JP2002090413A - Diagnostic apparatus for insulation abnormality of high- voltage device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a diagnostic apparatus, for the insulation abnormality of a high-voltage device, by which a discharged electric charge amount and a discharge part can be estimated even when a partial discharge is generated in the high-voltage device. SOLUTION: A vibration generated in the enclosure of the high-voltage device due to the partial discharge generated by the device is sensed by an AE sensor 15. A leakage current due to the partial discharge is sensed by a CT sensor 17. Sensed results from the AE sensor 15 and the CT sensor 17 and a power- supply signal from the power-supply line 19 of the device are measured by a measuring part 21, respectively. A signal processing operation in which information useful for an abnormality diagnosis is extracted from measured results is performed by a signal processing part 23. On the basis of a signal processed result, the electric charge amount of the partial discharge is calculated by an electric-charge-amount calculation part 27 by referring to an abnormality diagnostic database 25. On the basis of the sensed results by the AE sensor 15 and the CT sensor 17, the partial discharge is estimated by a discharge-part estimation part 31 by referring to a discharge-part estimation database 29.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device for diagnosing insulation abnormality of high-voltage equipment capable of realizing a quantitative evaluation of insulation abnormality. The present invention relates to a device for diagnosing insulation abnormality of high-voltage equipment capable of supporting a wide range of diagnosis target models.

[0002]

2. Description of the Related Art In recent years, the number of high-voltage power receiving and distributing facilities used over time has been increasing, and the need for an optimum maintenance plan (predictive maintenance) based on state monitoring is increasing. Therefore, there is an urgent need to develop an insulation abnormality diagnosis device that detects a partial discharge of a high-voltage device and monitors the state.

Conventionally, as a technique for detecting a partial discharge of a high-voltage device, a leak current detection technique for detecting a leak current at a ground wire using a CT sensor, and an electromagnetic wave generated at the time of discharge using an antenna are measured. Discharge electromagnetic wave measurement method,
There is a detection method for detecting partial discharge light using an optical fiber, and research and development are ongoing at present.

[0004] Aerial ultrasonic sound collectors that measure ultrasonic waves in the air generated by partial discharge of high-voltage equipment with a sound collector are also commercially available. In addition, a method of detecting a housing vibration due to a partial discharge with an AE sensor or a vibration sensor has been reported.

[0005]

The above-described partial discharge detection method cannot be completely compared because the detection principle and the rating and size of the target device are different. However, at present, in the field of insulation diagnosis, The object is to improve the sensitivity of discharge detection, noise separation (identification) for improving diagnosis accuracy, and practicality in performing field diagnosis.

However, the conventional method has the following problems.

In the case of the leakage current detection method, C
Without the ground wire to which the T sensor was attached, the leakage current due to partial discharge could not be measured. In addition, since the leakage current from the entire panel is measured, it is difficult to specify where the partial discharge is occurring in the equipment in the panel.

In the case of the electromagnetic wave measurement method, high-speed sampling is required to measure an electromagnetic wave, and thus there is a problem that a load on a measuring system is increased.

Further, in the method of detecting partial discharge light, there is a problem that even when an optical fiber is attached, there are positions where partial discharge light can be detected and positions where partial discharge light cannot be detected.

Furthermore, in the case of an aerial ultrasonic sound collector,
In a mode in which aerial ultrasonic waves are not generated due to the partial discharge, the partial discharge cannot be detected. In the detectable case, the collected data is subjected to FFT processing, and the level of insulation deterioration is determined based on the spectrum intensity. However, the amount of discharge charge actually generated from the device cannot be evaluated.

According to a method of diagnosing an insulation abnormality from a housing vibration caused by a partial discharge by attaching an AE sensor to a board, an AE sensor is mounted on a switchgear ceiling or a case of a circuit breaker, and the partial discharge is caused by ultrasonic waves. Although it is detected as mechanical vibration in the area, the discharge frequency band and intensity vary depending on the model, material, etc., so partial discharge in a frequency band specific to a wide range of models can be detected with only one AE sensor. Did not.

Although the detection signal from the AE sensor can be discriminated into field noise and partial discharge sound by using the wavelet transform, the amount of discharge charge cannot be estimated from the detection signal from the AE sensor. .

The present invention has been made in view of the above,
An object of the present invention is to provide an insulation abnormality diagnostic device for a high-voltage device that can estimate a discharge charge amount and a discharge portion even when a partial discharge occurs in the high-voltage device.

[0014]

According to the first aspect of the present invention,
In order to solve the above-mentioned problems, an AE sensor that detects vibration generated in a housing of a high-voltage device due to a partial discharge generated by the device, a CT sensor that detects a leakage current due to the partial discharge, the AE sensor, and a CT sensor Measuring means for measuring a detection result from the apparatus and a power signal from a power line of the device, a signal processing means for extracting useful information for abnormality diagnosis from the measurement result, and a signal processing result and a charge amount of the partial discharge. An abnormality diagnosis database that accumulates correlation information; a charge amount calculation unit that calculates the charge amount of the partial discharge by referring to the abnormality diagnosis database based on a signal processing result from the signal processing unit; and an AE sensor and a CT sensor. The discharge site estimation database stores the intensity of the detection result and the discharge site, and the discharge site estimation based on the detection results of the AE sensor and the CT sensor. And summarized in that and a discharge part estimation means by referring to the database to estimate the discharge site.

According to a second aspect of the present invention, there is provided a broadband vibration sensor for detecting a vibration generated in a housing of a high-voltage device due to a partial discharge generated by the device, and detecting the vibration from the wideband vibration sensor. Frequency analysis means for determining a frequency band characteristic of partial discharge from the result; and an AE sensor for partial discharge detection having high sensitivity characteristics for a frequency band characteristic of partial discharge obtained by the frequency analysis means. Sensor selection means for selecting, measurement means for measuring a detection signal from the AE sensor for partial discharge detection selected by the sensor selection means and a detection signal from the broadband vibration sensor, and information useful for diagnosis from the measurement result. A signal processing unit to be taken out, an abnormality diagnosis database that stores correlation information between the signal processing result and the charge amount of the partial discharge,
A charge amount calculation unit that calculates the charge amount of the partial discharge by referring to the abnormality diagnosis database based on a signal processing result from the signal processing unit; and a charge amount calculation unit that includes the broadband vibration sensor and the selected partial discharge detection AE sensor. A discharge portion estimation database for accumulating the intensity of the detection result and the discharge portion, and a discharge for estimating a discharge portion by referring to the discharge portion estimation database based on detection results of the broadband vibration sensor and the plurality of partial discharge detection AE sensors. The gist of the present invention is to include a region estimating unit.

[0016]

Embodiments of the present invention will be described below with reference to the drawings.

(First Embodiment) FIG. 1 shows an apparatus for diagnosing abnormal insulation of a high-voltage device 1 according to a first embodiment of the present invention.
FIG. 2 is a block diagram showing a schematic configuration of No. 1.

An insulation abnormality diagnosis device 11 for high voltage equipment includes an AE sensor 15 for detecting a partial discharge generated by the high voltage equipment as a vibration phenomenon of a metal container (housing), and a CT sensor for detecting a leakage current due to the partial discharge. 17 and AE sensor 15
Unit 13 for inputting the detection result of the CT sensor 17 and the applied voltage signal of the power supply LINE 19 and amplifying and amplifying each of them, a measuring unit 21 for measuring the output of the amplifier unit 13, Refer to the signal processing unit 23 for extracting information useful for diagnosis from the result, the insulation abnormality diagnosis database 25 storing the correlation function between the signal processing result from the signal processing unit 23 and the charge amount of the partial discharge, and the insulation abnormality diagnosis database 25. A charge amount calculation unit 27 for calculating the charge amount of the partial discharge; a discharge portion estimation database 29 for storing the intensity of the detection results of the AE sensor 15 and the CT sensor 17 and a discharge portion; an AE sensor 15 and a CT sensor 17 based on the detection result at 17
And a discharge portion estimating section 31 for estimating a discharge portion with reference to FIG.

Hereinafter, an insulation abnormality diagnosis apparatus 11 for high voltage equipment
Are described in detail, and the operation of the insulation abnormality diagnostic device 11 for high-voltage equipment will be described.

In FIG. 1, an AE signal from an AE sensor 15 and a CT signal from a CT sensor 17
A signal and a plurality of signals such as a voltage signal from the power supply line 19 are input. Each of the input signals is a filter circuit (not shown) provided inside the amplifier unit 13 and an amplifier circuit (not shown). Zu) via the desired sensor
The signal is converted into a signal that satisfies the amplifier characteristics and is output to the measurement unit 21.

In order to prevent AC power supply noise from being superimposed on the output signal from the AE sensor or the CT sensor, the amplifier 1
By inserting a noise cutter into the power supply line of No. 3, it is possible to prevent noise from being superimposed on the sensor output from the power supply line supplied locally. Further, in order to prevent AC power supply noise from being superimposed on the output signal from the AE sensor or the CT sensor, if the power supply of the amplifier unit 13 is entirely supplied by a battery, the noise with respect to the sensor output from the power supply line supplied locally is reduced. Can be prevented from overlapping.

Referring to FIG. 2, the measuring unit 21 includes a multi-channel input high-speed A / D conversion board 41 and a notebook computer 4.
3 to reduce the size of the notebook PC 43
By installing diagnostic software with a diagnostic function above, it becomes an effective tool for on-site diagnostics. When precision measurement is important, the measurement unit 21 may be configured to indirectly input measurement data to diagnostic software after measurement with a precision device such as an oscilloscope.

If a noise cutter is inserted into the power supply line of the measurement system in order to prevent superposition of AC power supply noise on the output signal from the AE sensor or CT sensor,
It is possible to prevent noise from being superimposed on sensor output from a power supply line supplied locally. In addition, in order to prevent AC power noise from being superimposed on the output signal from the AE sensor or the CT sensor, if the power of the measuring unit 21 is entirely supplied by a battery, the noise with respect to the sensor output from the power line supplied on site is reduced. Can be prevented from overlapping.

Here, with reference to FIG. 3, an outline of a diagnostic function by the measuring unit 21 will be described.

The AE sensor 15 is attached to the case of the high-voltage switchboard 45 or the housing of the circuit breaker 47 with a magnet or grease or the like. The housing vibration of the container 47 is detected. At this time, since the intensity of the output of the AE signal depends on the mounting position of the AE sensor 15, in order to realize a quantitative evaluation of the partial discharge, the AE sensor is previously determined by a preliminary experiment or the like for each model to be diagnosed. It is necessary to specify the position where 15 is to be attached.

Here, referring to FIG. 5, a magnetic circuit breaker (M
The mounting position of the AE sensor corresponding to BB) will be described.

With respect to the magnetic circuit breaker (MBB) shown in FIG. 5A, a measurement experiment is performed by attaching the AE sensor 15 to positions A, B and C.

As shown in FIG. 5B, the position A and the position B
Although the detection limits of the discharge starting voltage are almost equal, the detection sensitivities differ greatly. This is because the position A and the position B are one point on the same metal casing, but the amount of attenuation of the AE signal varies depending on the distance difference from the internal discharge occurrence position. However, since the position A is difficult to be actually mounted in the field due to safety in the field, the position B is better in consideration of actual operation.

The position C is the position where the AE sensor can be most easily mounted. However, since the position AE is structurally interposed between the connecting parts and the space, the intensity of the AE signal from the AE sensor is greatly attenuated at the joint. Problem.

As a result, in the magnetic circuit breaker (MBB) shown in FIG. 5A, it is preferable to mount the AE sensor 15 at the position B for measurement. In addition, since there is an optimal AE sensor mounting position for each diagnostic object, it is necessary to specify the AE sensor position in advance.

On the other hand, the CT sensor 17 is attached to a ground wire attached to a series of panel groups to detect a leakage current signal accompanying partial discharge. Further, as shown in FIG. 1, a local power supply line 19 is also detected in order to obtain phase shift information between the partial discharge generation phase and the applied power supply phase.

As shown in FIG. 3, the output of the plurality of sensors is converted into a desired signal by an insulating amplifier 13a.
After being converted into digital data by the / D card 41, the digital data is taken into the measuring unit 21. The insulating amplifier 13a has an electric circuit configuration that is electrically insulated so that even if a discharge occurs to the sensor during measurement, the discharge does not reach the measurement unit 21 from the sensor.

For example, a magnetic circuit breaker (MBB) shown in FIG.
In contrast, the center resonance frequency of the AE sensor 15 is as shown in FIG.
It is preferable to set the band to 20 to 40 kHz using a low-pass filter, a hand-pass filter, or the like, as in the characteristics of the sensor B shown in FIG. This is based on the frequency analysis result of the housing vibration obtained in the partial discharge test for various models and discharge modes.

As described above, since the AE sensor 15 or the CT sensor 17 has a high-sensitivity frequency response in a band of 20 kHz to 40 kHz, a certain type of magnetic circuit breaker (M
With regard to BB), highly sensitive partial discharge detection and accurate quantitative insulation diagnosis can be performed.

The measuring section 21 has a data collection condition setting function as shown in FIG.
Measurement waveforms (AE signal, CT signal, power signal), power frequency, device under test name, operator name, site name, manufacturer name,
Data collection conditions such as model, device name, serial number, and target facility name can be stored as a file in a storage unit (not shown) such as a hard disk.

As described above, the measuring unit 21 measures the date and time of measurement, temperature, humidity, measurement waveform (AE signal, CT signal, power signal), power frequency, name of device to be measured, name of measurer, site name, manufacturer name, Since a function for setting data collection conditions such as a model, a device name, a serial number, and a target facility name is provided, necessary information such as field information and environmental information at the time of on-site diagnosis can be input and stored.

The signals from the plurality of sensors are converted into A / D signals.
The data is simultaneously input to the card 41a, sampled at high speed at a sampling frequency that satisfies the sampling theorem, A / D converted, and stored in a file as data in, for example, a csv format.

As described above, the measuring section 21 has channels for measuring the power supply waveform and the partial discharge waveform applied to the high-voltage equipment, and performs sampling theorem for all the channels simultaneously and for the frequency band to be measured. Since the measurement is performed at a sampling frequency that satisfies, aliasing of the measured waveform is eliminated, and the time (phase) between each measurement signal is
Delay information and the like can be accurately grasped. From the accurate time information, it is possible to obtain information for estimating a discharge mode and a discharge position.

Further, the measuring unit 21 can perform measurement at a time designated by the measurer as a function of simultaneously collecting data on multiple channels. Also, for example, as shown in FIG.
There is also provided a trigger function for triggering to measure at a zero cross point from-to + with a pilot signal such as a power supply waveform. The measurement time is a setting function capable of measuring data for a desired period, for example, an integral multiple of the frequency (50 Hz or 60 Hz) of the commercial power supply.

As described above, the measuring section 21 has the trigger function of using the power supply waveform applied to the high-voltage equipment as a pilot signal and the function of measuring data of an integral multiple of the power supply cycle. It is possible to check in which phase the partial discharge occurs in the power supply.

Next, the data collected by the measuring section 21 is input to the signal processing section 23 shown in FIG. The signal processing unit 23 executes signal processing by diagnostic software according to the procedure shown in FIG.

In order to check the behavior of the AE signal for each period of the applied power supply, the signal processing unit 23 first determines in step S1
At 0, the AE data sampled at high speed by the high-speed A / D conversion board 41 is subjected to envelope processing (a low-pass filter at a desired frequency by digital filtering processing).
(High pass filter and band pass filter can be set.)
To detect the amplitude modulation component of the AE signal.

As described above, the signal processing unit 23 performs the envelope processing function (digital filter processing function) of the measured waveform data.
And the wavelet transform function, the amplitude modulation component can be extracted from the AE signal accompanying the partial discharge, and the time-frequency analysis of the AE amplitude modulation component can be performed, and the partial discharge included in the measured AE waveform can be performed. The feature component of the component can be detected as a peak on the time-frequency plane.

In step S20, the DC component of the data is removed by subtracting the average of the absolute values of the measured data from the result of the envelope processing.

As described above, since the signal processing unit 23 is provided with the function of removing the DC component of the waveform data measured and sampled, the AE unnecessary for the insulation abnormality diagnosis algorithm is provided.
The DC component of the signal can be cut, and the display of the wavelet transform result can be easily viewed.

Next, in step S30, data corresponding to one cycle of the sine wave of the power supply is cut out in order to check the relationship between the applied voltage and the generation phase of the partial discharge. Next, step S
In step 40, a wavelet transform is performed using a conversion formula for diagnosis, and the conversion result (three-dimensional data) is stored in a notebook computer 43.
Are displayed on the time-frequency plane on the display screen.

This display method is shown in FIGS.
There is a contour display method as shown in FIG. Normal,
The wavelet transform is defined as in equation (1), and takes an inner product of a localization function と called a mother wavelet and a waveform f (t) to be analyzed, and converts f (t) into a wavelet coefficient W ( a, b).

[0048]

(Equation 1) Here, as shown in the equation (2), the parameters of the mother wavelet ， are replaced with a = 2 ＾ j and b = k, and specialization is made for diagnosis so that the time resolution does not become coarse even on the low frequency side. I have.

[0049]

(Equation 2) In the mother wavelet transform, a plurality of mother wavelets are selected in advance because an optimal function differs depending on a diagnosis target. A typical example of a frequently used mother wavelet is a Mexican hat represented by the following equation (3).

[0050]

(Equation 3) Actually, as shown in FIG.
It is preferable to add to the diagnostic software a function of cutting out only the period and a function of selecting a mother wavelet (also called an analyzing wavelet).

As described above, since the signal processing unit 23 is provided with a plurality of mother wavelets for the wavelet transform, the signal processing unit 23 detects a partial discharge characteristic component using a mother wavelet suitable for each device to be diagnosed and each sensor output. An easy wavelet transform can be performed.

Next, with reference to FIG. 4A, an algorithm for making a diagnosis from the result of the wavelet transform will be described.

First, in steps S110 and S120,
Partial discharge noise is identified from the peak frequency component obtained as a result of the above-described wavelet transform. Partial discharge tends to occur near the positive and negative peaks of the applied voltage waveform, and the AE signal is expected to be amplitude-modulated in a half cycle of the applied voltage.

The noise has a characteristic frequency component for each type. If the discharge and the noise are not synchronized at all, the characteristic frequency component of the partial discharge that can distinguish the discharge and the noise from the frequency component of the wavelet transform result is: Since the partial discharge easily occurs at both the positive and negative peaks of the applied voltage, the applied frequency is set to twice the applied frequency.

Next, in steps S130 and S140,
For the partial discharge peak identified as noise, the partial discharge mode is specified by the above partial discharge and the discharge phase. In other words, each partial discharge mode such as creeping discharge, gap discharge, and void discharge has a specific signal strength and discharge phase.Therefore, focusing on the center of gravity phase of the discharge peak in the contour display of the wavelet transform result, the discharge mode is set. judge.

Next, in steps S150 and S160,
Attention is paid to the discharge peak intensity (discharge peak volume in contour display). This corresponds to the charge amount (discharge energy) of the generated partial discharge as shown in FIG. Therefore, the partial discharge charge amount can be quantitatively estimated from the calculation of the discharge peak intensity. An insulation abnormality is determined based on whether the estimated magnitude of the partial discharge deviates from a maker standard, a user standard, or some standard.

Therefore, in the signal processing section 23, as shown in FIG. 10, the power supply positive phase (0 ° -1
80 °) and the power supply negative phase (180 ° to 360 °), the effective value of the output data of the AE signal is calculated, and the peak intensity of the partial discharge component obtained as a result of the wavelet transform is also integrated by volume (within the discharge peak). (The sum of the included wavelet coefficients).

As described above, since the signal processing section 23 has a function of cutting out one cycle of the power supply waveform from the measurement waveform, data analysis can be performed for each power supply phase from 0 ° to 360 °, and the wavelet transform result can be obtained. Time series analysis becomes easier.

Next, the various numerical values calculated by the signal processing unit 23 are input to the charge amount calculating unit 27 shown in FIG. Originally, between the applied voltage and the output of the partial discharge charge and the AE signal, FIG.
There is a relationship as shown in FIG. Therefore, at the laboratory level, the applied power supply voltage, the amount of discharge charge (the amount of partial discharge measured by a charge amount measuring device), and the AE at the time of partial discharge are adjusted while adjusting the applied voltage to change the amount of partial discharge charge. Regarding data obtained by simultaneously measuring three items of the signal, a data set of the total charge amount of the group small discharge and the effective output value of the AE signal in the circled portion shown in FIG. A set of data of the total charge amount of the discharge and the data of the discharge peak intensity corresponding to the result of the wavelet transform may be stored in the database.

By performing this preliminary experiment, as the insulation abnormality diagnosis database 25, “correlation curve between AE signal intensity and partial discharge charge amount” or “AE
A master curve of "correlation curve between signal wavelet transform peak intensity and partial discharge charge amount" can be provided.

Therefore, if the electric charge amount calculating section 27 has a function of selecting a different electric charge amount estimation master curve from the diagnosis database 25 for each model to be diagnosed and a function of changing the evaluation threshold value, An optimal charge amount estimation master curve and an evaluation threshold value for diagnosis can be set for each model to be diagnosed, and a quantitative insulation diagnosis can be realized.

FIG. 12 shows a magnetic circuit breaker (MB) shown in FIG.
17 is a diagnostic database for B), showing a correlation curve when an AE sensor having the frequency characteristics shown in FIG. 17 is used. Accurate charge amount estimation with a correlation coefficient R close to 0.9 can be performed.

The charge amount calculation unit 27 refers to the correlation curve coefficient stored in the insulation abnormality diagnosis database 25, based on the output effective value of the AE signal and the wavelet intensity of the partial discharge calculated by the signal processing unit 23. It has a function of calculating the amount of partial discharge charge.

In FIG. 12, the total charge amount is 1500
The behavior of the correlation between the output of the AE signal and the charge amount is different from ２０００2000 pC as a boundary. FIGS. 12A and 12B
When the low charge amount region shown in FIG.
A correlation curve as shown in (d) can be obtained.

Here, in particular, when the total charge is 0 to 1500 pC
In the case of (low total charge amount region), it can be seen that linear approximation (proportional relationship) is established and the correlation (R ＾ 2 value) is further improved. In the region where the total electric charge is 200 pC or more, the output of the AE signal tends to be saturated. For the diagnosis, if the correlation function in the low total electric charge region where the total electric charge of the partial discharge is 0 to 1500 pC is used, the electric charge with high accuracy can be obtained. An estimate can be made.

In actual operation, for example, as shown in FIG.
If the correlation curve shown in (1) is prepared in advance as a database, in the field of low noise environment, partial discharge can be quantitatively evaluated only by measuring the output of the AE signal. what if,
Assuming that the effective value of the output of the AE signal from the sensor B (for a half cycle of the power supply) is measured as 300 mVrms, the partial discharge charge amount can be estimated to be 570 (pC) from the regression approximation formula.

As described above, the database stores the total value of the partial discharge charge amounts measured by the charge amount measuring device (not shown) and A
Since a correlation curve with the wavelet transform peak intensity of the E sensor output is provided, even if noise is mixed in the output of the AE signal of the housing (panel) vibration collected by the field measurement, the noise is identified by the wavelet transform. By calculating the characteristic peak intensity of the discharge, the partial discharge charge amount can be estimated from the output of the AE signal.

Further, the charge amount calculating section 27 calculates a low total charge amount region (eg, 0 to 1500) in the diagnostic database 25.
Since the linearity of the correlation curve at pC) is used, accurate charge amount estimation can be realized with a correlation having a correlation coefficient of 0.9 or more.

However, in actual field measurement, it is assumed that various noises are included. Therefore, in consideration of practicality, quantification of a partial discharge component obtained by removing a noise component from a measured AE signal using a wavelet transform is considered. In many cases, we use spatial correlation. Therefore, the partial discharge charge amount is estimated from the wavelet transform of the AE signal using the correlation between the wavelet peak intensity of the partial discharge characteristic component and the discharge charge amount as shown in FIG.

As described above, the charge amount calculation unit 27 determines the wavelet peaks in the frequency of 100 Hz appearing in the positive phase and the negative phase of the power supply voltage as a result of the wavelet transformation of the AE signal as the partial discharge components. In order to calculate the wavelet peak intensity by volume integration, it is possible to accurately evaluate the intensity of the wavelet transform peak of the partial discharge component that has identified the noise by volume integration, and to generate small groups at the positive and negative peaks of the applied voltage. For partial discharges with a strong tendency, the wavelet peak intensity can be evaluated for each amplitude modulation component of the output of the AE signal caused by each group.

Further, the charge amount calculator 27 has a function of calculating the effective value of the output of the AE signal for each of the positive-phase half cycle and the negative-phase half cycle of the power supply voltage. Can be evaluated for each amplitude modulation component of the output of the AE signal caused by a group of partial discharges, with regard to the partial discharges that tend to generate small groups at the peak of It becomes possible to analyze.

Further, steps S130 and S14 described above are performed.
0, the partial discharge mode was specified.
In the seventh example, if a function of calculating the phase center of the partial discharge component peak in the wavelet transform result is provided, information on the time at which the partial discharge occurs can be captured, and the partial discharge mode (surface discharge, gap discharge, void discharge, et.
c. ) Can be estimated, and the accuracy of specifying the partial discharge mode can be further improved.

Next, the discharge portion estimating section 31 and the discharge portion estimating database 29 will be described with reference to FIG.

First, FIG. 13A shows that the output of the AE signal is the smallest although the output of the CT signal is the largest among the three cases. Negative phase of the power supply cycle (180 to 36
0 °), it is assumed that the electron emission from the tip of the metal rod promotes the vibration of the metal rod. However, since the metal rod is not firmly fixed to the model electrode, the tape is fixed. , The vibration is absorbed and the vibration is not easily transmitted to the AE sensor.

On the other hand, FIG. 13B shows C in the negative phase.
Although the output of the T signal is less than half in FIG. 13A, the AE sensor 15 reliably detects the partial discharge. This is because the bolt electrode at the discharge site is firmly fixed, and the vibration generated at the discharge site is easily propagated. In the figure, in addition to the discharge in the positive phase of the power supply, the output of a high-level CT signal is also detected on the negative phase side. Is raised. Also in this case, since the discharge energy is large, the housing vibration clearly appears in the output of the AE signal.

FIG. 13C shows a structure in which a volt electrode protrudes from the ground side. Discharge occurs only when the applied voltage on the high voltage side has a positive phase, that is, when the electron concentration occurs at the tip of the volt. . This means that, due to the electron emission, the vibration propagates well to the housing due to the vibration of the electrode bolted to the ground side. Almost no discharge is observed in the negative phase, because the inside of the conductor of the plane electrode has a uniform negative potential and the electrons are not localized, so that the electrons cannot be emitted.

As described above, regarding the difference between the electrode shape and the fixed state of the electrode portion, the combination pattern of the output of the CT signal and the output of the AE signal as shown in FIGS. 13 (a), (b) and (c) is determined in advance. If the discharge part estimation database 29 is provided, the shape of the discharge part and the position of the discharge part can be estimated from the difference in the combination of the output intensity of the CT signal and the output of the AE signal and the phase shift.

The effect of the first embodiment of the present invention is that the AE sensor detects the vibration generated in the housing of the high-voltage equipment due to the partial discharge generated by the equipment, and the leakage current due to the partial discharge is detected by the CT sensor. AE sensor and CT
The signal processing for amplifying the detection result from the sensor and the power signal from the power supply line of the device via the amplifier unit and further performing filter processing, and measuring the result, and extracting information useful for abnormality diagnosis from the measurement result is performed. Then, the charge amount of the partial discharge is calculated by referring to the abnormality diagnosis database based on the signal processing result, and the discharge part is estimated by referring to the discharge part estimation database based on the detection results of the AE sensor and the CT sensor. Even when a partial discharge occurs in the voltage device, it is possible to estimate a discharge charge amount and a discharge portion.

(Second Embodiment) FIG. 14 is a block diagram showing a schematic configuration of an insulation abnormality diagnosing device 51 of a high-voltage device according to a second embodiment of the present invention. FIGS. 15 and 16 are diagrams illustrating examples in which the output of the broadband vibration sensor according to the second embodiment of the present invention is analyzed by the frequency analysis unit. FIG. 17 is a diagram illustrating an example of a characteristic frequency band of the AE sensor.

In FIG. 14, the insulation abnormality diagnosis device 51
Has a broadband vibration sensor 53. The ultrasonic frequency range, which is a feature of the housing (panel) vibration associated with the partial discharge, differs depending on the model to be diagnosed. Therefore, first, a broadband vibration sensor is attached to the housing of a power distribution board or a circuit breaker to detect board vibration and housing vibration in a wide frequency band. Next, the output of the broadband vibration sensor 53 is input to the frequency analysis unit 55.

The frequency analysis unit 55 has an FFT function and a wavelet transform function, and performs a frequency analysis of the input broadband vibration. For example, FIG.
FIG. 17 is a diagram of the six frequency characteristics shown in FIG. 16, and the frequency spectrum obtained according to the model and the partial discharge mode is slightly different. Therefore, based on the frequency spectrum obtained by the frequency analysis unit 55, the characteristic frequency components of the target partial discharge board vibration and the housing vibration are specified, and the analysis result is output to the sensor selection unit 57.

As described above, the frequency analysis unit 55
(Fast Fourier Transform) function or wavelet transform function enables to calculate the intensity distribution of the frequency component of the output of the broadband vibration sensor due to partial discharge generated from the device to be diagnosed, and to optimize the frequency characteristics for the object to be diagnosed. AE sensor can be selected.

The sensor selecting section 57 selects a sensor having a sensitivity characteristic in a frequency band in which partial discharge vibration can be detected with high sensitivity based on the analysis result from the sensor group 59. Then, the selected sensor is attached to the object to be diagnosed, and the ultrasonic vibration due to the partial discharge is detected and input to the amplifier unit 13.

As described above, the sensor selecting section 57 is provided with a plurality of high-sensitivity AE sensor groups having different frequency characteristics in advance, and has an AE sensor having high sensitivity in the characteristic frequency band of the partial discharge obtained from the frequency analysis result. Is selected, a sensor optimal for detecting partial discharge of the device to be diagnosed (having the most suitable frequency characteristic) can be used for diagnosis, and partial discharge can be detected with high sensitivity.

Further, if the mounting position of the sensor group is determined in advance for each target model, at the time of each diagnosis,
Measurement errors due to sensor displacement can be prevented, and quantitative evaluation using a diagnostic database becomes possible.

Hereinafter, the measuring unit 21, the signal processing unit 23, the charge amount calculating unit 27, and the discharge part estimating unit 31 after the amplifier unit 13 will be described.
Has the same configuration as that of the insulation abnormality diagnostic device 11 described in the first embodiment of the present invention, and a description thereof will be omitted.

The effect of the second embodiment of the present invention is that the vibration generated in the housing of the high-voltage device due to the partial discharge generated by the device is detected by the broadband vibration sensor, and the detection result from the broadband vibration sensor is used. A frequency band characteristic of partial discharge is determined, and an AE sensor for partial discharge detection having high sensitivity characteristics is selected for the obtained frequency band characteristic of partial discharge, and the selected AE sensor for partial discharge detection is selected.
It measures the detection signal from the sensor and the detection signal from the broadband vibration sensor, performs signal processing for extracting information useful for diagnosis from the measurement result, and refers to the abnormality diagnosis database based on the signal processing result to determine the charge amount of the partial discharge. By calculating the discharge location and estimating the discharge location by referring to the discharge location estimation database based on the detection results of the wideband vibration sensor and the plurality of AE sensors for partial discharge detection, even when a high-voltage device has a partial discharge, The amount of charge and the discharge site can be estimated.

In particular, it is possible to detect a partial discharge with high sensitivity using an optimum sensor for each diagnosis target and each discharge mode, and even when a partial discharge occurs in a high-voltage device, the amount of the discharged electric charge is increased. And the accuracy of estimation of the discharge site can be improved.

[0089]

According to the first aspect of the present invention, the vibration generated in the housing of the high-voltage device due to the partial discharge generated by the device is detected by the AE sensor, and the leakage current due to the partial discharge is detected by the CT sensor. Then, the detection result from the AE sensor and the CT sensor and the power supply signal from the power supply line of the device are measured, and signal processing for extracting information useful for abnormality diagnosis from the measurement result is performed. , The charge amount of the partial discharge is calculated, and AE is calculated.
By estimating the discharge site by referring to the discharge site estimation database based on the detection results of the sensor and the CT sensor, it is possible to estimate the discharge charge amount and the discharge site even when a partial discharge occurs in the high-voltage device.

According to the second aspect of the present invention, the vibration generated in the housing of the high-voltage equipment due to the partial discharge generated by the equipment is detected by the wide-band vibration sensor, and the vibration is detected based on the detection result from the wide-band vibration sensor. A frequency band characteristic of discharge is determined, and an AE sensor for partial discharge detection having high sensitivity characteristics is selected for the obtained frequency band characteristic of partial discharge, and the selected AE sensor for partial discharge detection is selected. And the detection signal from the broadband vibration sensor, and performs signal processing for extracting information useful for diagnosis from the measurement result, and refers to the abnormality diagnosis database based on the signal processing result to determine the charge amount of the partial discharge. By calculating and estimating the discharge site by referring to the discharge site estimation database based on the detection results of the broadband vibration sensor and the plurality of AE sensors for partial discharge detection, Even if the partial discharge in high voltage equipment has occurred, it is possible to estimate the discharge charge quantity discharge site.

In particular, it is possible to detect a partial discharge with high sensitivity by using an optimum sensor for each diagnosis target and each discharge mode, and even when a partial discharge occurs in a high-voltage device, a discharge charge amount can be increased. And the accuracy of estimation of the discharge site can be improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of an insulation abnormality diagnosis device for a high-voltage device according to a first embodiment of the present invention.

FIG. 2 is a diagram illustrating a configuration of a measurement unit of the insulation abnormality diagnosis device.

FIGS. 3A, 3B, and 3C are diagrams for explaining an outline of a diagnosis using the insulation abnormality diagnosis device. FIGS.

4A, 4B, 4C, and 4D are diagrams for explaining an insulation diagnosis algorithm performed by the insulation abnormality diagnosis device.

FIGS. 5A and 5B are diagrams showing the dependency of the mounting position of the AE sensor when the insulation abnormality diagnosis device is used. FIGS.

FIG. 6 is a diagram illustrating an example of a data collection condition setting function in a measurement unit of the insulation abnormality diagnosis device.

FIG. 7 is a diagram for explaining a trigger function and a function of measuring data of an integral multiple of a power supply cycle in a measurement unit of the insulation abnormality diagnosis device.

FIG. 8 is a diagram for explaining an outline of a signal processing function of the insulation abnormality diagnosis device.

FIG. 9 is a diagram for explaining a data cutout function and a plurality of mother wavelet reserve functions in a signal processing unit of the insulation abnormality diagnosis device.

FIG. 10 is a diagram illustrating an example of an evaluation result of insulation abnormality diagnosis using the insulation abnormality diagnosis device.

FIG. 11 is a diagram for explaining a correspondence relationship between a group small partial discharge pulse actually generated and an AE signal;

FIGS. 12A and 12B show examples of correlation curves stored in an insulation abnormality diagnosis database of the insulation abnormality diagnosis apparatus;
(B), (c), and (d).

FIGS. 13 (a), (b), and (c) show examples of a discharge part estimation database of a discharge part estimation unit of the insulation abnormality diagnosis device.
(C).

FIG. 14 is a block diagram illustrating a schematic configuration of an insulation abnormality diagnosis device for a high-voltage device according to a second embodiment of the present invention.

FIGS. 15 (a), (b), and (c) show examples in which the output of a broadband vibration sensor according to the second embodiment of the present invention is analyzed by a frequency analysis unit.

FIGS. 16 (a), (b), and (c) are views showing examples in which the output of the broadband vibration sensor according to the second embodiment of the present invention is analyzed by the frequency analysis unit.

FIG. 17 is a diagram illustrating an example of a characteristic frequency band of the AE sensor.

[Explanation of symbols]

 11, 51 Insulation abnormality diagnosis device 15 AE sensor 17 CT sensor 19 Power supply LINE 13 Amplifier unit 21 Measurement unit 23 Signal processing unit 25 Insulation abnormality diagnosis database 27 Charge amount calculation unit 29 Discharge site estimation database 31 Discharge site estimation unit 53 Broadband vibration sensor 55 frequency analysis unit 57 sensor selection unit 59 sensor group

Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat II (reference) // G01N 29/14 H02B 13/06 D (72) Inventor Takeshi Watanabe 1 Toshiba-cho, Fuchu-shi, Tokyo Toshiba Fuchu Co., Ltd. In-house (72) Inventor Hideki Nemoto 1 Toshiba-cho, Fuchu-shi, Tokyo Toshiba Corporation Fuchu Office (72) Inventor Katsuyuki Hiroi 1-1-1 Shibaura, Minato-ku, Tokyo Toshiba head office F-term ( Reference) 2G015 AA10 AA15 AA16 BA04 CA01 CA03 2G047 AC10 BA05 BC04 GA20 GG12 GG19 GG24 GG36 5G017 EE02 EE03 5G028 GG21 5G365 DN04

Claims (23)

[Claims] An AE sensor for detecting a vibration generated in a housing of the device due to a partial discharge generated by the high-voltage device; a CT sensor for detecting a leakage current due to the partial discharge; Measuring means for measuring the detection result and the power signal from the power line of the device, signal processing means for extracting information useful for abnormality diagnosis from the measurement result, and correlation information between the signal processing result and the charge amount of the partial discharge Diagnosis database that accumulates the data, charge calculation means for calculating the charge amount of the partial discharge by referring to the abnormality diagnosis database based on the signal processing result from the signal processing means, and detection results by the AE sensor and the CT sensor A discharge site estimation database that stores the intensity of the discharge and the discharge site, and the discharge site estimation data based on the detection results of the AE sensor and the CT sensor. Insulation abnormality diagnosis system of the high voltage equipment, characterized in that a discharge part estimation means with reference to the base for estimating the discharge site. 2. A wide-band vibration sensor for detecting vibration generated in a housing of a high-voltage device due to a partial discharge generated by the high-voltage device, and determining a frequency band characteristic of the partial discharge from a detection result from the wide-band vibration sensor. Frequency analysis means, sensor selection means for selecting an AE sensor for detecting partial discharge having high sensitivity characteristics for a frequency band characteristic of partial discharge obtained by the frequency analysis means, and selection by the sensor selection means A for partial discharge detection
A measuring means for measuring a detection signal from the E sensor and a detection signal from the broadband vibration sensor; a signal processing means for extracting information useful for diagnosis from the measurement result; and a correlation information between the signal processing result and the charge amount of the partial discharge. An abnormality diagnosis database to be accumulated; a charge amount calculation unit for calculating the charge amount of the partial discharge by referring to the abnormality diagnosis database based on a signal processing result from the signal processing unit; and the broadband vibration sensor and the selected portion A discharge portion estimation database that stores the intensity of the detection result and the discharge portion in the discharge detection AE sensor, the broadband vibration sensor and the plurality of partial discharge detection AEs.
An insulation abnormality diagnosis apparatus for a high-voltage device, comprising: a discharge part estimating means for estimating a discharge part by referring to the discharge part estimation database based on a result detected by a sensor. 3. The measuring means has a channel for measuring a power supply waveform and a partial discharge waveform applied to the high-voltage device, and performs a sampling theorem for all channels simultaneously and for a frequency band to be measured. 3. The apparatus according to claim 1, wherein the measurement is performed at a sampling frequency that satisfies the condition. 4. The measurement means has a trigger function of using a waveform of a power supply applied to the high-voltage device as a pilot signal, and a function of measuring data at a cycle that is an integral multiple of a cycle of the power supply. The apparatus for diagnosing abnormal insulation of a high-voltage device according to claim 1 or 2, wherein: 5. The measuring means includes a measurement date and time, a temperature, a humidity, a measurement waveform, a power supply frequency, a device to be measured, a person to be measured, a site name, a maker name, a model, a device name, a serial number, a target facility name, and the like. 3. The apparatus for diagnosing abnormal insulation of a high-voltage device according to claim 1, further comprising a function of setting data collection conditions. 6. The apparatus according to claim 2, wherein the frequency analysis unit has a fast Fourier transform function or a wavelet transform processing function. 7. The apparatus according to claim 1, wherein the signal processing unit has an envelope processing function and a wavelet transform processing function. 8. The apparatus according to claim 7, wherein the signal processing unit has a function of removing a DC component from the measured waveform data. 9. The apparatus according to claim 8, wherein the signal processing unit has a function of cutting out one cycle of the power supply waveform from the measurement waveform. 10. The apparatus according to claim 6, wherein the signal processing unit includes a plurality of mother wavelets used for wavelet transform. 11. The abnormality diagnosis database includes: a total value of partial discharge charge amounts measured by a charge amount measuring device;
3. The apparatus according to claim 1, further comprising a correlation curve of an effective value detected by the E sensor. 12. The abnormality diagnosis database according to claim 1, wherein: a total value of the charge amounts of the partial discharges measured by the charge amount measuring device;
12. The apparatus for diagnosing abnormal insulation of a high-voltage device according to claim 11, further comprising a correlation curve between a detection result of the AE sensor and a peak intensity obtained by performing a wavelet transform. 13. The sensor selection means includes a plurality of AE sensor groups having different frequency characteristics, and selects an AE sensor having high sensitivity to a frequency band characteristic of partial discharge obtained from the frequency analysis result. 3. The apparatus for diagnosing abnormal insulation of a high-voltage device according to claim 2, wherein: 14. The apparatus according to claim 2, wherein the AE sensor or the broadband vibration sensor has a high sensitivity band in a band of 20 kHz to 40 kHz. 15. The high-voltage device according to claim 1, wherein the charge amount calculating unit uses linearity of a correlation curve in a low total charge amount region accumulated in the abnormality diagnosis database. Insulation abnormality diagnostic equipment. 16. The charge amount calculating means has a function of calculating an effective value of an AE signal from an AE sensor for a half cycle on a positive phase side and a half cycle on a negative phase of a power supply voltage, respectively. The apparatus for diagnosing insulation abnormality of a high-voltage device according to claim 1 or 2. 17. The method according to claim 17, wherein the charge amount calculating unit calculates a wavelet transform of an AE signal from the AE sensor and calculates a partial discharge component of a wavelet peak in a frequency of 100 Hz appearing in a positive phase and a negative phase of a power supply voltage. The insulation abnormality diagnosis apparatus for a high-voltage device according to claim 6, wherein the wavelet peak intensity is calculated by performing volume integration for each peak. 18. The high-voltage equipment according to claim 17, wherein said charge amount calculating means has a function of calculating a phase center of a peak of a partial discharge component with respect to a result of said wavelet transform. Insulation abnormality diagnosis device. 19. The charge amount calculation means has a function of selecting a different charge amount estimation master curve from the abnormality diagnosis database for each diagnosis target model, and an evaluation threshold value changing function of changing an evaluation threshold value. 3. The apparatus for diagnosing abnormal insulation of a high-voltage device according to claim 1, further comprising: 20. The mounting position of the AE sensor group is determined in advance for each target model.
Insulation abnormality diagnosis device for high voltage equipment as described in the above. 21. The high-voltage device according to claim 1, wherein the discharge site estimating means calculates a combination of the detection result of the CT sensor and the intensity and phase shift of the detection result of the AE sensor. Insulation abnormality diagnostic equipment. 22. The apparatus for diagnosing abnormal insulation of a high-voltage device according to claim 1, further comprising a battery for supplying power to the measuring system. 23. The apparatus according to claim 1, further comprising a noise cutter for removing noise from a power supply line of the measurement system.