be detected. At this time, the comparison circuit 16 outputs the reset signal R to the maximum value holding circuit and prepares for input of the next detection output P.

As described above, in this device, the discharge pulse and the noise are detected from the detection output, which is a mixture of the discharge pulse and the noise of the partial discharge measurement, and the individual output is obtained by the frequency analysis without using the experience or the statistical method. The detection output P is discriminated.

Next, in order to verify the effect of this device, FIG.
As shown in (1), a test line 25 having a terminal 21 and a connection box 23 provided on the CV cable 20 was manufactured, and the device was connected to the connection box 23 of the test line 25 to perform a verification test.

At this time, as shown in FIG. 3, four band pass filters are used for each frequency component analysis filter circuit 10, and the cutoff (center) frequencies of each filter 10 are 3 MHz and 5 MHz, respectively. 8MHz, 13MHz
And Further, in this device, a part of the output of the frequency component analysis filter circuit 10 having a cutoff frequency of 3 MHz is used as an input signal to the gate signal generation circuit 14, and the frequency component analysis filter circuit 10 also serves as the trigger signal generation circuit 11. .. The other circuits used are the same as those described above. Then, a simulation pulse as a discharge pulse and a simulation pulse imitating pulse noise were alternately injected into the connection box 23 of the test line 25, and the discrimination result was examined.

At this time, an algorithm based on the discriminant of Formula 1 was used for the discrimination in the comparison circuit 16.

[0033]

[Equation 1]

That is, the value of x in the equation (1) is larger than the denominator in the numerator of the equation (1) having only the pass band up to several MHz because the frequency spectrum of the noise exhibits an attenuation characteristic at several MHz. Become. On the other hand, in the case of the discharge pulse, since the frequency spectrum is flat up to 10 MHz, x in equation (1)
The value of is smaller than that of noise. For this reason, in this measuring apparatus, the internal parameters of the comparison circuit 16 are set so that, for convenience, x ≧ 2, noise x <2, and discharge pulse.

As a result, when the simulated pulse was injected into the junction box 23 and when the simulated pulse was injected into the terminal portion 21, no error occurred in the discrimination output for discriminating between the two output from the discrimination circuit 12. It was also found that the discriminant used in the comparison circuit 16 has no error.

In this device, the so-called data acquisition time from the activation of the gate circuit 14 to the end of the discrimination until the detection of the next detection pulse is about 8 μsec.
Met. Therefore, commercial frequency (50Hz or 6
It was possible to distinguish more than 1000 detection pulses per cycle (0 Hz) in real time. For this reason, for example, this device is placed in a required location of a high-voltage power transmission line, constantly monitors the insulation state of the transmission line during power transmission, and from the monitoring result, an abnormality monitoring device that prevents a breakdown accident of the transmission line in advance. Can also be used as

[0037]

Since the present invention is configured as described above, it is possible to discriminate in real time whether each of the detection outputs of partial discharge measurement is a discharge pulse or noise.

Therefore, partial discharge measurement is possible even in an environment with a lot of external noise such as in a real field.

Therefore, it is effective to use it for a soundness test or abnormality monitoring of electric power equipment such as an electric power cable.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment.

FIG. 2 Action diagram

FIG. 3 is a block diagram showing another embodiment.

FIG. 4 is an operation diagram of a conventional example.

[Explanation of symbols]

 10 Frequency Component Analysis Filter Circuit 11 Trigger Signal Generation Circuit 12 Discrimination Circuit T Trigger Output P Detection Output

Claims (1)

[Claims] 1. A plurality of frequency component analysis filter circuits, each having a different pass band, into which a detection output of partial discharge measurement including a discharge pulse and noise is input, and a trigger signal generation circuit for detecting the presence or absence of the detection output. When a trigger output that detects the detection output is output from the trigger signal generation circuit, the outputs of the frequency component analysis filter circuits that are output within a fixed period after the trigger output is output are compared. , The frequency spectrum component ratio contained in the detection output of the partial discharge measurement is analyzed from the output of each of the compared filter circuits, and it is discriminated whether the detection output is a discharge pulse or noise from the analysis result, and the discrimination result thereof. A partial discharge measuring device having a discrimination circuit for outputting.