CN114707550A - Method for identifying direct current series arc fault at inverter end of photovoltaic system - Google Patents
Method for identifying direct current series arc fault at inverter end of photovoltaic system Download PDFInfo
- Publication number
- CN114707550A CN114707550A CN202210352668.7A CN202210352668A CN114707550A CN 114707550 A CN114707550 A CN 114707550A CN 202210352668 A CN202210352668 A CN 202210352668A CN 114707550 A CN114707550 A CN 114707550A
- Authority
- CN
- China
- Prior art keywords
- arc fault
- signal
- formula
- calculating
- current
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000001228 spectrum Methods 0.000 claims description 31
- BTCSSZJGUNDROE-UHFFFAOYSA-N gamma-aminobutyric acid Chemical compound NCCCC(O)=O BTCSSZJGUNDROE-UHFFFAOYSA-N 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000001514 detection method Methods 0.000 abstract description 10
- 238000010219 correlation analysis Methods 0.000 abstract description 3
- 239000000284 extract Substances 0.000 abstract 1
- 238000013016 damping Methods 0.000 description 13
- 238000004458 analytical method Methods 0.000 description 8
- 230000001052 transient effect Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 238000010891 electric arc Methods 0.000 description 5
- 230000010355 oscillation Effects 0.000 description 5
- 230000003595 spectral effect Effects 0.000 description 4
- 238000005070 sampling Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000010248 power generation Methods 0.000 description 2
- 239000003990 capacitor Substances 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000004451 qualitative analysis Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 230000035939 shock Effects 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)
Abstract
The invention relates to a method for identifying a direct current series arc fault at an inverter end of a photovoltaic system, and belongs to the technical field of direct current series arc fault detection of the photovoltaic system. The technical scheme of the invention extracts the arc fault current spectrogram characteristics by using the dimensionless data distribution statistics, and the dimensionless data distribution statistics can better reflect the spectrogram characteristics, unify the numerical range and better adapt to the starting working condition of the inverter. The grey correlation analysis has a strong ability to process random information. The method finds out the relevance of various characteristics and indexes to be analyzed and researched in different types of information through certain data processing, and can realize accurate and effective identification of the direct-current series arc fault under the starting working condition of the photovoltaic system inverter.
Description
Technical Field
The invention relates to a method for identifying a direct current series arc fault at an inverter end of a photovoltaic system, and belongs to the technical field of direct current series arc fault detection of the photovoltaic system.
Background
The traditional mode of power generation, of the extensive type, has posed a great number of environmental problems, with an ever-increasing population and an ever-expanding energy demand, going into the 21 st century. While the photovoltaic power generation technology benefits mankind, many faults, especially arc faults, seriously affect the safe operation of a photovoltaic power station. The existing protection methods mainly include a time domain protection method, a frequency domain protection method, a time-frequency domain protection method, and the like. The time domain protection method detects faults through signal waveform characteristics, and the technical implementation method is simple but has poor accuracy. And further excavating more arc fault characteristics for fault detection by the time domain and time-frequency domain protection method. However, in the existing arc fault research, attention is often paid to the comparison of fault characteristics and steady-state characteristics, and the influence of characteristics generated by current and voltage changes on the detection and protection of the direct-current arc fault under the normal working condition of the photovoltaic system is ignored. In actual environment, because the inverter internally comprises a nonlinear power electronic device, when the inverter works, particularly in the starting process, the sudden change characteristics generated by nonlinear fluctuation in the current rising process are easily confused with the current sudden change when an electric arc occurs, the existing time domain, frequency domain and time-frequency domain protection method is easily interfered, and the problem of misoperation of a direct current series arc fault detection device under the starting working condition of the inverter is not completely solved in the prior art. Therefore, the direct-current series arc fault is correctly identified under the starting working condition of the photovoltaic system inverter, the misoperation and the operation rejection probability are reduced, and the method has important significance for safe operation of the photovoltaic system.
Disclosure of Invention
The invention aims to solve the technical problem of misoperation of an arc fault detection method caused by current disturbance during the starting of an inverter, and provides a direct current series arc fault identification method which can be used for the inverter end close to a photovoltaic system, so that accurate and effective direct current series arc fault identification can be carried out under the starting working condition of the inverter.
In order to solve the problems, the technical scheme of the invention is as follows: provides a method for identifying a direct current series arc fault at an inverter end of a photovoltaic system, which sequentially comprises the following steps,
step 1: extracting arc fault current spectrogram characteristics by using dimensionless data distribution statistics mainly comprising kurtosis, skewness, peak factors, impact factors, margin factors and wave factors;
step 2: the arc fault data is classified using a grey correlation.
Further, step 1 comprises the following substeps:
step 1.1: calculating kurtosis ku of a signal spectrum1(x) The calculation formula is
In the formula, E { x4And E { x }2Are the fourth and second moments of the signal x;
step 1.2: calculating skewness skew (x) of signal spectrum, wherein the calculation formula is
Where μ is the expectation of signal x and σ is the standard deviation;
step 1.3: calculating the peak value factor C (x) of the signal spectrum by the formula
Wherein RMS (x) is the signal effective value;
step 1.4: calculating the impact factor I (x) of the signal frequency spectrum by the formula
Wherein the rectified mean value ARV (x) is the mean value of the absolute values of the signals and is calculated by
Step 1.5: calculating a margin factor M (x) of a signal spectrum by the formula
Where R (x) is the square root amplitude of the signal, is the square of the mean of the arithmetic square root, and is calculated as
Step 1.6: calculating the form factor F (x) of the signal spectrum by the formula
Further, step 2 comprises the following substeps:
step 2.1: carrying out dimensionless processing on the signal X, and recording the dimensionless quantity as XiFor xiAny one element x in (1)i(k) The computational expression of the dimensionless processing is
In the formula xi(1) Is the first data of the signal;
step 2.2: calculating the grey correlation coefficient by the formula
Wherein rho is a resolution coefficient and takes a value between 0 and 1,
referred to as the two-stage minimum difference,
referred to as the two-level maximum difference;
step 2.3: calculating the grey correlation degree according to the grey correlation coefficient, wherein the calculation formula is
In the formula of omegakIs the weight coefficient of the k index;
step 2.4: respectively calculating gray correlation degree p of the low frequency band and the high frequency band of the arc fault current data according to dimensionless data distribution statistics in the low frequency band of 1-20kHz and the high frequency band of 40-60kHz1；
Step 2.5: respectively calculating gray correlation degrees p of the low frequency band and the high frequency band of current data when the inverter is started according to dimensionless data distribution statistics in the low frequency band of 1-20kHz and the high frequency band of 40-60kHz2；
Step 2.6: when p is1≥p2Time of day is identified as an arc fault, p1＜p2An arc-free fault is identified.
The technical scheme of the invention utilizes the dimensionless data distribution statistics to extract the arc fault current spectrogram characteristics, the dimensionless data distribution statistics can better reflect certain characteristics of the spectrogram, and simultaneously, the dimensionless data distribution statistics unifies the numerical range and can better adapt to the starting working condition of the inverter, so the method is used for extracting the arc fault spectrogram characteristics. The grey correlation analysis has a strong ability to process random information. The method finds out the relevance of various characteristics and indexes to be analyzed and researched in different types of information through certain data processing, and determines the type of data to be tested according to main contradiction, main characteristics and main influence factors. Since the gray correlation analysis can reflect the overall trend of the characteristics without knowing typical rules, the requirement on the size of the sample volume is not high, and the final analysis result generally coincides with the qualitative analysis, so that the method has wide practicability.
Drawings
FIG. 1 is an RLC equivalent oscillation model of DC series arc fault
FIG. 2 is a schematic diagram of over-damped, critical damped, and under-damped oscillation waveforms
FIG. 3 is a schematic diagram of frequency spectra of over-damped, critically damped, and under-damped oscillation waveforms
FIG. 4 is a schematic circuit diagram of a DC series arc fault detection device
FIG. 5 shows a fault current and AD converted voltage signal according to a first embodiment of the present invention
FIG. 6 is a frequency spectrum of a fault current data window according to a first embodiment of the present invention
FIG. 7 is a diagram of a fault current and AD converted voltage signal according to a second embodiment of the present invention
FIG. 8 is a frequency spectrum of a fault current data window according to a second embodiment of the present invention
FIG. 9 shows a fault current and an AD converted voltage signal according to a third embodiment of the present invention
FIG. 10 is a frequency spectrum of a fault current data window in a third embodiment of the present invention
FIG. 11 is a diagram of a fault current and an AD converted voltage signal according to a fourth embodiment of the present invention
FIG. 12 is a frequency spectrum of a fault current data window in a fourth embodiment of the present invention
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic drawings that illustrate, by way of illustration only, the basic structure of the invention and, therefore, show only the components that are relevant to the invention.
The arc may be equivalent to an RLC oscillator circuit as shown in fig. 1. When arcing begins, a localized, isolated capacitively charged spot begins to appear between the two contact points. Under the condition of sufficiently high charge density, a strong electric field close to the edge of a light spot can cause electric breakdown, arc discharge with the length equal to that of current is generated, and capacitance C appears at two ends of a fault breakpointsLocal isolation point of (1), arc self-inductance LarcAnd arc resistance RarcAre connected in series.
Suppose self-inductance LarcAnd a resistance RarcApproximately time independent over the life of the arc, the arc can be equated with Larc、RarcAnd a capacitor CsThe second-order oscillation equivalent model is formed by the calculation formula of a differential equation
Wherein α ═ Rarc/2LarcIs the rate of the inductive damping,
is the natural frequency of the circuit, and alpha and omega are all non-negative numbers;
the differential equation corresponds to a characteristic equation of p2+2αp+ω2＝0；
For an over-damped transition state, the circuit i (t) general solution in the circuit can be expressed as
In the formula
Is the initial ramp rate of the arc current;
from the relationship between α and ω, it can be seen that when α > ω, i.e., α > ω
p has two different real numbers, and the current is an over-damping signal; when α ═ ω, p has two real roots which are equal, in this caseThe current is a critical damping signal; when alpha is less than omega, p has no real number, and the current is an underdamped signal;
FIG. 2 shows that the frequency spectrum voltage is converted into dimensionless normalized amplitude value through normalization processing, the bandwidths of over-damping, critical damping and under-damping are sequentially reduced, the arc current belongs to an over-damping signal, and three damped oscillation waveforms of over-damping, critical damping and under-damping and corresponding frequency spectrums are schematically shown in FIG. 3, wherein fsRepresenting the sampling frequency, where the over-damped resistance is 4 times the critical damping and the under-damping is 1/4 times the critical damping, so the arc current signal has a wider frequency band in the frequency domain.
The principle of the dc series arc fault detection device is shown in fig. 4, and mainly includes a current transformer, an impedance matching circuit, a signal amplifying and filtering circuit, a digital-to-analog conversion AD unit, and a controller. The AD uses SM73201 of TI company, and has double-end differential mode input, the double-end differential mode input has stronger noise suppression capability compared with single-end input, the maximum sampling frequency is 250kHz, and the sampling frequency of the oscilloscope is 250kHz as the characteristic frequency of the arc fault current can reach more than 100 kHz.
To facilitate understanding of the public, the technical solution of the present invention is further described in detail by the following specific examples.
The first embodiment is as follows:
when an arc fault occurs, a fault current IarcWhen transient change occurs, as shown by the dotted line in fig. 5, the operating voltage of the photovoltaic system is 600V, the arcing current is 14A, and data used by the arc fault detection device needs to be converted by the current transformer and the AD and then output as a voltage signal UarcThe converted voltage signal is shown in fig. 5(b), wherein a region I represents a signal waveform when the photovoltaic system normally operates, a dashed region II represents a signal waveform when an arc fault occurs, and a region III represents a signal waveform after the system is completely powered off. It can be seen from the figure that the arc fault signal represented within the red dashed line of region II has a transient abrupt change characteristic which may be used to identify an arc fault.
In order to facilitate analysis of the change condition of the fault current, 6 data windows in the region I and the region II before and after the fault are selected for Fourier analysis, 3 data windows selected in the region I are sequentially named as A1, A2 and A3 according to the time sequence, and 3 data windows selected in the region II are respectively named as A4, A5 and A6. Comparison of spectral amplitudes of a1-a6 is shown in fig. 6, the spectrograms of three groups of steady operating currents a1, a2 and A3 are (a), (b) and (c) in the graph, respectively, and the spectrums of arc fault data windows a4, a5 and a6 are shown as (d), (e) and (f) in the graph. By comparing the stable operation and the fault voltage signal frequency spectrogram, the frequency spectrum shapes and amplitudes of the A4, the A5 and the A6 are different from those of the A1, the A2 and the A3, and the amplitudes of the frequency spectrum shapes and amplitudes are different from those of the low frequency band of 1-20kHz and the high frequency band of 40-60 kHz.
The second embodiment:
the arc fault test working condition of the electric arc fault test of the graph 7 is different from that of the former group, the working voltage of a photovoltaic system is 600V, the arc discharge current is 10A, and the more important difference is that the frequency spectrum of the electric arc fault test working condition only contains fault characteristics between 1 kHz and 20 kHz. As can be seen from fig. 7(a), regions I and II also represent the photovoltaic system normal state and the arc fault state, respectively, and region III represents the system power-off state. It can be seen that the arc fault characteristics are substantially the same under different operating conditions, but the smaller the arc fault current, the larger the current transient amplitude, as shown in fig. 7(b), since the arc energy is small and the arc burning becomes unstable.
Three sets of data windows before and after the fault are selected from the regions I and II for analysis, wherein stable working data windows before the fault are named A7, A8 and A9 respectively, arc fault data windows are named A10, A11 and A12 respectively, the transient change amplitude of the A10 waveform is maximum, and the arc burning is seen to be gradually stabilized due to the transient change from A10 to A12. The frequency spectrum corresponding to the data window is shown in fig. 8, wherein the frequency spectrum change amplitude of a7-a9 representing the steady-state operating current is small, the characteristic frequency band of a10-a12 representing the arc fault is concentrated in the low frequency band of 1-20kHz, and particularly, the frequency spectrum of a10 is relatively remarkably improved and presents a broadband characteristic.
Example three:
fig. 9 is a waveform of current change at the time of starting operation of the inverter, and as can be seen from fig. 9(a), as time advances, the current first rises rapidly and then tends to be smooth. The working condition can be divided into 9 stages, the current has 3 jump processes in the areas II, IV and VI, the corresponding voltage signal in the graph 9(b) also has obvious transient shock sudden change, the jump can be mistaken for arc fault characteristics, the correct identification of the arc detection device to the fault is influenced, and false operation is caused.
From each of the four II, IV, VI and VIII regions 1 data window was selected for analysis, named M1, M2, M3 and M4, respectively, where the data window M4 was selected from the stationary current region VIII for comparison analysis with the transient current, and the total window length was 1024. FIG. 10 is a graph of spectra corresponding to four portions of the data window M1-M4, where the M1 spectral amplitude has a boost in the 0-60kHz band and the spectral shape has narrow band characteristics. The spectrum amplitude of M2 and M3 is increased in the range of 0-20kHz, and the shape of the part is close to that of an arc fault, so that the part is easy to interfere with the identification of the arc fault and also causes misoperation. The M4 part is in a stable phase with small current changes and therefore low spectral amplitude.
Example four:
fig. 11 is a current waveform under the same inverter starting condition, and is different from fig. 8 in that the current is gradually increased in the region II after the current starts to be gradually increased in the region I, and there is no multi-jump process until the current in the region III is stabilized after 10 seconds, so that the analysis can be divided into three typical regions.
One data window, named M5, M6, and M7, each 1024 in length, is selected from regions I, II and III, respectively. FIG. 12 is the frequency spectrum corresponding to M5-M7, and it can be seen from the figure that the frequency spectrum of M5 has a larger amplitude in the 20kHz band, while the frequency spectrum of M6 and M7 has a slightly increased amplitude in the 20-40kHz band. From spectrogram shape analysis, M5-M7 did not exhibit significant broadband characteristics.
Claims (9)
1. A method for identifying a direct current series arc fault at an inverter end of a photovoltaic system is characterized by extracting arc fault current spectrogram characteristics by using dimensionless data distribution statistics mainly comprising kurtosis, skewness, peak factors, impact factors, margin factors and wave form factors.
2. Calculating kurtosis ku of a signal spectrum1(x) The calculation formula is
In the formula, E { x4And E { x }2The fourth moment and the second moment of the signal x are calculated, and the skewness skew (x) of the signal frequency spectrum is calculated according to the formula
Where μ is the expectation of the signal x and σ is the standard deviation.
3. Calculating the peak value factor C (x) of the signal spectrum by the formula
Wherein RMS (x) is the effective value of the signal, calculating the impact factor I (x) of the signal spectrum
Wherein the rectified mean value ARV (x) is the mean value of the absolute values of the signals and is calculated by
4. Calculating a margin factor M (x) of a signal spectrum by the formula
Where R (x) is the square root amplitude of the signal, is the square of the mean of the arithmetic square root, and is calculated as
Calculating the form factor F (x) of the signal spectrum by the formula
5. The photovoltaic system inverter-side direct current series arc fault identification method of claim 1, wherein arc fault data is classified using a grey correlation.
6. Carrying out dimensionless processing on the signal X, and recording the dimensionless quantity as XiFor xiAny one element x in (1)i(k) The computational expression of the dimensionless processing is
In the formula xi(1) Is the first data of the signal.
7. Calculating the grey correlation coefficient by the formula
Wherein rho is a resolution coefficient and takes a value between 0 and 1,
referred to as the two-stage minimum difference,
called two-stage maximum difference, the grey correlation degree is calculated according to the grey correlation coefficient, and the calculation formula is
In the formula of omegakIs the weighting factor of the k index.
8. Respectively calculating gray correlation degree p of the low frequency band and the high frequency band of the arc fault current data according to dimensionless data distribution statistics in the low frequency band of 1-20kHz and the high frequency band of 40-60kHz1According to the invaluity in the low frequency band of 1-20kHz and the high frequency band of 40-60kHzCalculating gray correlation degree p between the current data and the current data at the start of the inverter according to the dimension data distribution statistics2。
9. When p is1≥p2Time of day is identified as an arc fault, p1＜p2An arc-free fault is identified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210352668.7A CN114707550A (en) | 2022-04-06 | 2022-04-06 | Method for identifying direct current series arc fault at inverter end of photovoltaic system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210352668.7A CN114707550A (en) | 2022-04-06 | 2022-04-06 | Method for identifying direct current series arc fault at inverter end of photovoltaic system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114707550A true CN114707550A (en) | 2022-07-05 |
Family
ID=82172449
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210352668.7A Pending CN114707550A (en) | 2022-04-06 | 2022-04-06 | Method for identifying direct current series arc fault at inverter end of photovoltaic system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114707550A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110501631A (en) * | 2019-08-19 | 2019-11-26 | 重庆大学 | A kind of online intermittent fault detection and diagnostic method |
| CN113552447A (en) * | 2021-07-27 | 2021-10-26 | 上海电机学院 | Series arc fault detection method based on random forest |
-
2022
- 2022-04-06 CN CN202210352668.7A patent/CN114707550A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110501631A (en) * | 2019-08-19 | 2019-11-26 | 重庆大学 | A kind of online intermittent fault detection and diagnostic method |
| CN113552447A (en) * | 2021-07-27 | 2021-10-26 | 上海电机学院 | Series arc fault detection method based on random forest |
Non-Patent Citations (1)
| Title |
|---|
| 陈进: "低压交流故障电弧的特征提取与识别", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》, 15 March 2022 (2022-03-15), pages 3 - 4 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8680873B2 (en) | Method and apparatus for fault identification in a power tansmission line | |
| Wang et al. | Classification of power quality events using optimal time-frequency representations-part 1: Theory | |
| Song et al. | A negative selection algorithm-based identification framework for distribution network faults with high resistance | |
| CN112730964A (en) | Lightning overvoltage identification method based on overvoltage waveform characteristics | |
| CN113191192A (en) | Breaker fault detection method based on wavelet analysis and fuzzy neural network algorithm | |
| CN109001602A (en) | Shelf depreciation Severity method based on extreme learning machine algorithm | |
| Wang et al. | Stator single-line-to-ground fault protection for bus-connected powerformers based on S-transform and bagging ensemble learning | |
| CN114707550A (en) | Method for identifying direct current series arc fault at inverter end of photovoltaic system | |
| Bonde et al. | Review on detection and classification of underlying causes of power quality disturbances using signal processing and soft computing technique | |
| CN112255495B (en) | Micro-grid high-resistance fault detection method | |
| CN111896890B (en) | Micro-grid line fault diagnosis method and system based on Hilbert-Huang transform | |
| Yao et al. | Pattern recognition for partial discharge using multi-feature combination adaptive boost classification model | |
| CN114113894A (en) | Repetitive fault identification method based on natural frequency characteristics | |
| Gao et al. | Quantitative research on accumulative effect of transformer winding deformation and its influence degree based on time-frequency analysis of vibration signal | |
| Gao et al. | Internal overvoltage identification of distribution network via time-frequency atomic decomposition | |
| CN106771850A (en) | Distribution network fault line selection method based on cascade bistable system | |
| Do et al. | Wavelet packet-based passive islanding detection method for grid connected photovoltaic inverters | |
| Wang et al. | Faulty line detection for cross‐line same‐phase successive ground faults in distribution network based on transient characteristics | |
| Wu et al. | A Novel Method for Extracting Features of Partial Discharge Pulse Current Signal | |
| CN115047240B (en) | Transformer excitation surge current discriminating method by utilizing wavelet detail component change characteristics | |
| CN113189446B (en) | Low-current ground fault line selection method | |
| CN113311219B (en) | Power distribution network temporary overvoltage identification method | |
| Shuai et al. | Pattern recognition of partial discharge ultrahigh frequency signal based on similar matrix BSS and deep learning CNN | |
| CN116500383B (en) | Method and system for identifying high-resistance faults and switching disturbance based on Lissajous curves | |
| CN114062833B (en) | Single-phase earth fault insulation degradation identification method based on fuzzy inference system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |