KR101553005B1 - A partial discharge monitoring and diagnosis system for power devices - Google Patents
A partial discharge monitoring and diagnosis system for power devices Download PDFInfo
- Publication number
- KR101553005B1 KR101553005B1 KR1020150055313A KR20150055313A KR101553005B1 KR 101553005 B1 KR101553005 B1 KR 101553005B1 KR 1020150055313 A KR1020150055313 A KR 1020150055313A KR 20150055313 A KR20150055313 A KR 20150055313A KR 101553005 B1 KR101553005 B1 KR 101553005B1
- Authority
- KR
- South Korea
- Prior art keywords
- partial discharge
- signal
- high frequency
- discharge
- sensor
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R23/00—Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
- G01R23/02—Arrangements for measuring frequency, e.g. pulse repetition rate; Arrangements for measuring period of current or voltage
- G01R23/06—Arrangements for measuring frequency, e.g. pulse repetition rate; Arrangements for measuring period of current or voltage by converting frequency into an amplitude of current or voltage
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R23/00—Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
- G01R23/16—Spectrum analysis; Fourier analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1227—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials
- G01R31/1263—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials of solid or fluid materials, e.g. insulation films, bulk material; of semiconductors or LV electronic components or parts; of cable, line or wire insulation
- G01R31/1272—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials of solid or fluid materials, e.g. insulation films, bulk material; of semiconductors or LV electronic components or parts; of cable, line or wire insulation of cable, line or wire insulation, e.g. using partial discharge measurements
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R35/00—Testing or calibrating of apparatus covered by the other groups of this subclass
Abstract
Description
The present invention estimates partial discharge detection by insulation failure, connection failure, or disconnection in a power device housing such as a high voltage cable, a transformer, a gas insulated switchgear (GIS), a switchgear, an electric distribution panel, By detecting a partial discharge generated during partial discharge by using a UHF (Ultra High Frequency) sensor and a high frequency current transformer (HFCT), it is possible to detect a partial discharge by using RBFNN (Radial Basis Function Neural Network) To a partial discharge monitoring diagnostic system of a power plant.
Further, the present invention is characterized in that an electric signal is measured by a partial discharge in an electric power facility of an electric power equipment, a characteristic used for judging a partial discharge is extracted from an electric signal, and the type of the partial discharge The present invention relates to a partial discharge monitoring diagnostic system of a power plant.
In addition, the present invention is constructed by a hierarchical SCADA system (centralized remote monitoring and control system), and a plurality of transient ground voltage sensors or microwave sensors are installed in an electric facility or the like, and the sensors are controlled by a
Generally, partial discharge refers to any of electric power equipment installed in various industries and power system substations, electric power equipment such as high voltage switchboard, high voltage cable, transformer, GIS (Gas Insulated Switchgear), switchgear, Discharge, a surface discharge occurring along the surface of an insulator, a void discharge occurring in a gap in an insulator, and the like.
It is very important to monitor the abnormality of the power equipment and to monitor the degree of deterioration of the insulator and to predict the repair time, and it is possible to predict and manage such by measuring and monitoring the partial discharge. For this purpose, a partial discharge of various power facilities such as high voltage cable, transformer, gas insulated switchgear (GIS), switchgear, power supply facility, high voltage panel, low voltage panel, motor control panel, Measuring devices are being used.
Electricity demand is continuously increasing due to the advancement of industry, and accidents caused by electric power facilities in high-voltage cables, transformers, GIS, power receiving facilities and power equipment systems are frequently occurring. Such accidents can cause technical losses as well as economic losses. Especially, high-voltage power equipment is regarded as a very important facility not only in the national infrastructure but also in civilian use, and the accident of power equipment is a nationally serious issue affecting the damage of human life, direct loss of electric power facilities, . Power equipment such as transformer, cable, breaker, switchgear, high voltage cable, transformer, GIS, PT, CT, arrester, etc. are installed in power system installed in power plants, substations and large factories. , Local discharge phenomena (partial discharge) caused by deterioration of the dielectric strength of the insulation, etc., are leading to accidents, which have a serious effect on the power supply system.
In addition, various kinds of insulators are used in such power devices in order to prevent a discharge phenomenon occurring in a high voltage situation. However, such insulation may cause gaps such as voids or delaminations during cooling and heating during operation for some reason or during operation. However, such a gap generates a partial electric discharge every time a high electric field is applied, and if such a partial discharge is repeated, the insulation is gradually eroded and the dielectric strength is reduced, resulting in serious dielectric breakdown. In order to solve this problem, it is desirable to remove the gap in the insulating material in advance to reduce the occurrence of partial discharge, but it is difficult to completely remove the gap in consideration of various reasons. In addition, the insulation characteristics of the insulator must be sufficiently inspected from the time of manufacture. Such inspections are effective for inspection of initial manufacturing defects, but insulative deterioration over time occurs during operation of the power equipment system, so that it is difficult to conduct substantial inspections. Therefore, in the past, the time interval between inspections increases, and it is impossible to accurately grasp the insulation characteristic at all times, resulting in an unexpected serious accident. In addition, the deterioration can be monitored by measuring the partial discharge.
Partial discharge in electric power equipment is a phenomenon that occurs when insulation deteriorates in electric power equipment, and most of them are generated in the last stage of insulation deterioration, and thus it is evaluated as the best method for deterioration diagnosis.
The partial discharge generated in the power equipment is not easy to detect due to the wide band of the electromagnetic wave signal and surrounding noise. As a detection method, a method using a UHF antenna has been proposed (
In addition, a technology for automatically analyzing a partial discharge using a neural network has been proposed [Patent Document 5].
On the other hand, among electric power facilities in electric power equipment, especially in the case of transformers, there is a very high probability that a partial electric discharge is likely to occur while being an important electric power facility of the water power distribution facility. In the case of an input transformer, effective measurement of the partial discharge signal generated by the transformer enables stable operation and cost reduction of the input transformer. However, the partial discharge detection method of the conventional inflow transformer includes a detection method using a coupling device, a partial discharge detection method using a gas analysis method, and an electromagnetic wave detection analysis method. Such a detection method according to the prior art has a disadvantage in that it is difficult to measure in real time and in the limit of the measurement, and in addition, the price is expensive and the competitiveness is remarkably deteriorated.
In addition, epoxy resin is the most used insulation material in the mold type transformer. Epoxy resins are widely used as insulating materials because they have high mechanical strength, excellent electrical insulation properties, and are relatively easy to mold. However, when the material is molded or a device is manufactured, foreign matter may enter, or micro voids or cracks may be generated. Deterioration diagnosis technique using partial discharge measurement has been actively studied, but research on partial discharge measurement and circuit technology suitable for electric power equipment system is still insufficient.
It is necessary to study the partial discharge measurement technology suitable for the electric power equipment such as the inflow transformer, the instrument transformer, the breaker, and the mold transformer used in the power equipment system, and it is necessary to develop an amplification circuit suitable for the measurement sensor. In other words, it is necessary to develop a technique for partial discharge detection method and amplification circuit generated in a power equipment system power facility so that it can be utilized for deterioration monitoring and diagnosis of a power equipment system.
SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems and to provide a power supply system in which a high-voltage cable, a transformer, a GIS (Gas Insulated Switchgear), a switchgear, Which detects a partial discharge occurring at the time of partial discharge by using a UHF sensor and a high frequency current sensor and determines the type of partial discharge using RBFNN (Radial Basis Function Neural Network) And to provide a partial discharge monitoring diagnostic system of the facility.
It is also an object of the present invention to provide a method and apparatus for measuring an electrical signal by partial discharge in a power plant of a power equipment system, extracting characteristics used for judging a partial discharge from an electrical signal, And to provide a partial discharge monitoring diagnostic system for a power plant.
Particularly, it is an object of the present invention to provide a high frequency current transformer (HFCT) for detecting an electromagnetic wave generated during a partial discharge by using a high frequency antenna sensor and a high frequency current transformer The present invention provides a partial discharge monitoring diagnostic system for a power plant that extracts characteristics from an electrical signal output from a sensor, and applies the extracted characteristics to RNFNN to determine types of void discharge, corona discharge, and surface discharge.
In order to accomplish the above object, the present invention provides a power supply system including a high-voltage cable, a transformer, a gas insulated switchgear (GIS), a switchgear, a power reception facility, A UHF detector for detecting an electrical signal of a very high frequency generated by a partial discharge in the power plant using microwave of 300 MHz to 3 GHz, And an HFCT (High Frequency Current Transformer) sensor for detecting an electrical signal of a high frequency current generated by a partial discharge, which is mounted on a ground line of a power facility provided inside the housing. And an electric signal of a high-frequency current detected by the sensor unit to obtain sampling data, extracting a characteristic quantity from the sampling data, and outputting the characteristic quantity to the RBFNN (Radial Basis Function Neural Network) And a fuzzy inference unit configured to determine whether a partial discharge is caused or a type of partial discharge according to an output value obtained by inputting a feature quantity, wherein the fuzzy inference unit comprises: a fuzzy inference unit for generating an electrical signal output from the UHF sensor or the HFCT sensor A signal detector for extracting and sampling a partial discharge signal and outputting sampling data of a partial discharge pulse signal by an electrical signal of a very high frequency or a high frequency current; A feature amount extracting unit for extracting a feature amount necessary for determining the type of partial discharge and partial discharge from the sampling data; The RBFNN neural network includes RBFNN inference (RBFNN), which is composed of a plurality of outputs so as to correspond to each of a plurality of partial discharge types, by applying an extracted characteristic quantity to an input to a RBFNN (Radial Basis Function Neural Network) part; And a partial discharge determination unit for determining the type of partial discharge or partial discharge according to output values of the plurality of outputs of the RBFNN neural network.
The signal detection unit extracts 5,000 data of each of the electrical signals of the very high frequency or the high frequency current at intervals of 0.5 seconds at a sampling rate of 10 kHz / s to generate sampling data .
The signal detection unit may comprise a coupling network and a low-noise high-frequency amplifier for amplifying a small signal to pass only partial discharge pulses of the very high frequency electrical signals to an amplifier, Wherein the low-noise high-frequency amplifier comprises a low-noise wide-band operational amplifier in two stages, configured to not limit the high-frequency cutoff frequency with high gain, and configured to amplify an electrical signal of the high- And the detection circuit is calibrated by using a discharge charge quantity of a known size in order to calculate the magnitude of an apparent charge of the partial discharge pulse in the electrical signal of the microwave or the high frequency current .
The present invention relates to a partial discharge monitoring diagnostic system for a power plant, wherein the feature quantity extracting unit is characterized by comprising: a feature quantity extracting unit for extracting, as the feature quantity, a kurtosis of the very high frequency sampling data obtained from the very high frequency electric signal, ) Converted into a frequency field and converted into a frequency field by FTT conversion to calculate an average value of frequency-based sizes calculated by converting the frequency field, a kurtosis of the high-frequency current sampling data obtained from the electrical signal of the high-frequency current, The asymmetry degree of the signal with respect to the waveform data of the high frequency current sampling data and the calculated total harmonic distortion factor after calculating the 1st to 15th harmonic with respect to the waveform data of the high frequency current sampling data, And extracts a feature quantity included in the feature quantity All.
The RBFNN inference unit comprises a neural network composed of an input layer, a hidden layer, and an output layer, wherein a radial basis function is used for the hidden layer, Wherein the input layer receives the feature quantity as a vector and outputs the sum of values obtained by multiplying output values of outputs output by the function of the hidden layer by connection weights in the output layer, do.
[Equation 1]
Where x is the input vector of the feature quantity, y j is the jth output of the output layer, c is the number of clusters corresponding to the fuzz rule of the hidden layer, and a ij is the jth Output connection weights, and v i is the center of the i-th cluster.
The present invention relates to a partial discharge monitoring diagnostic system of a power plant, wherein the center v i of the cluster is a sample input vector set X = {x 1 , x 2 , ..., x N }, x k ∈ R n characterized in that the center while modifying the u ik represents the center and by the membership degree (n is the size of the input vector) obtained by converging the objective function of the following [equation 2] as a minimum.
[Equation 2]
,
,
Where ∥ · ∥ denotes the weighted Euclidean distance, c is the number of clusters, m is the fuzzification coefficient, and X k is the k-th input vector.
In the partial discharge monitoring diagnosis system of the electric power facility, the partial discharge judgment unit calculates a maximum value among a plurality of output values calculated by the RBFNN speculation unit, judges normal if the maximum value is 0.5 or less, When the output having a maximum value of 0.5 and a maximum value is y 1 , a corona discharge is determined to be y 2 , and a surface discharge is determined to be y 3 .
As described above, according to the partial discharge monitoring diagnostic system of the electric power facility according to the present invention, the partial discharge occurring at the partial discharge is detected by using the UHF sensor and the high frequency current sensor, and the type of the partial discharge is judged using the RBFNN This makes it possible to more accurately and easily perform deterioration monitoring and diagnosis of the power equipment system.
1 is a block diagram of a configuration of a partial discharge monitoring diagnostic system of a power plant according to an embodiment of the present invention.
2 is a block diagram of a configuration of a monitoring apparatus according to an embodiment of the present invention;
3 is a block diagram of a configuration of a signal detection unit and a reasoning determination unit according to an embodiment of the present invention;
FIG. 4 is a graph illustrating frequency characteristics of a coupling network for allowing only partial discharge pulses of signals of a UHF sensor according to an embodiment of the present invention to pass through an amplifier. FIG.
5 is a circuit diagram (a) of an amplifier circuit for detecting breaker partial discharge according to an embodiment of the present invention, and FIG. 5 (b) is a circuit photograph.
6 is a configuration diagram of a calibration test system according to an embodiment of the present invention;
FIG. 7 is a graph showing an example of a detection calibration pulse waveform according to an embodiment of the present invention, and is a graph in which [50 mV / div, 500 ns / div] range.
8 illustrates a configuration of an experimental system according to an embodiment of the present invention.
Fig. 9 is a waveform graph of (a) power supply voltage-partial discharge pulse and (b) partial discharge pulse as an example of the detected waveform according to the experiment of the present invention, (50mV / div, 5ms / div), and Ch.2 is a partial discharge pulse (50mV / div, 5ms / div) .
10 is a circuit diagram (a) of the amplifying circuit for HFCT according to an embodiment of the present invention, and Fig. 10 (b) is a circuit photograph.
11 is a configuration diagram of a calibration experiment system according to an experiment of the present invention.
12 is a graph of a calibration pulse waveform (100 [pC]) according to an experiment of the present invention, with a graph in the range [100 mV / div, 2 μs / div].
13 is a diagram illustrating a partial discharge detection method using HFCT according to an experiment of the present invention.
FIG. 14 is a graph of a measured waveform according to an experiment of the present invention. The graph above is a graph for an applied voltage [5 kV / div, 5 ms / div] and the graph below shows a partial discharge pulse [50 mV / div, 5 ms / Graph for.
15 is a schematic diagram of a radiator low function neural network according to an embodiment of the present invention.
16 is an illustration of test data for a partial discharge test of the present invention.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the drawings.
In the description of the present invention, the same parts are denoted by the same reference numerals, and repetitive description thereof will be omitted.
First, a configuration of an overall system for implementing a partial discharge monitoring diagnostic system of a power plant according to an embodiment of the present invention will be described with reference to FIG.
1A, an overall system for implementing a system according to the present invention comprises a
Also, as shown in FIG. 1B, the power equipment system according to the present invention can be constructed as a hierarchical SCADA system (centralized remote monitoring and control system). That is, a plurality of
Preferably, one
The
When partial discharge occurs inside the power
The
The high frequency
Power equipment systems are installed in buildings or factories that use a large amount of electric power. They are divided into high voltage cables, transformers, GIS (Gas Insulated Switchgear), switchgear,
The discharge diagnosis system of the electric power system according to the present invention is provided with a sensing means for sensing the state of each device in the
Meanwhile, the
Next, the
That is, the
Preferably, the
The
The
Meanwhile, the
Further, the
Next, the configuration of the
2, the
The
The setting
Next, the
RBFNN (Radial Basis Function Neural Network) is an intelligent algorithm that is widely used for pattern classification, pattern recognition, and system modeling. It uses partition function based on FCM clustering as an active function. It is also represented by the "If-then" fuzzy rule and driven by a fuzzy reasoning mechanism. The conditional, conclusion, and reasoning parts form a network structure. The conditional partitions the input space using FCM clustering, and the conclusion part expresses the partitioned local area as a constant term. Finally, the final output of the network depends on the fuzzy reasoning of the reasoning part.
The
The
When the
If the real-time measurement signal or the characteristic amount or the pattern matches or is similar to the characteristic amount or pattern of the boundary, the alarm is stored in the database and the signal sampling data, the characteristic amount, and the like collected through the
Next, the detailed configuration of the
3, the
The
First, the
The
The
As an example, the
Partial discharge pulses are weak signals in the range of several hundreds of kHz to several hundreds of MHz and are extremely small in the power supply frequency and can not be directly detected because they are exposed to ambient electrical noise. The
5 is a circuit diagram of a low noise amplifier for amplifying a small partial discharge pulse connected to an output of a coupling network. The low-noise broadband (1.6 nV / √Hz, 330 [MHz]) operational amplifier should be configured in two stages to avoid limiting high-frequency cutoff frequencies with high gain. The gain of the designed and fabricated amplifier is 40 [dB], and the frequency band of -3 [dB] can amplify the partial discharge pulse delivered from the coupling network without attenuation from 100 [kHz] to 300 [MHz].
In order to calculate the magnitude of the apparent charge of the detected partial discharge pulse, it is necessary to calibrate the detection circuit using a discharge charge amount of a known size. In the present invention, a calibrated pulse is input between the fixed conductor of the circuit breaker and the enclosure using a dedicated calibration pulse generator (CAL1A, 1 to 100 pC, positive / negative) ).
Fig. 7 shows an example of the waveform detected at the output of the amplifier when a negative pulse of 100 [pC] is applied. Since the output voltage of the amplifier is 98 [mV], it can be seen that the sensitivity of the partial discharge measuring apparatus is 9.8 [mV] / 10 [pC]. Considering the measurement range of the oscilloscope, the oscillator can be stably measured at 10 [pC] (about 20 [mV]) or more.
In the experiment of generating the partial discharge of the circuit breaker, as shown in FIG. 8, a needle electrode was provided to generate partial discharge in the internal fixed electrode conductor, and a high voltage transformer (AC 220 [V] / 20 kV) And the AC 13.2 [kV] corresponding to the voltage was applied. FIG. 9 shows an example of the detected waveform. In this experiment, a maximum of 97 [pC] (190 [mV]) is detected, and a waveform of FIG. 9B corresponds to 76 [pC].
The designed and fabricated partial discharge measurement device was calibrated with 1.96 [mV] / [pC] as a result of calibrating experiment on the circuit breaker. It is possible.
In the application experiment, the partial discharge pulse can be observed by generating the partial discharge arbitrarily by applying 13.2 [kV] between the phase and the ground like the high voltage distribution line.
Next, the
The
The
As an example, the
Also, the
The amplifier circuit performs a characteristic evaluation as a ratio of an output signal to a sinusoidal input signal and has a gain of 8 [dB] and a frequency band of 100 [kHz] to 10 [MHz]. In order to calculate the magnitude of the apparent charge of the detected partial discharge pulse, it is necessary to calibrate the detection circuit using a discharge charge amount of a known size.
Therefore, a calibration experiment is performed using the calibration pulse generator (CAL1A, 1 to 100 [pC]) as shown in FIG. 11 and the sensitivity is calculated from the magnitude of the detection signal for the calibration pulse. An example of the detected waveform when a positive pulse of 100 [pC] is applied is shown in Fig. The output voltage of the amplifying circuit is 120 [mV], and the conversion sensitivity of the partial discharge measuring apparatus is 12 [mV] / 10 [pC].
After performing the sensitivity calculation of the amplification circuit, an experimental system is constructed as shown in FIG. 13, and a partial discharge is generated using a high voltage transformer (AC 220 [V] / 15 [kV]).
The measurement results are shown in Fig. Partial discharge occurred near the maximum value of the applied voltage and partial discharge of maximum 116 [pC] (140 [mV]) when 7.5 [kV] was applied.
Next, the feature
The feature
The first characteristic quantity is Kurtosis for the signal (or sampling data) of the
The second feature quantity is an average of the spectral magnitudes for the signal (or sampling data) of the
The third characteristic amount is a kurtosis with respect to the signal of the high frequency
The fourth characteristic quantity is the average of the spectrum magnitudes for the signals of the high frequency
The fifth characteristic quantity is the total harmonic distortion factor for the signal of the high frequency
The sixth feature is an asymmetry of the signal of the high-frequency
Next, the
The
The basic neural network structure consists of an input layer, a hidden layer, and an output layer. Generally, neural network uses a sigmoid function in hidden layer, and output layer consists of one node. The Radial Basis Function Neural Network (RBFNN) network of the present invention uses a Radial Basis Function for the hidden layer.
In general, the RBFNN or NN is composed of one node in the output layer, but the
The input layer receives the input vector x = [x 1 , x 2 , x 3 , ..., x 6 ] T and transfers it to the hidden layer. Herein, the input vector means a vector composed of six characteristic quantities obtained by the characteristic
The structure of the RBFNN in FIG. 15 can be expressed by the following equation. The number of rules can be determined by experience.
[Equation 1]
The equation for the ith rule: If x is R i then y j = f ji (x)
Here, x denotes an input vector, R i is a membership function of an i (= 1, ..., c) group by FCM (Fuzzy C Means) clustering, f ji (x) Is the constant term of the i-th fuzzy rule for the (= 1, ..., s) th output. The conditional part before "then" performs the function of the active function in the network structure and the function of the belonging function in the linguistic aspect by using FCM clustering. After the "then" conclusion, the constant term is the connection weight of the network, which acts as a local model of the fuzzy rule, and the final output of the network in the inference unit is obtained as the inference result of the fuzzy rule.
In the present invention, the FCM clustering algorithm is applied to the FCM algorithm to calculate the output value to the hidden layer. The conditional function of RBFNN divides the input space by the number of c clusters (number of fuzzy rules) and outputs the degree of membership of each domain as a fuzzy set to reflect the characteristics of learning data. The FCM cluster method assigns membership degree to the distance between each data and the center of a specific cluster, and classifies data according to the membership degree.
FCM clustering obtains the membership value for each cluster of input data by minimizing Equation (2) which is an objective function.
&Quot; (2) "
Here, c is the number of clusters (number of fuzzy rules), N is the number of input patterns, and m is a fuzzification coefficient. X k is the k-th input vector, and v i is the center of the i-th cluster. u ik is a real number between 0 and 1 indicating the degree of belonging to the kth data belonging to the i-th cluster and satisfies the conditions of Equations (3) and (4).
&Quot; (3) "
&Quot; (4) "
∥ · ∥ in Equation (2) uses a weighted Euclidean distance expressed by Equation (5).
&Quot; (5) "
Where j is the standard deviation of the j-th input dimension of the input patterns. Weighted cladian distances are widely used because they provide reasonable distance information that is not significantly affected by the data size distribution. n-dimensional Euclidean space, the input vector set composed of N number of the pattern on the X = {x 1, x 2 , ..., x N}, x k ∈R n, 1≤k≤N and cluster center v = {v 1, v 2, ..., v c} , v i ∈R n, can be expressed as the membership degree of the 1≤i≤c membership matrix is expressed as u = [u ik] ik u and v i equation 6 is And 7, respectively.
&Quot; (6) "
&Quot; (7) "
The FCM clustering modifies the membership matrix U and the center v i (i = 1, ..., c) of each cluster while iterating
The conclusion of the RBFNN structure is to form the rule after "then" in equation (1) by expressing each local region separated from the conditional part as a local session model of the polynomial function. F i (x) is omitted, the subscript j (= 1, ..., s ) for output neurons in ji f (x) of equation (1) each have a constant term, a linear equation form as Equation 8, and 9.
&Quot; (8) "
Finally, the final output of the proposed RBFNN structure is expressed by Equation (9) by fuzzy inference.
&Quot; (9) "
Here, u i is equal to R i (x) in Equation (1), and the sum of all clusters is 1 as shown in Equation (3).
RBFNN should be applied after performing learning using test data. That is, the connection weights a ij between the hidden layer and the output layer.
First, the test data collects input / output value data of RBFNN through partial discharge test, and learns using the collected test data to determine connection weight.
The test data is composed of a vector having the structure shown in FIG.
In the case of the void discharge, the corona discharge, and the surface discharge, the characteristic quantities are extracted for each of the cases. In the case of the void discharge, the data obtained is y1 = 1, the corona discharge = y2 = 1, The test data is obtained by setting y3 to 1 and all values to 0 in the normal case. For the learning of RBFNN, 100 or more experimental data are obtained and applied for each case.
The learning method using test data is as follows.
The learning for the connection weight calculation proceeds to minimize the error expressed by
&Quot; (10) "
Here, q is the error E, N for the second pattern is a q number of input patterns. t q is the target value for the q-th pattern. y q is the final output of the network for the qth pattern and is shown in Equation (11).
&Quot; (11) "
The connection weight a ij between the hidden layer and the output layer is calculated as many times as the number of test data as follows.
&Quot; (12) "
&Quot; (13) "
Where η is the learning rate, α is the momentum coefficient, p + 1 is the new value, p is the current value, and p-1 is the previous value.
a ij is the connection weight between the i th hidden layer and the j th output side, and the connection weights are learned in the same way for all connection weights.
The
Next, the partial discharge determination unit 343 will be described.
The partial discharge determination unit 343 determines the partial discharge from the output value calculated by the
Although the present invention has been described in detail with reference to the above embodiments, it is needless to say that the present invention is not limited to the above-described embodiments, and various modifications may be made without departing from the spirit of the present invention.
10: Power equipment system housing 11: Power equipment
20: Sensor part 21: UHF sensor
22: HFCT sensor
30: Monitoring device 31: Signal detection part
32: display section 33: setting section
34: Reasoning judgment part 35: Alarm part
36: Storage unit 40: Remote server
311: UHF detecting unit 312: HFCT detecting unit
341: Feature extraction unit 342: RBFNN Reasoning unit
343: partial discharge judgment unit
Claims (7)
A UHF (Ultra High Frequency) sensor installed inside the housing for detecting an electrical signal of a very high frequency generated by a partial discharge in the power facility using microwave of 300 MHz to 3 GHz, A sensor unit comprising a HFCT (High Frequency Current Transformer) sensor for detecting an electrical signal of a high frequency current generated by a partial discharge by being attached to a ground line of the equipment; And
A characteristic quantity is extracted from the sampling data by sampling an electrical signal of a microwave high frequency electric current and an electric signal of a high frequency current detected by the sensor unit, And a monitoring device that constitutes a fuzzy inference unit for determining whether the partial discharge is caused or not, according to an output value obtained by inputting the partial discharge,
Wherein the fuzzy inference unit comprises:
A signal detector for receiving an electrical signal output from the UHF sensor or the HFCT sensor to extract and sample a partial discharge signal and output sampling data of a partial discharge pulse signal by an electrical signal of a very high frequency or a high frequency current;
A feature amount extraction unit for extracting a feature amount necessary for determining the type of partial discharge and partial discharge from the sampling data;
The RBFNN neural network includes RBFNN inference (RBFNN), which is composed of a plurality of outputs so as to correspond to each of a plurality of partial discharge types, by applying an extracted feature quantity to an input to a RBFNN (Radial Basis Function Neural Network) part; And
And a partial discharge determination unit for determining a type of partial discharge or partial discharge according to an output value of a plurality of outputs of the RBFNN neural network.
Wherein the signal detecting unit extracts 5,000 pieces of data from the electrical signals of the very high frequency or the high frequency current at intervals of 0.5 seconds at a sampling period of 10 kHz / s to generate sampling data.
And a low-noise high-frequency amplifier for amplifying a small signal, wherein the low-noise high-frequency amplifier comprises a low-noise broadband operational amplifier in two stages, and the high gain So that the high-frequency cut-off frequency is not limited,
And an amplifying circuit for amplifying an electric signal of the high-frequency current, wherein a high-speed operational amplifier is used,
Characterized in that the detection circuit is calibrated using an amount of discharge charge of a known size to calculate the magnitude of the apparent charge of the partial discharge pulse in the electrical signal of the microwave or the high frequency current Discharge monitoring diagnostic system.
Wherein the feature quantity extracting unit extracts, as the feature quantity, a kurtosis of the very high frequency sampling data obtained from the electrical signal of the very high frequency, a magnitude of the frequency by a frequency field calculated by converting the very high frequency sampling data into a frequency field by FTT (Fast Fourier Transformation) Frequency of the high-frequency current sampling data obtained from the electrical signal of the high-frequency current, an average value of the frequency-by-frequency magnitudes calculated by converting the high-frequency current sampling data into a frequency field by FTT conversion, Extracts a characteristic quantity including a total harmonic distortion factor calculated after the 1st to 15th harmonics with respect to the data and an asymmetry degree of the signal with respect to the waveform data of the high frequency current sampling data. Partial discharge monitoring diagnostic system.
Wherein the RBFNN reasoning unit forms a neural network composed of an input layer, a hidden layer, and an output layer, wherein the hidden layer uses a Radial Basis Function, and the output layer includes at least three outputs, And outputs the output value as a sum of values obtained by multiplying the output values of the output by the function of the hidden layer by the connection weight, wherein the output value is calculated by the following equation Partial discharge monitoring diagnostic system of power facilities.
[Equation 1]
In addition, x is a feature quantity input vector, y j is a j-th output of the output layer, c is the number of clusters corresponding to the fuzz rule of the hidden layer, and a ij is a jth V i is the center of the i-th cluster, and m is the fuzzification coefficient.
The center v i of the cluster is centered by a set of sample input vectors X = {x 1 , x 2 , ..., x N }, x k ∈ R n (where n is the size of the input vector) Is a center obtained by converging the objective function Q of Equation (2) to minimize while modifying u ik indicating the degree of affiliation.
[Equation 2]
,
,
Where c is the number of clusters, m is the fuzzification coefficient, X k is the k-th input vector, N is the number of input vectors, v j is the center of the jth cluster.
Wherein the partial discharge determination unit calculates a maximum value among a plurality of output values calculated by the RBFNN inference unit, and determines that the maximum value is less than or equal to 0.5, and determines that the output having a maximum value greater than 0.5 and a maximum value is y 1 Wherein the determination unit determines that the discharge is a void discharge, a corona discharge if y 2 , and a creeping discharge if y 3 .
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020150055313A KR101553005B1 (en) | 2015-04-20 | 2015-04-20 | A partial discharge monitoring and diagnosis system for power devices |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020150055313A KR101553005B1 (en) | 2015-04-20 | 2015-04-20 | A partial discharge monitoring and diagnosis system for power devices |
Publications (1)
Publication Number | Publication Date |
---|---|
KR101553005B1 true KR101553005B1 (en) | 2015-10-01 |
Family
ID=54338377
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
KR1020150055313A KR101553005B1 (en) | 2015-04-20 | 2015-04-20 | A partial discharge monitoring and diagnosis system for power devices |
Country Status (1)
Country | Link |
---|---|
KR (1) | KR101553005B1 (en) |
Cited By (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101641515B1 (en) | 2016-04-05 | 2016-07-22 | (주)서전기전 | The distribution panel including a diagnosis apparatus using HFCT |
KR101715079B1 (en) * | 2016-11-11 | 2017-03-09 | 주식회사 현대콘트롤전기 | A partial discharge detection and diagnosis appratus for a distributing board with the acoustic emission sensor |
CN106772007A (en) * | 2016-12-19 | 2017-05-31 | 天津市品通电力科技有限公司 | A kind of device for high-voltage switch gear live detection |
KR20170088736A (en) * | 2016-01-25 | 2017-08-02 | 한국전자통신연구원 | Electricity-safety managing server, apparatus and method for managing of electricity-safety |
KR101814147B1 (en) * | 2017-02-28 | 2018-01-30 | 골드텍 주식회사 | Distributing board with sensing function for partial discharge |
KR20180042671A (en) | 2016-10-18 | 2018-04-26 | 현대일렉트릭앤에너지시스템(주) | Fault diagnosis apparatus using a dissolved gas generation pattern of the transformer |
CN108732478A (en) * | 2018-09-10 | 2018-11-02 | 清华四川能源互联网研究院 | Partial discharge of switchgear on-line monitoring system based on superfrequency punching current sensor and method |
KR101925338B1 (en) * | 2016-12-22 | 2018-12-05 | 엘에스산전 주식회사 | Diagnostic apparatus for switchgear |
KR20190020233A (en) | 2017-08-17 | 2019-02-28 | 한국전력공사 | Simulated signal generating Apparatus and method for partial discharge simulation |
KR101982623B1 (en) | 2018-02-28 | 2019-06-28 | (주)인피니어 | Partial discharge measuring device and method of high voltage cable joint |
CN110108993A (en) * | 2019-05-31 | 2019-08-09 | 国网青海省电力公司检修公司 | A kind of multiband three-dimensional joint test method for power equipment shelf depreciation |
CN110531228A (en) * | 2019-09-03 | 2019-12-03 | 国网湖南省电力有限公司 | GIS partial discharge extent of disease severity method, system and medium based on principal component dimensionality reduction and clustering |
KR102058230B1 (en) * | 2018-08-27 | 2019-12-20 | 연세대학교 산학협력단 | Device, method and system to predict fault of cable |
KR102079874B1 (en) | 2018-10-05 | 2020-04-13 | 한국전력공사 | Apparatus and method for diagnosing partial discharge |
KR20200119431A (en) * | 2019-04-02 | 2020-10-20 | 현대일렉트릭앤에너지시스템(주) | Apparatus for diagnosis of partial discharge |
CN112924821A (en) * | 2021-01-25 | 2021-06-08 | 广东电网有限责任公司广州供电局 | Composite detection system and method for electric power equipment discharge and heating defect detection |
KR102295214B1 (en) * | 2021-05-31 | 2021-08-31 | 지투파워(주) | Partial discharging detection-diagnosis system of distribution panel using high frequency current transformer |
CN113848435A (en) * | 2021-08-31 | 2021-12-28 | 四川大学 | Direct current partial discharge signal classification and identification method based on frequency spectrum energy distribution probability |
CN114113943A (en) * | 2021-11-25 | 2022-03-01 | 广东电网有限责任公司广州供电局 | Transformer partial discharge detection system, method and equipment based on current and ultrasonic signals |
KR102377939B1 (en) * | 2021-12-24 | 2022-03-23 | 지투파워(주) | Partial discharge monitoring and diagnosis system for distribution board using ultra frequency and high frequency current transformer signal |
KR102377936B1 (en) * | 2021-12-24 | 2022-03-23 | 지투파워(주) | Partial discharge monitoring and diagnosis system for intelligent distribution board using ultra frequency signal |
KR20220089422A (en) * | 2020-12-21 | 2022-06-28 | (주)에이피엠테크놀러지스 | Partial discharge monitoring system and patial discharge monitoring method |
KR20220096957A (en) * | 2020-12-31 | 2022-07-07 | 주식회사 큐아이티 | Method and apparatus for detecting partial discharge in arc-detecting system in power system |
KR20220126930A (en) * | 2021-03-10 | 2022-09-19 | 한전케이디엔주식회사 | Apparatus and method for predicting risk of partial discharge |
CN115128534A (en) * | 2022-06-28 | 2022-09-30 | 苏州纳芯微电子股份有限公司 | Detection method and detection system for current sensor cavity defect |
KR20230013423A (en) * | 2021-07-19 | 2023-01-26 | 효성중공업 주식회사 | Fire Preventive and Diagnostic System for Battery |
KR102547467B1 (en) * | 2022-12-23 | 2023-06-26 | 주식회사 솔라리치 | Distribution board with partial discharge diagnosis function using ultra high frequency signal |
US11846665B2 (en) | 2020-08-31 | 2023-12-19 | General Electric Company | Online and offline partial discharge detection for electrical drive systems |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101426792B1 (en) | 2014-01-03 | 2014-08-05 | (주) 동보파워텍 | UHF partial discharge and temperature monitoring system embedded switchgear system |
-
2015
- 2015-04-20 KR KR1020150055313A patent/KR101553005B1/en active IP Right Grant
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101426792B1 (en) | 2014-01-03 | 2014-08-05 | (주) 동보파워텍 | UHF partial discharge and temperature monitoring system embedded switchgear system |
Cited By (38)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20170088736A (en) * | 2016-01-25 | 2017-08-02 | 한국전자통신연구원 | Electricity-safety managing server, apparatus and method for managing of electricity-safety |
KR102433819B1 (en) * | 2016-01-25 | 2022-08-19 | 한국전자통신연구원 | Electricity-safety managing server, apparatus and method for managing of electricity-safety |
KR101641515B1 (en) | 2016-04-05 | 2016-07-22 | (주)서전기전 | The distribution panel including a diagnosis apparatus using HFCT |
KR20180042671A (en) | 2016-10-18 | 2018-04-26 | 현대일렉트릭앤에너지시스템(주) | Fault diagnosis apparatus using a dissolved gas generation pattern of the transformer |
KR101715079B1 (en) * | 2016-11-11 | 2017-03-09 | 주식회사 현대콘트롤전기 | A partial discharge detection and diagnosis appratus for a distributing board with the acoustic emission sensor |
CN106772007A (en) * | 2016-12-19 | 2017-05-31 | 天津市品通电力科技有限公司 | A kind of device for high-voltage switch gear live detection |
US10444284B2 (en) | 2016-12-22 | 2019-10-15 | Lsis Co., Ltd. | Diagnostic apparatus for switchgear |
KR101925338B1 (en) * | 2016-12-22 | 2018-12-05 | 엘에스산전 주식회사 | Diagnostic apparatus for switchgear |
KR101814147B1 (en) * | 2017-02-28 | 2018-01-30 | 골드텍 주식회사 | Distributing board with sensing function for partial discharge |
KR20190020233A (en) | 2017-08-17 | 2019-02-28 | 한국전력공사 | Simulated signal generating Apparatus and method for partial discharge simulation |
KR101982623B1 (en) | 2018-02-28 | 2019-06-28 | (주)인피니어 | Partial discharge measuring device and method of high voltage cable joint |
KR102058230B1 (en) * | 2018-08-27 | 2019-12-20 | 연세대학교 산학협력단 | Device, method and system to predict fault of cable |
CN108732478A (en) * | 2018-09-10 | 2018-11-02 | 清华四川能源互联网研究院 | Partial discharge of switchgear on-line monitoring system based on superfrequency punching current sensor and method |
KR102079874B1 (en) | 2018-10-05 | 2020-04-13 | 한국전력공사 | Apparatus and method for diagnosing partial discharge |
KR20200119431A (en) * | 2019-04-02 | 2020-10-20 | 현대일렉트릭앤에너지시스템(주) | Apparatus for diagnosis of partial discharge |
KR102183462B1 (en) * | 2019-04-02 | 2020-11-26 | 현대일렉트릭앤에너지시스템(주) | Apparatus for diagnosis of partial discharge |
CN110108993B (en) * | 2019-05-31 | 2021-05-07 | 国网青海省电力公司检修公司 | Multi-frequency-band three-dimensional combined test method for partial discharge of power equipment |
CN110108993A (en) * | 2019-05-31 | 2019-08-09 | 国网青海省电力公司检修公司 | A kind of multiband three-dimensional joint test method for power equipment shelf depreciation |
CN110531228A (en) * | 2019-09-03 | 2019-12-03 | 国网湖南省电力有限公司 | GIS partial discharge extent of disease severity method, system and medium based on principal component dimensionality reduction and clustering |
US11846665B2 (en) | 2020-08-31 | 2023-12-19 | General Electric Company | Online and offline partial discharge detection for electrical drive systems |
KR102413919B1 (en) * | 2020-12-21 | 2022-06-29 | (주)에이피엠테크놀러지스 | Partial discharge monitoring system and patial discharge monitoring method |
KR20220089422A (en) * | 2020-12-21 | 2022-06-28 | (주)에이피엠테크놀러지스 | Partial discharge monitoring system and patial discharge monitoring method |
KR20220096957A (en) * | 2020-12-31 | 2022-07-07 | 주식회사 큐아이티 | Method and apparatus for detecting partial discharge in arc-detecting system in power system |
KR102600662B1 (en) * | 2020-12-31 | 2023-11-10 | 주식회사 큐아이티 | Method for detecting partial discharge in arc-detecting system and a partial discharge detecting device for preveting arc in power system |
CN112924821A (en) * | 2021-01-25 | 2021-06-08 | 广东电网有限责任公司广州供电局 | Composite detection system and method for electric power equipment discharge and heating defect detection |
KR20220126930A (en) * | 2021-03-10 | 2022-09-19 | 한전케이디엔주식회사 | Apparatus and method for predicting risk of partial discharge |
KR102513271B1 (en) | 2021-03-10 | 2023-03-23 | 한전케이디엔주식회사 | Apparatus and method for predicting risk of partial discharge |
KR102295214B1 (en) * | 2021-05-31 | 2021-08-31 | 지투파워(주) | Partial discharging detection-diagnosis system of distribution panel using high frequency current transformer |
KR102586368B1 (en) * | 2021-07-19 | 2023-10-10 | 효성중공업 주식회사 | Fire Preventive and Diagnostic System for Battery |
KR20230013423A (en) * | 2021-07-19 | 2023-01-26 | 효성중공업 주식회사 | Fire Preventive and Diagnostic System for Battery |
CN113848435A (en) * | 2021-08-31 | 2021-12-28 | 四川大学 | Direct current partial discharge signal classification and identification method based on frequency spectrum energy distribution probability |
CN113848435B (en) * | 2021-08-31 | 2023-12-15 | 四川大学 | Direct-current partial discharge signal classification and identification method based on frequency spectrum energy distribution probability |
CN114113943A (en) * | 2021-11-25 | 2022-03-01 | 广东电网有限责任公司广州供电局 | Transformer partial discharge detection system, method and equipment based on current and ultrasonic signals |
KR102377936B1 (en) * | 2021-12-24 | 2022-03-23 | 지투파워(주) | Partial discharge monitoring and diagnosis system for intelligent distribution board using ultra frequency signal |
KR102377939B1 (en) * | 2021-12-24 | 2022-03-23 | 지투파워(주) | Partial discharge monitoring and diagnosis system for distribution board using ultra frequency and high frequency current transformer signal |
CN115128534B (en) * | 2022-06-28 | 2023-12-05 | 苏州纳芯微电子股份有限公司 | Detection method and detection system for current sensor cavity defect |
CN115128534A (en) * | 2022-06-28 | 2022-09-30 | 苏州纳芯微电子股份有限公司 | Detection method and detection system for current sensor cavity defect |
KR102547467B1 (en) * | 2022-12-23 | 2023-06-26 | 주식회사 솔라리치 | Distribution board with partial discharge diagnosis function using ultra high frequency signal |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR101553005B1 (en) | A partial discharge monitoring and diagnosis system for power devices | |
KR101574615B1 (en) | A partial discharge monitoring and diagnosis system for power devices by using signal detection based on statistical phase-angle patterns | |
KR101574613B1 (en) | A detection and diagnosis system with remote configuration function for partial discharge by detecting UHF electrical signal | |
Li et al. | Condition monitoring and diagnosis of power equipment: review and prospective | |
Montanari et al. | Partial discharge diagnostics: from apparatus monitoring to smart grid assessment | |
Zeng et al. | Feature extraction and severity assessment of partial discharge under protrusion defect based on fuzzy comprehensive evaluation | |
KR102192609B1 (en) | High voltage distributing board, low voltage distributing board, distributing board, motor control board having monitoring watching function using induced voltage sensor | |
JP2015078882A (en) | Insulation diagnostic device | |
Wang et al. | Reflectometry-based cable insulation aging diagnosis and prognosis | |
CN108051364A (en) | A kind of EPR nuclear energy cable residue lifetime estimation method and prediction EPR nuclear energy cable remaining life methods | |
CN110647924B (en) | GIS equipment state evaluation method based on support vector description and K-nearest neighbor algorithm | |
CN108548999B (en) | Cable insulation state evaluation method | |
KR101550689B1 (en) | A arc or corona detection system for a distributing board with the acoustic emission sensor and noise removal function | |
Zhao et al. | Experimental evaluation of detecting power transformer internal faults using FRA polar plot and texture analysis | |
Wang et al. | Application of extension neural network algorithm and chaos synchronization detection method to partial discharge diagnosis of power capacitor | |
Xu et al. | Integrated decision‐making method for power transformer fault diagnosis via rough set and DS evidence theories | |
Zhang et al. | Detection of the ratio error drift in CVT considering AVC | |
Khalyasmaa et al. | Fuzzy neural networks' application for substation integral state assessment | |
Taki et al. | Soft fault diagnosis in wiring networks using reflectometry and Principal Component Analysis | |
Hussain et al. | Review on Partial Discharge Diagnostic Techniques for High Voltage Equipment in Power Systems | |
CN109375037A (en) | Delay unloading a ship in one kind oceangoing ship Single-phase Earth Fault of Power System alarm designs method | |
Chang et al. | Fuzzy theory-based partial discharge technique for operating state diagnosis of high-voltage motor | |
Montanari | Insulation diagnosis of high voltage apparatus by partial discharge investigation | |
Sahoo et al. | A review on condition monitoring of GIS | |
Cselko et al. | Challenges of partial discharge diagnostics of low-voltage cables |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
E701 | Decision to grant or registration of patent right | ||
GRNT | Written decision to grant | ||
FPAY | Annual fee payment |
Payment date: 20180906 Year of fee payment: 4 |
|
FPAY | Annual fee payment |
Payment date: 20191105 Year of fee payment: 5 |