JP4139022B2 - How to remove external noise - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a method for removing external noise, and more particularly, to a method for removing external noise during partial discharge measurement of a power system.
[0002]
[Prior art]
As a method for removing external noise by digital signal processing when measuring partial discharge of electric power equipment, for example, Technical Report No. 593 of the Institute of Electrical Engineers of Japan or the 1997 Annual Conference of the Institute of Electrical Engineers of Japan, Energy No. Methods using digital filters, as reported in 655, are known.
[0003]
According to these reports, as a digital filter, an adaptive rejection filter, an adaptive predictor filter, a Kalman filter, or the like has been studied, and each has the following advantages.
[0004]
That is, the adaptive rejection filter has an advantage that it is effective for a continuous noise (PD) signal without changing the original partial discharge signal, and the adaptive predictor filter is a continuous filter. The Kalman filter has an advantage that it is effective against noise and can be applied to different noises. The Kalman filter has a short calculation time and a small change in the shape of the PD pulse. Further, it is highly sensitive to the selection of filter coefficients. It has the advantage of being.
[0005]
[Problems to be solved by the invention]
However, the noise removal method using a conventional digital filter simply suppresses continuous and periodic noise, requires knowledge about the signal and noise in advance, and requires a long calculation time in real time. However, there are various problems, such as the inability to perform the above process, the distortion of the original waveform of the detected signal, or the complicated process of selecting partial discharge and external noise.
[0006]
Accordingly, an object of the present invention is to provide a method for removing external noise that can remove external noise in real time by a simple method without distorting a signal waveform.
[0007]
[Means for Solving the Problems]
In order to achieve the above object, the present invention is a method of removing external noise components from a partial discharge measurement signal in a power cable line, power equipment, etc., and sampling the measurement signal at a predetermined sampling frequency. Accordingly, a predetermined number of digital data is generated in a plurality of sections set in the measurement cycle, and the weighted average value of the section corresponding to the measurement cycle two times before the digital data corresponds to the measurement cycle one before. An external noise characterized in that a threshold value is set based on a maximum value of a section, the digital data is compared with the threshold value, and the digital data larger than the threshold value is output as a partial discharge signal. The removal method of this is provided.
[0008]
The external noise removal method according to the present invention is performed, for example, as follows. First, a predetermined number of digital data is generated in a plurality of sections set for measurement sampling by sampling a measurement signal from a power cable line, a power device or the like at a predetermined sampling frequency.
[0009]
Next, in such a setting, the measurement cycle k (k = 1, 2, 3,...) Is divided into sections j (j = 1, 2, 3,...) Having a predetermined time width, and k cycles. The i-th digital data in the section j of the eye is P ^k (j · i), the weighted average value of the digital data in the section j of the k-th cycle is P _a0 ^k (i), and in the section j of the k-th cycle. When the maximum value of digital data is P _a1 ^k (j) and A, B, C, and D are constants,
Vth = A · P _a0 ^k−1 (j) + B (1)
However, P _a0 ^k (j) = C · P _a0 ^k−1 (j) + D · P _a1 ^k (j) (2)
Sets the threshold value Vth. Next, digital data satisfying P ^k (j, i)> Vth is selected and the others are set to zero, and the selected digital data is output as a partial discharge signal, thereby reliably removing external noise components. Will be.
[0010]
DETAILED DESCRIPTION OF THE INVENTION
Next, an embodiment of the external noise removal method according to the present invention will be described.
FIG. 1 shows an external noise elimination device, where 1 is a power cable line to be measured, 2 is an insulation connection part thereof, 3 is a foil electrode attached to the insulation connection part 2, and 4 is measured from a foil electrode 3. The detection impedance for detecting the signal, 5 indicates a tuning amplifier that tunes and amplifies the detected measurement signal at the optimum frequency.
[0011]
6 is an A / D (analog / digital) converter for sampling the measurement signal tuned and amplified in the tuning amplifier 5 at a predetermined sampling frequency for A / D, and 7 is a signal for removing external noise by signal processing. A processing circuit 8 is a monitor personal computer that displays and records the partial discharge signal from which external noise has been removed.
[0012]
FIG. 2 shows a circuit for obtaining the optimum frequency in the tuning amplifier 5. Foil electrodes 3 and 3 ′ are affixed to the insulating connection 2. A calibration pulse oscillator 9 is connected to the foil electrode 3 ′, and a spectrum analyzer 10 is connected to the detection impedance 4.
[0013]
An insulating cylinder 11 for cutting the conductive outer sheath of the power cable line 1 is provided at the central position of the insulating connection portion 2, and the calibration pulse oscillator 9 is passed through the foil electrodes 3 ′ on both sides of the insulating cylinder 11. A signal pulse for calibration is injected from.
[0014]
The detection impedance 4 detects a signal output based on the injection of the calibration signal pulse and sends it to the spectrum analyzer 10. The spectrum analyzer 10 analyzes the frequency of the detection signal according to the following procedure.
[0015]
(1) A signal having a predetermined magnitude such as 10 pC is injected from the calibration pulse oscillator 9, and the frequency spectrum S (f) of the detection signal at this time is obtained.
(2) The calibration pulse oscillator 9 is stopped, and the frequency spectrum N (f) of the noise signal detected in a state where no signal is injected is obtained.
(3) The frequency at which the ratio between the frequency spectrum S (f) of the detection signal and the frequency spectrum N (f) of the noise signal is maximized is selected as the optimum frequency. Usually, the spectrum analyzer 10 has an SN ratio calculation function, which is used.
[0016]
FIG. 3 shows a specific example of the noise removal method.
FIG. 3A shows a measurement signal detected by the detection impedance 4 and sent to the A / D converter 6 via the tuning amplifier 5 and digitally sampled at the A / D converter 6 at a rate of 2 Msample / second. A part of the digital data 12 generated at the time of execution is shown, which corresponds to one period of a commercial frequency of 50 Hz (one cycle of 20 ms), divided into 200 sections, and each section is set to 100 ms.
[0017]
FIG. 3B shows the weighted average values [P _a0 ^k−2 (j)] and (k−1) cycles up to two cycles before this cycle, that is, the corresponding section of the (k−2) cycle. The threshold value Vth for each digital data of each section calculated from the maximum value [P _a1 ^k−1 (j)] of the corresponding section is shown. The signal circuit 7 calculates each threshold value Vth. Do. Here, in equations (1) and (2), A was set to 1.0, B was set to 200 mV, and C and D were set to 0.5.
[0018]
In the signal processing circuit 7, the digital data 12 [P ^k (j · i)] is compared with the threshold value Vth, and the digital data 12 satisfying P ^k (j, i)> Vth is output to the monitor personal computer 8, Other digital data is zeroed.
[0019]
FIG. 3C shows the result of selecting only a digital data signal larger than the threshold value Vth, and there are a total of ten partial discharge signals 13 exceeding the threshold value Vth. The partial discharge signal 13 is displayed and recorded on the monitor personal computer 8. When the partial discharge signal 13 is output, digital data having a predetermined time width before and after the partial discharge signal 13 is also added.
[0020]
【The invention's effect】
As described above, according to the external noise removal method of the present invention, digital data is generated in a plurality of sections set in a measurement cycle by sampling a measurement signal from a power cable line or the like at a predetermined frequency. The threshold value is set based on the weighted average value of the interval corresponding to the measurement cycle two times before the digital data and the maximum value of the interval corresponding to the measurement cycle immediately before the digital data. Since the digital data larger than the threshold value is output as the partial discharge signal by comparing the two, the noise removal method is simple, and real-time noise removal can be performed without distorting the waveform.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram showing a measuring instrument used in an embodiment of a method for removing external noise according to the present invention.
FIG. 2 is a circuit diagram used to select an optimum frequency in the tuned amplifier of FIG. 1;
FIGS. 3A and 3B are explanatory diagrams of signals in the embodiment of the external noise removing method according to the present invention, wherein FIG. 3A is digital data generated by digital sampling from a measurement signal, and FIG. 3B is digital data of FIG. (C) shows the partial discharge signal extracted from the digital data and threshold value comparison of (b).
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Power cable line 2 Insulation connection part 3, 3 'Foil electrode 4 Detection impedance 5 Tuning amplifier 6 A / D converter 7 Signal processing circuit 8 Personal computer for monitor 9 Calibration pulse oscillator 10 Spectrum analyzer 11 Insulating cylinder 12 Waveform 13 Partial discharge Signal Vth threshold

Claims (3)

In the external noise removal method for removing external noise components from the measurement signal of partial discharge in power cable lines, power equipment, etc.,
Sampling the measurement signal at a predetermined sampling frequency to generate a predetermined number of digital data in a plurality of sections set in a measurement cycle,
A threshold is set based on the weighted average value of the interval corresponding to the previous measurement cycle of the digital data and the maximum value of the interval corresponding to the previous measurement cycle;
Comparing the digital data with the threshold value;
A method for removing external noise, wherein the digital data larger than the threshold value is output as a partial discharge signal. When the threshold value is Vth, the setting is that the measurement cycle k (k = 1, 2, 3,...) Is a section j (j = 1, 2, 3,...) Having a predetermined time width. Divided into
The i-th digital data in the interval j of the k-th cycle is P ^k (j · i),
The weighted average value of the digital data in the section j of the k-th cycle is P _a0 ^k (j),
The maximum value of the digital data in the section j of the k-th cycle is P _a1 ^k (j)
When A, B, C, and D are constants,
Vth = A · P _a0 ^k−1 (j) + B,
However, P _a0 ^k (j) = C · P _a0 ^k−1 (j) + D · P _a1 ^k (j)
Made by
2. The method according to claim 1, wherein the partial discharge signal is output by selecting an output of the digital data that satisfies P ^k (j, i)> Vth. The partial discharge signal is output by adding digital data having a predetermined time width before and after the digital data to the digital data of P ^k (j, i) in the relational expression. 3. A method for removing external noise according to item 2.

Images