CN114720818A - Alternating current series fault arc detection method based on time-frequency feature screening - Google Patents
Alternating current series fault arc detection method based on time-frequency feature screening Download PDFInfo
- Publication number
- CN114720818A CN114720818A CN202210338968.XA CN202210338968A CN114720818A CN 114720818 A CN114720818 A CN 114720818A CN 202210338968 A CN202210338968 A CN 202210338968A CN 114720818 A CN114720818 A CN 114720818A
- Authority
- CN
- China
- Prior art keywords
- current
- frequency
- fault arc
- time
- series fault
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 25
- 238000012216 screening Methods 0.000 title claims abstract description 23
- 210000002569 neuron Anatomy 0.000 claims abstract description 12
- 238000002474 experimental method Methods 0.000 claims abstract description 9
- 238000001228 spectrum Methods 0.000 claims description 23
- 238000012549 training Methods 0.000 claims description 17
- 238000005070 sampling Methods 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 6
- 230000003595 spectral effect Effects 0.000 claims description 6
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 108010076504 Protein Sorting Signals Proteins 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- -1 hydrogen Chemical class 0.000 claims 1
- 239000001257 hydrogen Substances 0.000 claims 1
- 230000002265 prevention Effects 0.000 claims 1
- 238000000034 method Methods 0.000 description 5
- 230000017105 transposition Effects 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 4
- 241000282414 Homo sapiens Species 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000010891 electric arc Methods 0.000 description 2
- 238000012356 Product development Methods 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
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
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/081—Locating faults in cables, transmission lines, or networks according to type of conductors
- G01R31/086—Locating faults in cables, transmission lines, or networks according to type of conductors in power transmission or distribution networks, i.e. with interconnected conductors
-
- 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
- G01R31/08—Locating faults in cables, transmission lines, or networks
- G01R31/088—Aspects of digital computing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/088—Non-supervised learning, e.g. competitive learning
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)
Abstract
The invention belongs to the field of line arc fault detection, and particularly discloses an alternating current series fault arc detection method based on time-frequency feature screening, which comprises the following steps: s1: building an alternating current series fault arc experiment acquisition platform; s2: collecting normal current waveforms and current waveforms when arc faults occur in single-load and multi-load circuits, and constructing a series fault arc current database; s3: calculating time domain characteristics and frequency domain characteristics; s4: corresponding the time domain and frequency domain characteristics to the original data one by one to construct a characteristic database; s5: performing feature screening through an XGboost algorithm to screen out important features; s6: setting the number of MLP layers and the number of neurons; s7: selecting an optimizer of the model and setting a learning rate; s8: regularization of the MLP by Dropout prevents overfitting of the model; s9: and storing the trained model.
Description
Technical Field
The invention relates to the field of line arc fault detection.
Background
With the development and scientific progress of society, the quality of life of human beings gradually rises, and the demand for power resources is increasing. Nowadays, the use of low-cost power resources by human beings is gradually increased, and electric equipment is diversified, and brings convenience to the daily life of human beings. However, as the number of electrical devices increases, the line connection structure of the residential area becomes complicated and complicated, electric arcs may be generated due to line aging or loose connection, and high temperatures of 2000-. In order to improve the detection capability of the fault arc and enhance the power utilization safety of the alternating current power supply system, relevant standards are proposed at home and abroad. In 1999, the U.S. proposed a product development for an arc fault detector that is suitable for the UL1699 standard under 120V/50Hz AC operating power. In 2013, the international electrotechnical commission has set an IEC62606 international standard by researching the characteristics and diagnosis of fault arcs under a 240V power supply system. In 2014, the national standard GB14287.4 was established in China by combining international standards and national research status.
The position of the fault arc is unknown, the installation position of the sensor is difficult to determine, and the signal acquisition is easily influenced by factors such as illumination intensity, environmental noise and the like. It is difficult to adapt to the change of the load type by setting a threshold value by a voltage current signal or the like. The frequency spectrum of the current signal is obtained through FFT conversion, the frequency spectrum resolution is low, and a large window is needed if the resolution is to be improved. According to part of researches, original current waveforms are identified by combining a deep learning algorithm after being processed by some signal processing methods, so that the identification accuracy can be improved, but the method has many redundant characteristics, and needs to consume a large amount of calculation cost or bring much interference information. At present, most of researches aim at single-load fault arc detection, and the fault detection precision of multiple-load complex conditions is difficult to guarantee.
Disclosure of Invention
The invention aims to provide an MLP-based alternating current series fault arc detection method which improves the frequency spectrum resolution of an arc current signal based on CZT conversion, avoids redundant features by combining with an XGboost algorithm for feature screening and is self-adaptive to learning.
In order to solve the above problems and achieve the above object, the present invention provides the following technical solutions:
1. an alternating current series fault arc detection method based on time-frequency feature screening is characterized by comprising the following steps:
s1: building an alternating current series fault arc experiment acquisition platform;
s2: collecting normal current waveforms and current waveforms when arc faults occur in single-load and multi-load circuits, and constructing a series fault arc current database;
s3: calculating time domain and frequency domain characteristics;
s4: constructing a feature database by using the calculated time domain and frequency domain features;
s5: performing feature screening through an XGboost algorithm to screen out important features;
s6: setting the number of MLP layers and the number of neurons;
s7: selecting an optimizer of the model and setting a learning rate;
s8: regularization of the MLP by Dropout prevents overfitting of the model;
s9: and storing the trained model.
Further, the step S1 specifically includes: the method comprises the steps that an alternating current series fault arc experiment acquisition platform is built, and the alternating current series fault arc experiment acquisition platform comprises an arc generation transposition, a sensor, an oscilloscope, a switch, a sampling resistor and different types of loads; wherein, the arc generating device is used for simulating fault arc in the line; the sensor is used for sensing the change of current in the line; the oscilloscope is used for displaying and storing current data; the switch is used for controlling the access of the load; the sampling resistor is 100 omega; the different types of loads are used for simulating line faults of different types of loads and multiple loads.
Further, the step S2 specifically includes: collecting current waveforms of normal and arc faults in a single-load and multi-load circuit, and constructing a series fault arc current database, wherein one or more loads are connected through a control switch, the arc is adjusted to generate a transposition to collect the current waveforms of the normal and arc faults of the single load or the multi-load, the collected current data of the normal and fault are used as a group of data according to a period length, and the series fault arc current waveform database is constructed;
further, the step S3 specifically includes: calculating time domain and frequency domain characteristics, wherein the time domain characteristics comprise a current fluctuation value, a current effective value, a current integral and a mean value of a window maximum difference value; the expression is as follows:
wherein Δ I represents a current fluctuation value; x (n) represents a current signal sequence of one cycle; max [ x (n)]，min[x(n)]Respectively representing the maximum value and the minimum value of the current; i isrmsRepresents the effective value of the current; x is the number ofiRepresents the magnitude of the current at time i; n represents the number of data points of the current in one period; i is∫Representing the result of the integration; i isWMDRepresents the mean of the maximum differences of the windows, and the current value of each sampling point is marked as xiThe sliding window size was chosen to be 20 data points.
Dividing the frequency spectrum of the frequency band of 0-4kHz into 4 equal parts through CZT, setting the frequency spectrum resolution as 1Hz, and then setting the frequency domain characteristics to include the mean value of the narrow-band frequency spectrum amplitude and the ratio of the frequency spectrum amplitude; the expression is as follows:
wherein s isiRepresents the sum of the spectral magnitudes of the ith segment,
represents the mean of the ith segment; n is a radical ofiRepresenting the number of i-th band spectral points. r isiRepresenting the ith section of spectrum ratio; m represents the total number of segments of the refined spectrum.
Further, the step S4 specifically includes: and constructing a feature database by using the calculated time domain and frequency domain features, wherein each group of data has 12 time domain and frequency domain features in total, the number of the normal data is 0, and the number of the fault data is 1.
Further, the step S5 specifically includes: and screening the characteristics through an XGboost algorithm to screen out important characteristics, wherein the learning rate in the XGboost algorithm is set to be 0.1, the depth is set to be 5, the number of trees is set to be 15, and the first 5 characteristics of the scores are selected as final characteristics through training.
Further, the step S6 is specifically: the number of MLP layers and the number of neurons are set, wherein the number of hidden layers is set to be 2, 16 neurons are set in the first layer, and 8 neurons are set in the second layer.
Further, the step S7 specifically includes: selecting an optimizer of the model and setting a learning rate, wherein the optimizer selects Adam and the learning rate is set to 0.01.
Further, the step S8 specifically includes: regularization of the MLP by Dropout prevents overfitting of the model, specifically, the data set is divided into a training set and a test set with a ratio of 7:3, and the training set is input to the 2 nd hidden layer in the MLP for Dropout.
Further, the step S9 specifically includes: and storing the trained model, wherein the training is stopped when the loss function and the correct rate curve are converged and the number of times of training reaches a set number of times or the loss function value is smaller than a set threshold value in the training process, the model is stored, and the stored model is used for actual arc detection.
The invention has the beneficial effects that:
(1) according to the invention, by introducing CZT conversion, the frequency spectrum resolution is improved, and the problems of low frequency spectrum resolution of FFT conversion and the like are solved;
(2) aiming at the extracted features, redundancy exists, the XGboost algorithm is utilized to screen out the first 5 important features, and the redundant features are prevented from bringing a large amount of calculation cost or interference;
(3) through MLP self-adaptation learning, the threshold value does not need to be set manually, and the identification rate of the electric arc can be improved.
Drawings
FIG. 1A is an AC series fault arc experiment acquisition platform
FIG. 2 is a flow chart of an AC series fault arc detection method based on time-frequency feature screening
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
The data acquisition experiment platform of the invention is shown in figure 1, and the equipment comprises an electric arc generation transposition, a sensor, an oscilloscope, a switch, a sampling resistor and different types of loads; wherein, the arc generating device is used for simulating fault arc in the line; the sensor is used for sensing the change of current in the line; the oscilloscope is used for displaying and storing current data; the switch is used for controlling the access of the load; the sampling resistor is 100 omega; the different types of loads are used for simulating line faults of different types of loads and multiple loads.
Fig. 2 is a flow chart of the alternating current series fault arc detection method based on time-frequency feature screening, and as shown in the figure, the method specifically includes the following steps:
1. an alternating current series fault arc experiment acquisition platform is built, and needed equipment comprises an arc generation transposition, a sensor, an oscilloscope, a switch, a sampling resistor and different types of loads.
2. Collecting current waveforms of normal and arc faults in a single-load and multi-load circuit, and constructing a series fault arc current database, wherein one or more loads are connected through a control switch, the arc is adjusted to generate a transposition to collect the current waveforms of the normal and arc faults of the single load or the multi-load, the collected current data of the normal and fault are used as a group of data according to a period length, and the series fault arc current waveform database is constructed;
3. calculating time domain and frequency domain characteristics, wherein the time domain characteristics: the average value of the current fluctuation value, the current effective value, the current integral and the window maximum difference value; the time domain feature expression is:
wherein Δ I represents a current fluctuation value; x (n) represents a current signal sequence of one cycle; max [ x (n)]，min[x(n)]Respectively representing the maximum value and the minimum value of the current; i isrmsRepresents the effective value of the current; x is the number ofiRepresents the magnitude of the current at time i; n represents the number of data points of the current in one period; i is∫Represents the result of the integration; i isWMDRepresents the mean of the maximum differences of the windows, and the current value of each sampling point is marked as xiThe sliding window size was chosen to be 20 data points.
Dividing the frequency spectrum of the 0-4kHz frequency band into 4 equal parts through CZT, setting the frequency spectrum resolution as 1Hz, and calculating the frequency domain characteristics: the average value of the narrow-band frequency spectrum amplitude and the ratio of the frequency spectrum amplitude. The frequency domain characteristic expression is:
wherein s isiRepresents the sum of the spectral magnitudes of the ith segment,
represents the mean of the ith segment; n is a radical ofiRepresenting the number of i-th segment spectral points. r isiRepresenting the ith section of spectrum ratio; m represents the total number of segments of the refined spectrum.
4. And constructing a feature database by using the calculated time domain and frequency domain features, wherein each group of data has 12 time domain and frequency domain features in total, the number of the normal data is 0, and the number of the fault data is 1.
5. And screening the characteristics through an XGboost algorithm to screen out important characteristics, wherein the learning rate of the XGboost algorithm is set to be 0.1, the depth is set to be 5, the number of trees is set to be 15, and the first 5 characteristics of the scores are selected as final characteristics through training.
6. The number of MLP layers and the number of neurons are set, wherein the number of hidden layers is set to be 2, 16 neurons are set in the first layer, and 8 neurons are set in the second layer.
7. Selecting an optimizer of the model and setting a learning rate, wherein the optimizer selects Adam and the learning rate is set to 0.01.
8. Regularizing the MLP by Dropot to prevent overfitting of the model, specifically, dividing a data set into a training set and a test set with a ratio of 7:3, and inputting the training set into a layer 2 hidden layer in the MLP for Dropot.
9. And storing the trained model, specifically, stopping training when the loss function and the correct rate curve are converged in the training process and the training times reach the set times or the loss function value is smaller than the set threshold, storing the model, and using the stored model for actual arc detection.
Claims (10)
1. An alternating current series fault arc detection method based on time-frequency feature screening is characterized by comprising the following steps:
s1: building an alternating current series fault arc experiment acquisition platform;
s2: collecting normal current waveforms and current waveforms when arc faults occur in single-load and multi-load circuits, and constructing a series fault arc current database;
s3: calculating time domain and frequency domain characteristics;
s4: constructing a feature database by using the calculated time domain and frequency domain features;
s5: performing feature screening through an XGboost algorithm to screen out important features;
s6: setting the number of MLP layers and the number of neurons;
s7: selecting an optimizer of the model and setting a learning rate;
s8: regularization of the MLP by Dropout prevents overfitting of the model;
s9: and storing the trained model.
2. The alternating-current series fault arc detection method based on time-frequency feature screening of claim 1, wherein in the step S1, an alternating-current series fault arc experiment acquisition platform is built, which includes an arc generation transpose, a sensor, an oscilloscope, a switch, a sampling resistor, and different types of loads; wherein, the arc generating device is used for simulating fault arc in the line; the sensor is used for sensing the change of current in the line; the oscilloscope is used for displaying and storing current data; the switch is used for controlling the access of the load; the sampling resistor is 100 omega; the different types of loads are used for simulating line faults of different types of single loads and multiple loads.
3. The alternating-current series fault arc detection method based on time-frequency feature screening as claimed in claim 1, wherein in the step S2, current waveforms of normal and arc faults occurring in single-load and multi-load circuits are collected to construct a series fault arc current database, wherein one or more loads are connected through a control switch, an arc generator is adjusted to collect current waveforms of normal and arc faults occurring in single-load or multi-load circuits, collected current data of normal and fault are grouped into a group of data according to a period length, and a series fault arc current waveform database is constructed.
4. The time-frequency feature screening-based alternating current series fault arc detection method according to claim 1, wherein time domain and frequency domain features are calculated in the step S3, wherein the time domain features: the average value of the current fluctuation value, the current effective value, the current integral and the window maximum difference value; the time domain characteristic expression is:
wherein Δ I represents a current fluctuation value; x (n) represents a current signal sequence of one cycle; max [ x (n)]，min[x(n)]Respectively representing the maximum value and the minimum value of the current; i isrmsRepresents the effective value of the current; x is the number ofiRepresents the magnitude of the current at time i; n represents the number of data points of the current in one period; i is∫Represents the result of the integration; I.C. AWMDRepresents the mean of the maximum differences of the windows, and the current value of each sampling point is marked as xiThe sliding window size was chosen to be 20 data points.
Dividing the frequency spectrum of the 0-4kHz frequency band into 4 equal parts through CZT, setting the frequency spectrum resolution as 1Hz, and calculating the frequency domain characteristics: the average value of the narrow-band frequency spectrum amplitude and the ratio of the frequency spectrum amplitude. The frequency domain characteristic expression is:
wherein s isiRepresents the sum of the spectral magnitudes of the ith segment,
represents the mean of the ith segment; n is a radical ofiRepresenting the number of i-th segment spectral points. r is a radical of hydrogeniRepresenting the ith section of spectrum ratio; m represents the total number of segments of the refined spectrum.
5. The time-frequency feature screening-based alternating current series fault arc detection method as claimed in claim 1, wherein in the step S4, a feature database is constructed from the calculated time domain and frequency domain features, wherein each group of data has a total of 12 time domain and frequency domain features, the normal data number is 0, and the fault data number is 1.
6. The alternating-current series fault arc detection method based on time-frequency feature screening as claimed in claim 1, wherein in step S5, feature screening is performed through an XGBoost algorithm to screen out important features, wherein the learning rate of the XGBoost algorithm is set to 0.1, the depth is set to 5, the number of trees is set to 15, and the first 5 features are selected as final features through training.
7. The alternating-current series fault arc detection method based on time-frequency feature screening of claim 1, wherein the number of MLPs and the number of neurons are set in step S6, specifically, the number of hidden layers is set to 2, the first layer is set to 16 neurons, and the second layer is set to 8 neurons.
8. The alternating-current series fault arc detection method based on time-frequency feature screening as claimed in claim 1, wherein an optimizer of the model is selected and a learning rate is set in step S7, wherein Adam is selected by the optimizer and the learning rate is set to 0.01.
9. The time-frequency feature screening-based alternating current series fault arc detection method as claimed in claim 1, wherein in the step S8, the MLP regularization prevention model is overfitted by Dropout, specifically, a data set is divided into a training set and a test set in a ratio of 7:3, and the training set is input to the 2 nd hidden layer in the MLP for Dropout.
10. The alternating-current series fault arc detection method based on time-frequency feature screening as claimed in claim 1, wherein in the step S9, the trained model is saved, specifically, when the loss function and the correctness curve converge and the training times reach the set times or the loss function value is smaller than the set threshold value, the training is stopped, the model is saved, and the saved model is used for actual arc detection.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210338968.XA CN114720818A (en) | 2022-04-01 | 2022-04-01 | Alternating current series fault arc detection method based on time-frequency feature screening |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210338968.XA CN114720818A (en) | 2022-04-01 | 2022-04-01 | Alternating current series fault arc detection method based on time-frequency feature screening |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114720818A true CN114720818A (en) | 2022-07-08 |
Family
ID=82241335
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210338968.XA Pending CN114720818A (en) | 2022-04-01 | 2022-04-01 | Alternating current series fault arc detection method based on time-frequency feature screening |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114720818A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116908632A (en) * | 2023-07-19 | 2023-10-20 | 常熟理工学院 | Low-voltage alternating-current series arc fault characteristic self-adaptive optimizing detection method and system |
| CN117454166A (en) * | 2023-10-11 | 2024-01-26 | 国网四川省电力公司电力科学研究院 | Method for identifying arc faults of ignition based on EffNet lightweight model |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190293701A1 (en) * | 2018-03-20 | 2019-09-26 | Whisker Labs, Inc. | Detection of electric discharges that precede fires in electrical wiring |
| CN111665819A (en) * | 2020-06-08 | 2020-09-15 | 杭州电子科技大学 | Deep learning multi-model fusion-based complex chemical process fault diagnosis method |
| CN112904157A (en) * | 2021-01-19 | 2021-06-04 | 重庆邮电大学 | Fault arc detection method based on integrated machine learning |
| CN112904156A (en) * | 2021-01-19 | 2021-06-04 | 重庆邮电大学 | Fault arc detection method based on frequency domain classification |
| CN113049922A (en) * | 2020-04-22 | 2021-06-29 | 青岛鼎信通讯股份有限公司 | Fault arc signal detection method adopting convolutional neural network |
| CN113378778A (en) * | 2021-06-30 | 2021-09-10 | 东南大学 | On-load tap-changer fault diagnosis method based on self-encoder |
| CN113820574A (en) * | 2021-09-29 | 2021-12-21 | 南方电网数字电网研究院有限公司 | SoC (system on chip) architecture and device for arc detection |
-
2022
- 2022-04-01 CN CN202210338968.XA patent/CN114720818A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190293701A1 (en) * | 2018-03-20 | 2019-09-26 | Whisker Labs, Inc. | Detection of electric discharges that precede fires in electrical wiring |
| CN113049922A (en) * | 2020-04-22 | 2021-06-29 | 青岛鼎信通讯股份有限公司 | Fault arc signal detection method adopting convolutional neural network |
| CN111665819A (en) * | 2020-06-08 | 2020-09-15 | 杭州电子科技大学 | Deep learning multi-model fusion-based complex chemical process fault diagnosis method |
| CN112904157A (en) * | 2021-01-19 | 2021-06-04 | 重庆邮电大学 | Fault arc detection method based on integrated machine learning |
| CN112904156A (en) * | 2021-01-19 | 2021-06-04 | 重庆邮电大学 | Fault arc detection method based on frequency domain classification |
| CN113378778A (en) * | 2021-06-30 | 2021-09-10 | 东南大学 | On-load tap-changer fault diagnosis method based on self-encoder |
| CN113820574A (en) * | 2021-09-29 | 2021-12-21 | 南方电网数字电网研究院有限公司 | SoC (system on chip) architecture and device for arc detection |
Non-Patent Citations (1)
| Title |
|---|
| 王毅 等: "基于Stacking模型融合的串联故障电弧检测", 《电子技术应用》, vol. 47, no. 11, 6 November 2021 (2021-11-06), pages 53 - 57 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116908632A (en) * | 2023-07-19 | 2023-10-20 | 常熟理工学院 | Low-voltage alternating-current series arc fault characteristic self-adaptive optimizing detection method and system |
| CN117454166A (en) * | 2023-10-11 | 2024-01-26 | 国网四川省电力公司电力科学研究院 | Method for identifying arc faults of ignition based on EffNet lightweight model |
| CN117454166B (en) * | 2023-10-11 | 2024-05-10 | 国网四川省电力公司电力科学研究院 | Method for identifying arc faults of induced thermal power based on EffNet lightweight model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shaik et al. | Power quality assessment and event detection in hybrid power system | |
| CN114720818A (en) | Alternating current series fault arc detection method based on time-frequency feature screening | |
| CN108508320B (en) | Arc grounding fault identification method based on harmonic energy and wave distortion feature | |
| Yılmaz et al. | A real-time UWT-based intelligent fault detection method for PV-based microgrids | |
| CN109444667B (en) | Power distribution network early fault classification method and device based on convolutional neural network | |
| CN107064752A (en) | A kind of distinguished number of aviation fault electric arc detection | |
| Chen et al. | Intelligent identification of voltage variation events based on IEEE Std 1159-2009 for SCADA of distributed energy system | |
| CN113917294B (en) | Intelligent self-adaptive arc detection method based on wavelet decomposition and application device thereof | |
| CN113376474A (en) | Neural network fault arc identification system and method based on generalized S transformation | |
| CN116526969A (en) | Photovoltaic fault arc detection method based on linear frequency modulation Z transformation and wavelet packet decomposition | |
| Wang et al. | An on-line detection method and device of series arc fault based on lightweight CNN | |
| CN109635430B (en) | Power grid transmission line transient signal monitoring method and system | |
| CN112949563B (en) | Non-invasive load identification method based on variable point detection and improved KNN algorithm | |
| CN118114019A (en) | Automatic identification method and system for power distribution network topology based on data analysis | |
| Zhu et al. | Aiming to Complex Power Quality Disturbances: A Novel Decomposition and Detection Framework | |
| Wang et al. | A novel series arc fault detection method based on mel-frequency cepstral coefficients and fully connected neural network | |
| Lin et al. | An earth fault diagnosis method based on online dynamically calculated thresholds for resonant ground systems | |
| CN113094983B (en) | Online simulation method for multi-dimensional time-varying characteristics of direct-current fault electric arc of photovoltaic system | |
| CN113610119A (en) | Method for identifying power transmission line developmental fault based on convolutional neural network | |
| CN116842425A (en) | Method for detecting fault arc in low-voltage alternating current system | |
| CN115808600B (en) | DC arc detection method based on mathematical morphology and pattern recognition | |
| Couto et al. | High impedance fault detection on microgrids considering the impact of VSC based generation | |
| CN117556371A (en) | Low-voltage series arc fault identification method, equipment and medium based on multidimensional features and XGBoost algorithm | |
| CN113363979B (en) | Non-contact power load decomposition sensing method and system | |
| Varghese P et al. | Application of signal processing techniques and intelligent classifiers for high-impedance fault detection in ensuring the reliable operation of power distribution systems |
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 |