JPH05107302A - Partial discharge detecting method for power line - Google Patents

Abstract

PURPOSE:To provide a partial discharge pulse detecting method with a raised S/N by effectively using an averaging process of signals. CONSTITUTION:Partial discharge pulse signals are concurrently detected at two different frequencies, and the pulse signal detected at a second frequency higher than a first frequency is averaged by using a pulse signal detected at the first frequency. Whether there is a partial discharge pulse or not is judged by means of the thus averaged signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting a partial discharge on a power line, and more particularly to a method for accurately detecting a partial discharge on a power line which has a large amount of noise and whose fluctuations are large.

[0002]

2. Description of the Related Art A method of electrically detecting a partial discharge pulse signal generated in a power line and averaging simultaneously detected acoustic pulse signals with the electrically detected pulse to enhance the detection sensitivity of the acoustic pulse signal. Is known.
-6742).

[0003]

However, since the attenuation of the acoustic pulse signal is more significant than that of the electrical pulse signal, the above-mentioned method uses the averaging process when the position of the detecting element is separated from the signal source by a certain amount or more. No improvement in sensitivity is obtained.
The propagation speed of the acoustic pulse signal is 340 m / se in air.
c, which is 6 orders of magnitude slower than 3 × 10 ⁸ m / sec of the electric pulse signal. Therefore, a large time difference occurs between the signal electrically detected and the signal acoustically detected during signal processing. In addition, the frequency characteristic of the acoustic signal element is limited, and if a steep pulse is also detected by the acoustic signal element, it becomes blunt. Further, what is detected by the acoustic signal element is an acoustic pulse signal that is reflected back and forth in a complicated manner in the sample system. These factors cause the signal waveform to change significantly.
For these reasons, it is actually difficult to process electrical signals and acoustic pulse signals with the same signal system. An object of the present invention is to realize a partial discharge pulse detection method in which the signal averaging process is effectively used to enhance the S / N ratio.

[0004]

In order to realize a partial discharge pulse detection method in which the signal averaging process is effectively used and the S / N ratio is increased, the present invention uses partial discharge pulse signals with two different values. Detected at the same frequency, the pulse signal detected at the first frequency is used to average the pulse signal detected at the second frequency higher than the first frequency, and the pulse signal detected at the second frequency is detected. In addition, the detection sensitivity of the partial discharge pulse signal is improved.

The first frequency is preferably chosen at a relatively low frequency, generally around 10 MHz, where external noise dominates, and the second frequency is relatively at which internal noise generally dominates over external noise. High frequency, eg 200
It is preferable to select at least MHz.

Individual detectors may be used to detect the partial discharge pulse signals at two different frequencies at the same time. However, the detectors may be detected by one detector and the detected signals may be frequency-separated. You may each detect the component of two frequencies. If separate detectors are used, they may be heterogeneous detectors (eg, tuned pulse detector and wideband pulse detector).

The phrase "averaging the pulse signal detected at the second frequency using the pulse signal detected at the first frequency" means that a predetermined signal is detected in the signal detected at the first frequency. Whenever a pulse signal exceeding the level is detected, this pulse signal is used as a trigger pulse to capture the pulse signal detected at the second frequency, and this is repeated N times, and the captured detection signal at the second frequency is captured. Is added to perform an averaging process (the arithmetic average is calculated by the number of additions).

[0008]

The partial discharge pulse signal is simultaneously detected at two different frequencies, and the pulse signal detected at the first frequency is used to average the pulse signal detected at the second frequency to obtain a signal. It is possible to reduce the noise component contained in and to improve the S / N ratio (in an ideal case, 1 / (N) ^1/2 is obtained when the number of averaging is N).

[0009]

EXAMPLES The present invention will be described more concretely with reference to the following examples. [Embodiment 1] Insulation diagnosis of a power line was performed by the circuit configuration shown in FIG. That is, in the insulating connection portion 2 of the power cable 1, a pair of metal foil electrodes 3a are fixed to the outside of the insulating sheaths 2b on both sides of the insulating tube 2a, a detection impedance 4 is connected between them, and both ends thereof are tuned high frequency. It was connected to measuring instruments 5a and 5b. The outputs of the tunable high-frequency measuring instruments 5a and 5b are the averaging processing devices 6a and 6 respectively.
b and their outputs are connected to the oscilloscope 7. Another pair of metal foil electrodes 3b was fixed to the insulating connection portion 2, and the calibration oscillator 8 was connected thereto. The power cable 1 is a 275 kV power cable.

The operation of the above configuration will be described below. When a partial discharge pulse is generated in the insulation connection portion 2 of the cable 1, a pulse voltage is generated between the pair of metal foil electrodes 3a provided outside the insulation connection portion 2 and the potential difference between both ends of the detection impedance 4 is obtained. It is detected by the tunable high frequency measuring instruments 5a and 5b having different tuning frequencies. The tuning type high frequency measuring instruments 5a and 5b respectively output only the signals of the tuning frequency and are input to the averaging processing devices 6a and 6b. These signals are added and averaged by the averaging processing devices 6a and 6b for a predetermined period (for example, until one signal is detected a predetermined number of times) or a predetermined number of times. The result is displayed on the oscilloscope 7.

An AC voltage of 159 kV is applied to the power cable 1, a pulse signal of 11 MHz is detected by the tunable high frequency measuring instrument 5a, and a pulse signal of 230 MHz is detected by the tunable high frequency measuring instrument 5b, and the averaging process is performed. With the devices 6a and 6b, the pulse detected by the tunable high frequency measuring instrument 5a was used to average the pulse detected by the tunable high frequency measuring instrument 5b. The averaging process is a tuned high frequency measuring instrument 5
The number of pulses detected in a is the number of times shown in FIG.
It was performed by adding the pulses detected by the tunable high frequency measuring instrument 5b during the input (48 times). The output was observed with the oscilloscope 7. Tuning type high frequency measuring instrument 5a
As for the pulse detected in, 64 times by itself, 20
The averaging treatment was performed 48 times. The result is shown in FIG.

The locus a in FIG. 2A is a tuned high frequency measuring instrument 5.
a non-averaged signal with a frequency of 11 MHz detected by a is shown, and the trace b is a tuned high frequency measuring instrument 5b.
Shows an un-averaged signal with a frequency of 230 MHz detected by FIG. 2B shows the result of averaging each signal while the signal detected at the frequency of 11 MHz was counted 64 (2 ⁶ ) times, and FIG. 2C shows the result of averaging 2048 (2 ¹¹ ) times. Track a has a frequency of 11M
The locus b shows a signal detected at Hz and averaged with the detected pulse, and a locus b shows a signal detected at a frequency of 230 MHz and averaged with a pulse detected at a frequency of 11 MHz. The latter (trajectory b) has its amplitude enlarged 200 times that of the former (trajectory a).

The signal detected at a frequency of 11 MHz (trajectory a) has several peaks when it is not averaged, but it is added every time a pulse is detected, and it is repeated 64 times and averaged. , Except for the pulse indicated by C near the center. In the averaging of 2048 times, the distribution other than the pulse C is extremely flat.

The signal measured at a frequency of 230 MHz (trajectory b) contains a large number of small pulses due to internal noise when not averaged, but is added every time a pulse is detected at 11 MHz and repeated 64 times. And averaging significantly reduces the amplitude of many small pulses,
In the averaging of 2048 times, almost all the signals except the signal pulse indicated by D near the center disappear. The noise of frequency 230 MHz is mostly thermal noise and Schottky noise due to the measuring instrument itself, and since these generally have irregular waveforms and periods, the effect of averaging greatly appears (in an ideal case). The theoretical value of the averaging effect is 1 /
(N) ^1/2 ).

FIG. 2C, which is the result of 2048 averagings.
It is highly reliably determined that the pulse C and the pulse D of 1 are pulses based on partial discharge. Even if the noise changes considerably during the measurement, the occurrence of partial discharge can be identified if a peak is observed in the signal which is detected at 230 MHz and averaged.

In the present example, the first frequency is selected to be 11 MHz and the second frequency is selected to be 230 MHz, but the selection of the frequency for such measurement and averaging is made in consideration of the frequency distribution of the partial discharge signal and noise. It FIG. 3 qualitatively shows the frequency distribution of the partial discharge signal and the noise. Generally, both external noise and internal noise decrease with increasing frequency, whereas external noise is dominant at low frequencies,
Internal noise dominates at high frequencies. This tendency is found not only in electrical signals but also in acoustic signals and optical signals.

[Embodiment 2] In Embodiment 1, electric pulses generated by two pairs of metal foil electrodes are detected at two frequencies and averaged at different frequencies. In this embodiment, two acoustic microphones are used. The acoustic signals were detected at each of the four frequencies, and they were averaged using the signals detected at the lower frequencies. That is, as shown in FIG. 4, acoustic microphones 21a and 21b are provided outside the insulating sheath 2b in the insulating connection portion 2 of the cable 1. The output of the acoustic microphone 21a is connected to the amplifier 22a via the filter 71a, and the output of the acoustic microphone 21b is connected to the amplifier 22b via the filter 71b.

The output of the amplifier 22a is the averaging processor 23.
The output of the amplifier 22b is connected to the averaging processor 23b. The output of the amplifier 22a is also connected to the trigger input terminal of the averaging processor 23b. Each output of the averaging processing devices 23a and 23b is the oscilloscope 7
It is connected to the.

Amplifier 22a by averaging processor 23a
The output of is averaged in the same manner as in FIG. The averaging processing device 23b averages the output of the amplifier 22b by the output signal of the amplifier 22a. When the latter does not exist, the averaging processing device 23b does not operate even if there is the output of the amplifier 22b.

[0028]

According to the present invention, the electrical discharge or the acoustic signal detected at the first frequency and the second frequency higher than the first frequency is averaged by the former to obtain S / S for partial discharge detection. The N ratio can be effectively increased. It is possible to avoid the influence of the significant deformation of the waveform of the signal detected by the acoustic signal detection element, such as the case where the electrical signal and the acoustic signal are averaged by the same signal system.

[Brief description of drawings]

FIG. 1 is an explanatory diagram of a circuit configuration used in an embodiment of a partial discharge detection method according to the present invention.

2A, 2B and 2C show signal waveforms observed in an embodiment of the partial discharge detection method according to the present invention.

FIG. 3 is a graph qualitatively showing frequency distributions of a partial discharge signal and noise.

FIG. 4 is an explanatory diagram of a circuit configuration used in another embodiment of the partial discharge detection method according to the present invention.

[Explanation of symbols]

1 power cable 21
Acoustic microphone 2 Insulated connection parts 21a, 21b
Acoustic microphone 2a Insulation tube 22
Amplifier 2b Insulation sheath 22a, 22b
Amplifier 3 A pair of metal foil electrodes 23a, 23b
Averaging device 3a, 3b A pair of metal foil electrodes 71a, 71b
Filter 4 Detection impedance 5 Tuning type high frequency measuring instrument 5a, 5b Frequency analyzer 6 Averaging processor 6a, 6b Averaging processor 7 Oscilloscope 8 Calibration oscillator

Claims (4)

[Claims] 1. A pulse signal detected at the first frequency by simultaneously electrically or acoustically detecting a partial discharge pulse signal generated in a power line at a first frequency and a second frequency higher than the first frequency. By using, to average the pulse signal detected at the second frequency, to determine the presence or absence of partial discharge occurrence from the signal subjected to the averaging process, the partial discharge detection method of the power line .. 2. The partial discharge detection of a power line according to claim 1, wherein the first frequency is set to a frequency where external noise is dominant, and the second frequency is set to a frequency where thermal noise and Schottky noise are dominant. Method. 3. The method according to claim 1, wherein the first and second frequencies are in a radio frequency range. 4. The method for detecting partial discharge in a power line according to claim 1, wherein the first and second frequencies are in an ultrasonic range.