KR20160093452A - Device for detecting insulation defect of mold transformer and method thereby - Google Patents

Abstract

A method for determining the insulation defect of a mold transformer according to the present invention comprises: a measurement step of obtaining the degradation signal of a mold transformer by detecting partial discharge generated in the mold transformer; a signal processing step of removing noise existing in the degradation signal; and an analysis step of determining the insulation defect of the mold transformer from the waveform of the degradation signal having the noise removed therefrom. According to the present invention, a mold transformer accident can be prevented in advance as whether the insulation of the mold transformer is defective or not can be determined from a partial discharge signal.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a device for determining an insulation defect of a mold transformer,

The present invention relates to an apparatus and method for determining an insulation defect in a mold transformer, and more particularly, to an apparatus and method for determining an insulation defect using a partial discharge signal generated in a mold transformer in a live line state.

In domestic electric facilities, power transformers are very important in heavy electric equipment. Especially, since domestic mold transformer was commercialized in Ewha Electric Industry in 1981, it has been rapidly spreading to cope with inflow transformer until recently. The mold transformer is advantageous in that it is less likely to ignite due to the nature of the installation environment, has a high mechanical strength, and is molded with a mold in a vacuum state and hardly causes deterioration in insulation due to moisture absorption. It is estimated that such a mold transformer is used in more than 200,000 locations of domestic power receiving facilities under 22.9kV in Korea.

The replacement period of the mold transformer was announced by the Japan Electrical Manufacturers' Association (JEMA) with an average life span of 15 to 20 years, and the Korea Electric Safety Corporation (KESCO) offered 20 years. Therefore, it can be inferred that the replacement cycle of the mold transformer installed in the field has come almost since the mid 1980s. Therefore, the importance of inspection cycle and safety management of the mold transformer is emphasized, and it is necessary to make a preventive diagnosis to detect an abnormal signal of an accident sign in the mold transformer at an early stage.

The most common types of mold transformer accidents are insulation breakdown due to voids or foreign matter in epoxy resin, which is a solid insulation, and surface discharge accident caused by epoxy crack. In the early days, mold transformers manufactured in Korea were frequently caused by many manufacturing defects due to manufacturing technology problems. However, due to the development of technology, accidents caused by maintenance factors and environmental factors have occurred frequently .

Methods of diagnosing the state of the mold transformer include insulation resistance measurement, electric strength test method, ultrasonic measurement method, partial discharge measurement method, and temperature (heat) measurement method. The mold transformer has a high initial insulation performance because the main part of the device is made of solid insulation, but there is a problem that the visual inspection or state analysis of the internal state of the insulation is difficult due to the structure of the device. Unlike an inflow transformer, a molded transformer is sealed with a solid insulator, epoxy resin. Therefore, there are many limitations to the method of checking the deterioration state and predicting an accident.

Conventionally, for the maintenance of the mold transformer, only very limited items such as visual inspection and thermal management using a thermal camera and noise management were checked. In addition, temperature management using thermal cameras, which are mostly implemented in the field, can not detect the deterioration of the insulation, and is disadvantageous in that it is distorted by the influence of the load or the environment.

In the precise diagnosis, the partial discharge of the mold transformer was tested to determine whether it was deteriorated. However, there is a problem that it is difficult to predict the mold transformer accident beforehand with these inspection items. As an inspection item for accident prevention of mold transformer, insulation deterioration can be measured through dielectric tangent test in offline state. However, this method is only possible at the annual inspection due to problems such as power failure, and has a disadvantage in that a large amount of preparation time is required. Therefore, it is effective to measure the partial discharge of the mold transformer in operation, but it requires a technique to remove ambient noise during measurement and a criterion to judge insulation defect from partial discharge. In this way, accident prevention technology of mold transformer is very limited compared to inflow transformer.

Korea Electrical Safety Corporation, which conducts the inspections of electrical facilities, introduced the uninterruptible power system of electric power facilities from 2011 and carried out the safety inspection of the mold transformer by the partial discharge test in live condition. However, there are many problems in the test judgment because the test judgment criterion does not match the condition of the field where the external noise signal can be inputted in a large quantity.

Research trends in related fields at home and abroad are at the level where the deterioration state of the mold transformer is measured and the analysis technique is presented in the insulation defect condition of the laboratory, and it is difficult to utilize it in the field. In order to diagnose insulation deterioration of mold transformer, complementary measurement method and insulation state of mold transformer are required. In this paper, And an analytical method capable of determining the presence or absence of an abnormality.

Korean Patent No. 1358049

SUMMARY OF THE INVENTION The present invention is directed to an apparatus and method for determining whether an insulation defect of a mold transformer is present or absent. It is another object of the present invention to provide an apparatus and a method for detecting a partial discharge generated in a mold transformer and eliminating a noise of a detected partial discharge signal to determine whether or not an insulation defect exists.

According to an aspect of the present invention, there is provided a method of determining an insulation defect of a mold transformer, comprising: a measurement step of detecting a partial discharge generated in a mold transformer to acquire a deterioration signal of the mold transformer; A signal processing step of removing noise existing in the deteriorated signal; And an analysis step of judging an insulation defect of the mold transformer from the waveform of the deteriorated signal from which the noise is removed.

Preferably, the measuring step may detect the partial discharge in a non-contact manner from the mold transformer using a helical antenna.

Preferably, the signal processing step may remove noise present in the degraded signal using a normal wavelet transform technique.

Preferably, the signal processing step comprises: (a) setting a level of decomposition to filter noise based on frequency; (B) setting a threshold value to filter noise based on the size; And (c) inverse wavelet transforming the filtered deteriorated signal.

Preferably, the analysis step can determine the insulation defect of the mold transformer using the peak value and the pulse value of the deteriorated signal. In this case, the analysis step determines an insulation defect of the mold transformer through a rate of change of the partial discharge coefficient, and the partial discharge coefficient can be derived by the following equation.

[Mathematical Expression]

는 부분방전 계수이고,

은 상기 열화신호의 함수이고,

는 상기 열화신호의 1 주기에서 양수 범위의 펄스 값이고,

은 상기 열화신호의 1 주기에서 음수 범위의 펄스 값이고,

은 상기 측정단계에서 열화신호를 취득한 시간의 길이이고,

는 상기 열화신호의 주파수.here

Is a partial discharge coefficient,

Is a function of the deteriorated signal,

Is a positive pulse value in one cycle of the deteriorated signal,

Is a pulse value in a negative range in one cycle of the deteriorated signal,

Is a length of time obtained by acquiring a deteriorated signal in the measuring step,

Is the frequency of the deteriorated signal.

Preferably, the analysis step can determine the mold transformer as an insulation defect if the rate of change of the partial discharge coefficient is greater than zero.

According to another aspect of the present invention, there is provided an apparatus for determining an insulation defect in a mold transformer, comprising: a helical antenna for detecting a partial discharge generated in a mold transformer and acquiring a deterioration signal of the mold transformer; A signal processing module for removing noise existing in a deteriorated signal by using a normal wavelet transform (Stationary Wavelet Transform) technique; And an analysis module for determining an insulation defect of the mold transformer based on the waveform of the deteriorated signal from which the noise is removed.

Preferably, the analysis module determines the insulation defect of the mold transformer through the rate of change of the partial discharge coefficient, and the partial discharge coefficient can be derived from the above equation.

According to the present invention, it is possible to determine whether the mold transformer is insulated from a partial discharge signal, thereby advantageously preventing an accident of the mold transformer.

Further, according to the present invention, the partial discharge of the mold transformer is detected in a non-contact manner by using a helical antenna, so that it is easy to judge the generation site of the discharge, and no noise is introduced from other facilities.

The present invention is advantageous in that noise included in a deterioration signal of a mold transformer can be effectively removed by using a stationary wavelet transform technique, which is mainly applied to a bio-signal analysis technique.

Further, the present invention can provide a criterion for determining the insulation defect of the mold transformer by analyzing the waveform of the signal-processed deteriorated signal.

FIG. 1 is a conceptual diagram of an insulation defect determination apparatus for a mold transformer according to an embodiment of the present invention. Referring to FIG.
2 is a view showing an algorithm for determining an insulation defect of a mold transformer according to an embodiment of the present invention.
3 shows a waveform of a deteriorated signal acquired from four sensors in an oscilloscope envelope mode according to Experimental Example 1. FIG.
4 shows a waveform of a deteriorated signal acquired from four sensors in the oscilloscope peak mode according to Experiment 1.
FIG. 5 shows a result of signal processing of the deteriorated signal according to Experimental Example 2 using various wavelet transform techniques.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited to or limited by the exemplary embodiments. Like reference numerals in the drawings denote members performing substantially the same function.

The objects and effects of the present invention can be understood or clarified naturally by the following description, and the purpose and effect of the present invention are not limited by the following description.

The objects, features and advantages of the present invention will become more apparent from the following detailed description. In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail.

1 is a conceptual diagram of an insulation defect determination apparatus 1 of a mold transformer according to an embodiment of the present invention. Referring to FIG. 1, an insulation defect determination apparatus 1 of a mold transformer may include a sensor 20, a signal processing module 301, and an analysis module 303. The sensor 20 can detect the partial discharge of the mold transformer 10 and acquire the deterioration signal of the mold transformer 10. [

The term "partial discharge" is a generic term for a discharge occurring in a part without occurring between the electrode and the electrode. Partial discharge is a local electrical discharge phenomenon that occurs in the insulation system between conductors. This can be understood as a very short dielectric breakdown process, which is caused by the formation of a high electric field inside and outside of the insulator. If there are defects such as voids or foreign matter inside the insulator, the electric field will concentrate on the part, and the electric tree will grow, which may eventually lead to insulation breakdown. These partial discharges can be classified into corona discharge, surface discharge, and internal discharge depending on the electric field concentration region. When a partial discharge occurs, initial signs are shown by electric pulse current, electromagnetic wave, sound, and chemical reaction, and the degree of deterioration of the electric power facility can be confirmed by detecting a signal at a place where a partial discharge occurs.

Hereinafter, a signal detected at a place where a partial discharge is generated is referred to as a deteriorated signal. In the embodiment of the present invention, the degree of deterioration is determined by using the frequency of occurrence of the deterioration signal, the magnitude of the discharge, and the like. This can be useful in the field as an accident predicting technology for electric power facilities composed of solid insulators such as mold transformers.

As the sensor 20 for acquiring the deteriorated signal, an HFCT sensor, a TEV sensor, or a helical antenna may be used. The HFCT sensor and the TEV sensor should be installed in contact with the ground wire or the outer surface of the mold transformer 10. Contact sensors, such as HFCT sensors and TEV sensors, can introduce large amounts of noise from ground lines or contact surfaces. Further, due to the characteristics of the contact type sensor, it is difficult to distinguish the noise introduced from the cable or the noise generated in the mold transformer 10.

In this embodiment, the partial discharge measurement of the mold transformer 10 can use the helical antenna 20. The helical antenna 20 is a receiver mainly used in RF communication, and there is no example yet applied to a mold transformer. The helical antenna 20 has a frequency band of 500 MHz, is simple to install and measure, and is inexpensive and suitable as a measurement sensor according to the present embodiment.

The helical antenna 20 can detect the partial discharge in a noncontact manner within a distance of about 25 cm from the mold transformer 10 to receive the deteriorated signal. When the helical antenna 20 is used as a sensor, the noise introduced from the peripheral equipment of the mold transformer 10 can be cut off. In addition, when the helical antenna 20 is used as a sensor, the signal strength according to the sensing distance is confirmed, and the portion where the partial discharge is generated can be easily detected.

The detection principle of the partial discharge is as follows. Partial discharge of the epoxy resin, which is an insulator of the mold transformer 10, due to a void or a foreign substance causes micro discharge in which charge is accumulated on the initial void surface. As shown in [Figure 1], the microdischarge progresses deterioration with the electric tree over time.

[Figure 1]

The void and discharge phenomenon inside the solid insulator can be expressed as an electrical equivalent circuit consisting of capacitors as shown in [Figure 2].

[Figure 2]

가 인가되는 동안 방전현상이 발생하지 않으면 커패시턴스 c에 가해지는 전압은

와 같다.

가 방전개시전압에 도달하면 커패시턴스 c에서 정극성 방전이 발생하며 매우 짧은 방전 지속시간(

s)에 전압

가 강하된다. 이 같은 부분방전 과정은 정현파를 가진 인가전압에 의해 지속 반복된다. 보이드 표면에 전하가 축적되고 방전개시전압과 함께 축적된 전하는 방전펄스 전류로 흐르게 된다. 이처럼 부분방전의 방전개시전압, 방전전하량, 방전펄스의 빈도수, 방전 에너지 등에 관한 다양한 정보를 검출함으로써 부분방전을 측정할 수 있다.Insulation defects such as voids can be represented by a capacitance c divided by the breakdown path, a capacitance b connected in parallel, and a capacitance a, which is a healthy part. In the mold transformer 10,

If the discharge phenomenon does not occur while the voltage is applied, the voltage applied to the capacitance c

.

Reaches a discharge start voltage, a positive discharge occurs in the capacitance c and a very short discharge duration (

s)

Lt; / RTI > This partial discharge process is continuously repeated by an applied voltage having a sinusoidal wave. Charge is accumulated on the surface of the void, and the electric charge accumulated together with the discharge start voltage flows into the discharge pulse current. As described above, the partial discharge can be measured by detecting various information on the discharge start voltage, the discharge charge amount, the frequency of the discharge pulse, the discharge energy, and the like.

The signal processing module 301 may receive a deterioration signal from the sensor 20. [ The deterioration signal detected from the partial discharge generally includes a large number of noise in a low frequency band. In this case, it is difficult to judge an insulation defect such as insulation breakdown of the mold transformer 10 in a deteriorated signal including noise. The signal processing module 301 can remove noise existing in the deteriorated signal so as to determine whether the mold transformer 10 is insulated.

In this embodiment, the signal processing module 301 can remove noise existing in the deteriorated signal using a normal wavelet transform (Wavelet Transform) technique. Wavelet Transform (Wavelet Transform) is a technique proposed by Morlet, which extracts partial scale components of a signal by applying a function that changes its size in the time-frequency domain. This is advantageous for simultaneous signal analysis in time domain and frequency domain and for analyzing discontinuous signals. Discrete wavelet transform (DWT), wavelet packet transform (WPT), and stationary wavelet transform (SWT) are used for wavelet transformation.

The discrete wavelet transform (DWT) down-samples the degraded signal by applying a low-pass filter and a high-pass filter. That is, the deteriorated signal is divided into an approximated coefficient signal in which a low frequency band is filtered and a detailed coefficient signal in which a high frequency band is filtered. Thereafter, the low-pass filter and high-pass filter are applied again by the number of subcarriers whose size changes (basis function). The number of times of performing the degradation signal decomposition is determined according to the decomposition level. The decomposition level can be set so that the low frequency corresponding to the noise is sufficiently removed. However, when the discrete wavelet transform (DWT) technique is used to remove noise, there is a disadvantage in that the number of sampled signals is reduced and a deteriorated signal is lost.

On the other hand, the normal wavelet transform (SWT) technique may be preceded by an up-sampling process on each scale before the convolution with the number of the downsampled movers, so that the number of samples of the degraded signal can be preserved. Accordingly, it is effective that the signal processing module 301 removes the noise existing in the deteriorated signal using the normal wavelet transform (SWT) technique. The signal processing module 301 can set the decomposition level and filter the noise present in the degraded signal based on the frequency. Thereafter, the signal processing module 301 may set a threshold value to filter noise existing in the degraded signal based on the size. The signal processing module 301 may inversely wavelet-transform the deteriorated signal after the filtering to convert the deteriorated signal into a time domain. The degraded signal converted into the time domain can be transmitted to the analysis module 303. The filtering principle of the signal processing module 301 will be described later with reference to FIG.

The analysis module 303 can determine whether the mold transformer 10 is insulated from the waveform of the deteriorated signal from which the noise is removed. In this case, the degraded signal is a function of the inverse wavelet transformed time domain. The analysis module 303 can determine the insulation defect of the mold transformer through the peak value of the deteriorated signal and the rate of change of the partial discharge coefficient derived from the pulse value. The definition and judgment principle of the partial discharge coefficient will be described later with reference to FIG.

2 is a view showing an algorithm for determining an insulation defect of the mold transformer 10 according to the embodiment of the present invention. Referring to FIG. 2, a method of determining an insulation defect of the mold transformer 10 may include a measurement step S10, a signal processing step S20, and an analysis step S30.

In the measurement step S10, a partial discharge generated in the mold transformer 10 is detected to acquire a deteriorated signal. The deterioration signal can be acquired using the measurement sensor 20 such as the above-described HFCT, TEV, or helical antenna. In the present embodiment, it is preferable that the measuring step S10 detects partial discharge in a non-contact manner from the mold transformer 10 by using the helical antenna 20. [

Generally, the method of detecting the partial discharge in the mold transformer 10 requires a different detection method and approach from the heavy electric machine such as the GIS, the power cable, and the inflow transformer because of the insulating property of the epoxy which is a solid insulator. The test method for partial discharge is specified in the international standard IEC 60270, and it is specified that the partial discharge is tested using a coupling capacitor sensor as an electrical detection method. In this standard, the upper limit frequency is recommended to be 500 kHz or less for the measurement frequency band, and the upper limit frequency is recommended to be 1 MHz for the narrow frequency band. Detection at this frequency is not consistent with the field conditions because of the large external noise input such as noise in the field. Recently, IEC 62478, a standard for measuring partial discharge by non-electrical methods such as acoustic and electromagnetic waves, has been defined. In IEC 62478, unlike IEC 60270, the partial discharge test can be measured by acoustic and electromagnetic waves. The frequency is divided into the HF / VHF band (3 ~ 300MHz) and the UHF band (300MHz ~ 3GHz) have. Therefore, when measuring the partial discharge in the mold transformer (10), frequency band is selected in accordance with the test method of IEC 62487, and accurate partial discharge can be measured by using a detection sensor corresponding thereto.

In the field, various measurement sensors have been applied as a method of detecting partial discharge of electric power facilities. In the embodiment according to the present invention, TEV, HFCT, and helical antenna are used to provide a new measuring method for measuring the insulation defect of the mold transformer 10. The measuring step S10 can measure the deteriorated signal on the high-voltage side by applying a rated voltage to the secondary winding of the mold transformer 10 and applying a high voltage to the primary side.

Experiment 1 - Partial discharge detection

In Experimental Example 1, a single-phase 30 kVA (input: 220 V, output: 0 to 300 V) was used as an input voltage source in a mold transformer 10 having a single-phase 30 kVA capacity and a voltage was injected into the secondary side of the mold transformer 10 in an induction voltage regulator . In this case, 22.9 kV is induced on the first high-voltage side of the mold transformer 10, so that a partial discharge occurs at the defective portion. A capacitor, a HFCT, a TEV, and a helical antenna are used as the sensor 20 for detecting a deteriorated signal generated from a defect.

The measurement step S10 includes a coupling capacitor connected to the primary side of the transformer as shown in Fig. 3, an HFCT installed on the ground side of the enclosure 10 of the mold transformer 10, a TEV installed on the upper- The deterioration signal can be measured from a helical antenna disposed at a suitable distance (about 25 cm) from the high-pressure epoxy resin of the transformer 10.

[Figure 3]

In the presence of an insulation defect in the mold transformer 10, when a voltage is applied to the secondary winding, a high voltage is induced in the primary side, and a deterioration signal is generated in the insulation defect portion. In this experiment, in the measurement step (S10), four measurement sensors including a coupling capacitor, a TEV, a HFCT, and a helical antenna detect a partial discharge generated at 22.9 kV induced on the primary side to acquire a deteriorated signal.

The data obtained in the measuring step S10 were measured under two conditions, respectively, in the envelope mode and the peak mode of the oscilloscope. In the Envelope mode, the acquisition time of data was set to 4 ms / div, and 10 sampling data were stored in 100 ms (6 cycles).

Results of Experiment 1

3 shows waveforms of the deteriorated signals acquired from the four sensors in the oscilloscope envelope mode. 4 shows waveforms of the deteriorated signals acquired from the four sensors in the oscilloscope peak mode. 3 and 4, the partial discharge generated by the foreign substance in the epoxy of the mold transformer 10 repeats charging and discharging at the discharge start voltage and the discharge decay voltage repeatedly for a discharge duration of 16.67 ms ( It can be confirmed that the same discharge pattern is formed every one cycle.

It is verified that even if the helical antenna 20 is used as a sensor, the deterioration signal of the mold transformer 10 can be measured in the same manner as the other three sensors.

The signal processing step S20 is a step performed in the signal processing module 301, and removes noise existing in the deteriorated signal acquired in the measuring step S10. In this case, as an algorithm for removing noise in the signal processing step S20, a stationary wavelet transform technique may be used. In the present embodiment, the signal processing step S20 includes the steps of (a) filtering noise on the basis of a frequency by setting an decomposition level using a normal wavelet transform technique, filtering a noise based on a size by setting a threshold value b) performing an inverse wavelet transform on the filtered impulse signal, and (c) inverse wavelet transforming the filtered impulse signal.

In the step (a), the degraded signal is decomposed into a low frequency band and a high frequency band according to the wavelet transform. The decomposition process can be performed more than once, and thus the low frequencies can be removed step by step. The decomposition level means the number of times the decomposition process is performed. The user can appropriately select the decomposition level so as to eliminate the noise in the low frequency band without seriously distorting the original signal. For example, the decomposition level may be set to 5 or less.

After eliminating low-frequency noise existing in the degraded signal, step (b) may perform a second-order filtering process based on the size of the noise. In this case, a threshold setting to be a reference of the magnitude is required. Applying the threshold can remove more noise, but it also removes the original deterioration signal inherent in the noise. Therefore, it is necessary to set an optimal threshold value so as not to disturb the deteriorated signal, while removing more noise.

In the present embodiment, the threshold value of the signal processing step S20 may be determined according to the following Equation 1 by applying the universal threshold rule proposed by Donoho and Johnstone.

[Equation 1]

,

,

는 임계값,

는 j번째 분해레벨에서 계수의 길이를 의미한다. 임계값을 적용한 후 임계값 이상의 노이즈가 제거되면 (c)단계는 역 웨이블릿 변환을 통해 열화신호를 시간영역으로 복원할 수 있다.
Where MAD is the median absolute deviation coefficient,

Lt; / RTI >

Is the length of the coefficient at the jth decomposition level. If the noise above the threshold value is removed after applying the threshold value, step (c) may restore the degraded signal to the time domain through inverse wavelet transform.

Experimental Example  2 - noise  remove

In order to confirm the effect of the noise removal in the degraded signal in the signal processing step S20, three experiments were performed using discrete wavelet transform (DWT), wavelet packet transform (WPT) and normal wavelet transform (SWT) . The wavelet transform simulation is shown in Table 1 below.

Method Transform Wavelet Threshold A DWT db4 Universal thr B DWT db4 Thr = 0.09 C WPT db4 Universal thr D WPT db4 Thr = 0.09 E SWT Bior2.2 Universal thr F SWT Biro2.2 Thr = 0.09

Under the above conditions A and B were selected as db4 (Daubechies) as the noise reduction condition of the DWT, and the threshold value was simulated under the condition of 0.09 according to [Equation 1]. C and D are the noise removal conditions of WPT, and set the number and the threshold to be the same as A and B. E and F are the noise removal conditions by SWT, and Bior2.2 (Coiflets) is selected as the number of the noise, and the threshold value is applied same as the previous condition. The decomposition level was set to 5.

Results of Experiment 2

The deteriorated signals generated by the insulation defects of the mold transformer 10 were analyzed by DWT, WPT, and SWT to remove noise. The following [Figure 4] shows the deterioration signal including the noise acquired in the measurement step S10.

[Figure 4]

Fig. 5 shows the result of signal processing of the degraded signal of [Fig. 4] measured by the method of DWT, WPT, and SWT. Referring to FIG. 5, it can be seen that the noise cancellation effect is excellent when the normal wavelet transform (SWT) technique is used.

The analysis step S30 is a process performed in the analysis module 303. In the signal processing step S20, the insulation defect of the mold transformer 10 is determined from the waveform of the deteriorated signal from which the noise is removed. Referring to FIG. 5, a discharge signal having a predetermined peak value in a baseband is detected as a deteriorated signal from which noise is removed.

In the embodiment of the present invention, as a criterion for judging whether or not an insulation defect of the mold transformer 10 is determined from such a discharge signal, a concept of a partial discharge coefficient calculated through a peak value of a deteriorated signal and a parameter of a pulse value Respectively. The analysis step S30 can determine the insulation defect of the mold transformer 10 based on the rate of change of the partial discharge coefficient derived from the following equation (2).

&Quot; (2) "

는 부분방전 계수이고,

은 상기 열화신호의 함수이고,

는 상기 열화신호의 1 주기에서 양수 범위의 펄스 값이고,

은 상기 열화신호의 1 주기에서 음수 범위의 펄스 값이고,

은 상기 측정단계에서 열화신호를 취득한 시간의 길이이고,

는 상기 열화신호의 주파수를 의미한다. 즉, 부분방전 계수(

)는 노이즈가 제거된 열화신호의 피크 값(

)과 펄스 값(

) 성분으로부터 도출될 수 있다. here

Is a partial discharge coefficient,

Is a function of the deteriorated signal,

Is a positive pulse value in one cycle of the deteriorated signal,

Is a pulse value in a negative range in one cycle of the deteriorated signal,

Is a length of time obtained by acquiring a deteriorated signal in the measuring step,

Means the frequency of the deteriorated signal. That is, the partial discharge coefficient (

) Is the peak value of the deteriorated signal from which the noise is removed

) And the pulse value (

) ≪ / RTI >

,

의 잔류하는 노이즈의 고려 및 시스템적 설계를 고려하여 카운팅 되는 펄스의 양수범위와 음수범위는 다음과 같이 설정될 수 있다. In this embodiment,

,

The positive range and the negative range of the pulses counted in consideration of the consideration of the residual noise of the signal and the systematic design can be set as follows.

)의 변화율이 0보다 큰 경우 절연결함으로 판단할 수 있다. 본 발명에 따른 실시예에서 하나 이상의 측정 센서(20)를 이용할 경우, 복수의 측정 센서(20)로부터 도출된 부분방전 계수 중 2개 이상이 0보다 큰 변화율을 보이는 경우 절연결함으로 판정할 수 있다. The analysis step S30 is a step of calculating the partial discharge coefficient (

) Is greater than 0, it can be judged as an insulation defect. When at least one measurement sensor 20 is used in the embodiment according to the present invention, it can be judged as an insulation defect when two or more of the partial discharge coefficients derived from the plurality of measurement sensors 20 show a rate of change larger than 0 .

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. will be. Therefore, the scope of the present invention should not be limited to the above-described embodiments, but should be determined by all changes or modifications derived from the scope of the appended claims and equivalents of the following claims.

1: Insulation defect judgment device of mold transformer
10: Mold transformer
20: Helical antenna
301: Signal processing module
303: Analysis module

Claims (9)

A measuring step of detecting a partial discharge generated in the mold transformer and acquiring a deterioration signal of the mold transformer;
A signal processing step of removing noise existing in the deteriorated signal; And
And an analysis step of determining an insulation defect of the mold transformer from the waveform of the deteriorated signal from which the noise is removed.
The method according to claim 1,
Wherein the measuring step comprises:
Wherein a partial discharge is detected in a non-contact manner from the mold transformer by using a helical antenna.
The method according to claim 1,
Wherein the signal processing step comprises:
And the noise existing in the deteriorated signal is removed using a normal wavelet transform (Stationary Wavelet Transform) technique.
The method of claim 3,
Wherein the signal processing step comprises:
(a) setting a decomposition level to filter noise based on a frequency;
(b) filtering the noise based on the size by setting a threshold value; And
(c) performing an inverse wavelet transform on the filtered deteriorated signal.
The method according to claim 1,
Wherein the analyzing step comprises:
Wherein an insulation defect of the mold transformer is determined using a peak value and a pulse value of the deteriorated signal.
6. The method of claim 5,
Wherein the analyzing step comprises:
Determining an insulation defect of the mold transformer through a change rate of a partial discharge coefficient,
Wherein the partial discharge coefficient is derived by the following equation.
[Mathematical Expression]

here

Is a partial discharge coefficient,

Is a function of the deteriorated signal,

Is a positive pulse value in one cycle of the deteriorated signal,

Is a pulse value in a negative range in one cycle of the deteriorated signal,

Is a length of time obtained by acquiring a deteriorated signal in the measuring step,

Is the frequency of the deteriorated signal.
The method according to claim 6,
Wherein the analyzing step comprises:
Wherein the mold transformer is determined as an insulation defect when the rate of change of the partial discharge coefficient is greater than zero.
A helical antenna for detecting a partial discharge generated in the mold transformer and acquiring a deterioration signal of the mold transformer;
A signal processing module for removing noise existing in the deteriorated signal using a normal wavelet transform (Wavelet Transform) technique; And
And an analysis module for determining an insulation defect of the mold transformer from the waveform of the deteriorated signal from which the noise is removed.
9. The method of claim 8,
Wherein the analysis module determines an insulation defect of the mold transformer through a change rate of a partial discharge coefficient,
Wherein the partial discharge coefficient is derived by the following equation.
[Mathematical Expression]

here

Is a partial discharge coefficient,

Is a function of the deteriorated signal,

Is a positive pulse value in one cycle of the deteriorated signal,

Is a pulse value in a negative range in one cycle of the deteriorated signal,

Is a length of time obtained by acquiring a deteriorated signal in the measuring step,

Is the frequency of the deteriorated signal.

Images