CN115980515A - Single-phase earth fault positioning method - Google Patents
Single-phase earth fault positioning method Download PDFInfo
- Publication number
- CN115980515A CN115980515A CN202310156920.1A CN202310156920A CN115980515A CN 115980515 A CN115980515 A CN 115980515A CN 202310156920 A CN202310156920 A CN 202310156920A CN 115980515 A CN115980515 A CN 115980515A
- Authority
- CN
- China
- Prior art keywords
- fault
- phase
- node
- layer
- neural network
- 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
Images
Classifications
-
- 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
- Locating Faults (AREA)
Abstract
The invention relates to the technical field of distribution network fault positioning, and discloses a single-phase earth fault positioning method, which comprises the following steps: s1, generating single-phase earth fault waveform data as a training set and a test set based on Simulink simulation; s2, acquiring field fault recording data as a verification set; s3, unifying sampling rates of the training set, the test set and the verification set; s4, designing a neural network; and S5, inputting the verification set into the trained neural network, and judging the relation between the position of the fault recording device and the fault position, so that the fault position can be positioned according to the circuit topological graph. The method utilizes the capability of deep learning to automatically extract the characteristics, avoids the problems of insufficient fault information, insufficient universality and low accuracy of the traditional method for extracting the characteristics by using a manual characteristic extractor, realizes a more accurate fault positioning algorithm, and can assist power grid operation and maintenance personnel to quickly and effectively process and solve the single-phase earth fault accident of the power distribution network.
Description
Technical Field
The invention relates to the technical field of distribution network fault positioning, in particular to a single-phase earth fault positioning method.
Background
In a power distribution network, single-phase earth faults are most likely to occur, and due to the fact that fault currents are small, lines can operate with faults, and power supply reliability is improved. If the single-phase earth fault is discovered slowly, an interphase short-circuit fault can be developed, the equipment is damaged, the fault range is enlarged, and the reliability of power supply is seriously influenced. Because an effective loop cannot be formed after the power distribution network is grounded in a single phase, the fault capacitance current is weak, and the fault representation information is not obvious particularly under the superposition influence of multiple factors such as over-compensation/under-compensation of an arc suppression coil, grounding transition resistance, three-phase imbalance of the power grid, function/precision limitation of a measuring device and the like. Due to the influences of factors such as grounding impedance change, intermittent arc restrike, fault evolution, fault concurrency and the like, the uncertainty of fault characteristics is high, so that accurate positioning of the grounding fault is very difficult, and the reliability of the power distribution network is seriously threatened.
The existing research results have a certain effect in single-phase earth fault positioning, but most of the results only select partial characteristics of a power distribution network, namely, a manual characteristic extractor is used for analyzing the specific attributes of certain faults, so that the fault information description is insufficient, the universality is insufficient, the accuracy is not high, and a dispatcher is not facilitated to make targeted fault treatment measures. The deep learning is more and more widely applied in the engineering field, is very good at automatically learning complex and useful characteristics from high-dimensional data sets, and can directly realize end-to-end task training or extract abstract characteristics to fully describe fault information compared with a plurality of excellent manual characteristic extractors in the past, thereby ensuring the generalization and the accuracy of a fault positioning algorithm.
Disclosure of Invention
Aiming at the defects and shortcomings of the prior art, the invention provides a single-phase earth fault positioning method, aiming at the problem of inaccurate fault positioning caused by insufficient characteristic information extracted by using a manual characteristic extractor in the traditional method, the invention realizes automatic extraction of fault waveform characteristics, fully extracts the characteristics and ensures the accuracy of fault positioning.
The purpose of the invention can be realized by the following technical scheme:
a single-phase earth fault positioning method comprises the following steps:
s1: generating single-phase earth fault waveform data by using Simulink simulation as a training set and a test set;
s2: installing each fault recording device into a medium-voltage transmission line;
when the power transmission line has a single-phase earth fault, a fault wave recording device in the power transmission line automatically records and stores three-phase voltage signal data and three-phase current signal data before and after the fault occurs, and records the three-phase voltage signal data and the three-phase current signal data as fault wave recording data;
taking fault recording data as a verification set;
s3: processing the training set, the test set and the verification set by utilizing a decimal sampling rate conversion algorithm, and unifying the sampling rates of the three-phase voltage signals and the three-phase current signals in the training set, the test set and the verification set;
s4: constructing a neural network and training the network by utilizing a training set and a test set;
s5: and inputting the verification set into the trained neural network, and judging the relation between the position of the fault recording device and the fault position, so that the fault position can be positioned according to the line topological graph.
Further, S1 specifically is:
s1-1: creating a Simulink line topology simulation model according to the medium-voltage transmission line topology domain;
s1-2: simulating single-phase earth fault waveform data under different parameters by using a Simulink simulation model;
the parameters include: the method comprises the following steps of low-voltage side load, transition resistance, fault position, fault occurrence phase, fault duration and sampling frequency;
the fault waveform data comprises three-phase voltage signals and three-phase current signal data;
s1-3: and randomly dividing the obtained fault waveform data into a training set and a testing set according to the proportion of 3.
Further, the fault recording data in S2 specifically include three-phase voltage signal and three-phase current signal data.
Further, the fractional sampling rate conversion algorithm in S3 specifically includes:
s3-1: the Fourier interpolation method is utilized to improve the sampling rate of three-phase voltage and current signals in the training set, the testing set and the verification set;
s3-2: and reducing and unifying the sampling rate of the three-phase voltage and current signals in the training set, the testing set and the verification set by using an integral multiple extraction method.
Further, the neural network in S4 includes four parts:
the first part is a three-phase data characteristic synthesis module which is an independent double-branch structure, each branch independently processes three-phase voltage and current signal data, and each branch comprises two layers of convolution layers; the first layer of the convolutional layer takes three-phase voltage and current signal input data as 3-channel one-dimensional data, the three-phase voltage and current signal data are expanded to 8 channels by utilizing convolutional kernel convolution operation, and the output characteristic length of the first layer of the convolutional layer is consistent with the length of the input data; the input of the second layer of convolution layer is the output of the first layer of convolution layer, the second layer of convolution layer compresses the input characteristic to 1 channel by utilizing convolution kernel convolution operation, and the output characteristic length of the second layer of convolution layer is consistent with the input characteristic length; splicing the output characteristics of the double branches together to serve as subsequent input;
the second part is a voltage-current characteristic comparison module which is formed by stacking a plurality of convolution layers, each convolution layer further compares the output characteristics of the convolution layer on the upper layer, and the final output characteristic has the channel number of 1 and the length of one fourth of the input length of the module;
the third part is a characteristic self-adaptive selection module which consists of a multilayer perceptron; a segment of features is divided into a plurality of segments of features after being input into the module; the multilayer perceptron calculates the importance degree of each segment of characteristics, and then synthesizes the multiple segments of characteristics into the whole segment of characteristics according to the calculation result, namely the output of the characteristic self-adaptive selection module, and the process is expressed by the following formula:
in the above formula, MLP is a multilayer perceptron; fe _ in i Inputting features for the ith segment; fe _ num is the number of input feature segments; fe _ out is the output of the characteristic self-adaptive selection module;
the fourth part is a classifier module which is composed of 4 layers of full connection layers, wherein the output of the first three layers adopts a LeakyRelu activation function, and the output of the last layer adopts a Softmax activation function; specifically, the method comprises the following steps:
in is the input value of the first three full-connected layers, z i And the Class _ num is the output value of the ith node of the last layer of the full connection layer, and is the number of the nodes.
Further, in the step S4, the neural network constructed by the Adam optimizer is trained, the learning rate is set to 0.0001, and the loss function uses a cross entropy function;
when the neural network training process reaches 1000 rounds or the accuracy of the test set is improved by less than 0.1% in any continuous 20 training rounds, the training is stopped and the parameter model with the highest accuracy of the test set is stored.
Further, the S5 specifically is:
s5-1, establishing a network description matrix D for the power distribution network system according to the topological relation; the assignment principle of the elements in the matrix is as follows:
if there is a line connection between node i and node j and node number i<j, then d when the direction of the node i pointing to the node j is the same as the reference direction ij =1; the node is a fault recording device;
if there is a line connection between node i and node j and node number i<j, then d when the direction of the node i pointing to the node j is different from the reference direction ij =0；
Otherwise let d ij =0；
S5-2, based on the verification set, generating a single-phase earth fault information sequence E = [ E ] by using the trained neural network 1 ，e 2 ，…，e n ](ii) a N represents the number of nodes, and if the neural network judges that the node i is positioned in the fault section, e i =1, otherwise e i =0；
The fault information sequence E = [ E ] 1 ，e 2 ，…，e n ]Substituting diagonal elements of the matrix D to obtain a fault discrimination matrix P;
s5-3, judging and distinguishing P in the matrix P if the fault occurs ii =1 and for all p in the same row element ij Each j corresponding to =1 has p jj And =1, it is determined that a single-phase ground fault occurs between the node i and the node j, that is, the line between the node i and the node j is a fault section.
Further, i in S5-3 is not equal to j.
The invention has the beneficial technical effects that: by utilizing the capability of deep learning and automatic feature extraction, the problems of insufficient fault information, insufficient universality and low accuracy caused by the fact that a manual feature extractor is used for extracting features in the traditional method are solved, a more accurate fault positioning algorithm is realized, and the single-phase grounding fault accident of the power distribution network can be rapidly and effectively solved by power grid operation and maintenance personnel.
Drawings
FIG. 1 is a general flow diagram of the present invention.
Fig. 2 is an overall diagram of a Simulink line topology simulation model in the embodiment of the present invention.
FIG. 3 is a diagram of a neural network architecture of the present invention.
Fig. 4 is a topology structure diagram of a power distribution network system in an embodiment of the present invention.
Fig. 5 is a part 1 of a Simulink line topology simulation model according to an embodiment of the present invention.
Fig. 6 is a part 2 of a Simulink line topology simulation model in the embodiment of the present invention.
Fig. 7 is a part 3 of a Simulink line topology simulation model according to an embodiment of the present invention.
Fig. 8 is a part 4 of a Simulink line topology simulation model according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the invention.
Example (b):
when a single-phase earth fault frequently occurs in a medium-voltage distribution network, if the fault cannot be timely eliminated, an interphase short-circuit fault can be developed, equipment is damaged, the fault range is enlarged, and the reliability of power supply is seriously influenced. The single-phase earth fault positioning algorithm based on the mechanism logic architecture neural network can assist operation and maintenance personnel to quickly position fault positions, eliminate faults and ensure power supply reliability.
As shown in fig. 1, a single-phase earth fault location method includes the following steps:
s1: generating single-phase earth fault waveform data by using Simulink simulation as a training set and a test set; specifically, the method comprises the following steps:
s1-1: a Simulink line topology simulation model is created from the medium voltage transmission line topology domain as shown in figures 2, 5, 6, 7, 8.
S1-2: and simulating single-phase earth fault waveform data under different parameters by using a Simulink simulation model.
The parameters include: low-voltage side load, transition resistance, fault location, fault occurrence phase, fault duration and sampling frequency.
The fault waveform data includes three-phase voltage signal and three-phase current signal data.
S1-3: and randomly dividing the obtained fault waveform data into a training set and a testing set according to the proportion of 3.
S2: acquiring fault recording data acquired by a fault recording device as a verification set; the method specifically comprises the following steps: the fault recording data is three-phase voltage and current signal data of fault recording stored and detected on an actual medium-voltage operating line by using a fault recording device.
S3: and processing the training set, the testing set and the verification set by utilizing a decimal sampling rate conversion algorithm, and unifying the sampling rates of the three-phase voltage signals and the three-phase current signals in the training set, the testing set and the verification set.
The decimal sampling rate conversion algorithm specifically comprises the following steps:
s3-1: improving the sampling rate of the signal sequence by utilizing a Fourier interpolation method; the specific process comprises the following steps:
three-phase voltage and current signals X in the time domain are converted into a frequency domain for interpolation, interpolation points are filled with 0, interpolation factors are utilized for reasonable scaling so as to keep the symmetry on the frequency domain, then the inversion is converted back to the time domain, and the data at the moment are represented. As shown in the following formula:
in the above formula, S is a frequency domain signal,
The signal length is l for a frequency domain signal after the X Fourier transform of an original time domain signal; />
Is a zero point interpolated in the frequency domain and has the length of n (l-1) -l; />
The interpolated data is transformed back to a time domain signal through inverse Fourier transform; the FFT and IFFT represent the fourier transform and inverse fourier transform functions, respectively.
S3-2: reducing and unifying the signal sampling rate by an integral multiple extraction method; the specific process is as follows:
for time domain signal sequence
Performing an integer M-fold extraction, the extracted sequence->
Comprises the following steps:
the above formula represents the pair sequence
One sample point is kept every M-1 values.
S4: constructing a neural network and training the network by utilizing a training set and a test set;
the neural network includes four parts, as shown in fig. 3:
the first part is a three-phase data characteristic synthesis module which is an independent double-branch structure, each branch independently processes three-phase voltage and current signal data, and each branch comprises two layers of convolution layers; the first layer of the convolutional layer takes three-phase voltage and current signal input data as 3-channel one-dimensional data, the three-phase voltage and current signal data are expanded to 8 channels by utilizing convolutional kernel convolution operation, and the output characteristic length of the first layer of the convolutional layer is consistent with the length of the input data; the input of the second layer of convolutional layer is the output of the first layer of convolutional layer, the second layer of convolutional layer compresses the input characteristic to 1 channel by using convolutional kernel convolution operation, and the output characteristic length of the second layer of convolutional layer is consistent with the input characteristic length; the output characteristics of the dual branches are spliced together as subsequent inputs.
The second part is a voltage-current characteristic comparison module which is formed by stacking a plurality of convolution layers, each convolution layer further compares the output characteristics of the convolution layer on the upper layer, and the number of channels of the final output characteristics is 1, and the length of the final output characteristics is one quarter of the input length of the module.
The third part is a characteristic self-adaptive selection module which consists of a multilayer perceptron; a segment of features is divided into a plurality of segments of features after being input into the module; the multilayer perceptron calculates the importance degree of each segment of characteristics, and then synthesizes the multiple segments of characteristics into the whole segment of characteristics according to the calculation result, namely the output of the characteristic self-adaptive selection module, and the process is expressed by the following formula:
in the above formula, MLP is a multilayer perceptron; fe _ in i Inputting features for the ith segment; fe _ num is the number of input feature segments; fe out is the output of the feature adaptive selection module.
The fourth part is a classifier module which is composed of 4 layers of full connection layers, wherein the output of the first three layers adopts a LeakyRelu activation function, and the output of the last layer adopts a Softmax activation function; specifically, the method comprises the following steps:
in is the input value of the first three full-connected layers, z i And the Class _ num is the output value of the ith node of the last layer of the full connection layer, and is the number of the nodes.
S4-2: the training process of the neural network constructed based on the mechanism logic is as follows: based on the training set of S1, an Adam optimizer is used for training a neural network constructed based on mechanism logic, the learning rate is set to be 0.0001, a cross entropy function is selected and used as a loss function, and the training round is set to be 1000 rounds.
And stopping the neural network training process when the accuracy of the neural network reaches 1000 rounds or the accuracy of the test set is improved by less than 0.1% in any continuous 20 training rounds, and storing the parameter model with the best performance of the neural network in the test set.
S5: and inputting the verification set into the trained neural network, and judging the relation between the position of the fault recording device and the fault position, so that the fault position can be positioned according to the line topological graph. The method specifically comprises the following steps:
s5-1: a network description matrix D is established for the power distribution network system according to the nodes and the topological relation of the power distribution network system, the topological structure of the power distribution network system is shown in figure 4, the nodes 1 to 6 exist, and the assignment principle of elements in the matrix is as follows:
if there is a line connection between node i and node j and node number i<j, then d when the direction of the node i pointing to the node j is the same as the reference direction ij =1; the node is a fault recording device;
if there is a line connection between node i and node j and node number i<j, then d when the direction of the node i pointing to the node j is different from the reference direction ij =0；
Otherwise let d ij =0。
Based on fig. 4 and according to the above assignment principle, a network description matrix D can be established as follows:
s5-2: generating a fault discrimination matrix P according to the internal and external information of the fault, specifically:
based on the verification set, generating a single-phase earth fault information sequence E = [ E ] by using the trained neural network 1 ，e 2 ，…，e n ](ii) a Wherein n represents the number of nodes, and if the neural network judges that the node i is positioned in the fault section, e i =1, otherwise e i =0；
The fault information sequence E = [ E ] 1 ，e 2 ，…，e n ]And substituting diagonal elements of the matrix D to obtain a fault discrimination matrix P.
If E = [1, 0] is substituted into the diagonal elements of the network description matrix D, the fault discrimination matrix P is
S5-3: if fault discrimination matrix P ii =1 and for all p in the same row element ij J corresponding to =1 all have p jj And =1, it is determined that a single-phase ground fault occurs between the node i and the node j, that is, the line between the node i and the node j is a fault section.
I in S5-3 is not equal to j.
According to the fault discrimination matrix P in the present embodiment, it is determined that the line between the node 1 and the node 2 is a fault section.
The above-mentioned embodiments are illustrative of the specific embodiments of the present invention, and are not restrictive, and those skilled in the relevant art can make various changes and modifications to obtain corresponding equivalent technical solutions without departing from the spirit and scope of the present invention, so that all equivalent technical solutions should be included in the scope of the present invention.
Claims (7)
1. A single-phase earth fault positioning method is characterized by comprising the following steps:
s1: generating single-phase earth fault waveform data by using Simulink simulation as a training set and a test set;
s2: installing each fault recording device into a medium-voltage transmission line;
when the power transmission line has a single-phase earth fault, a fault recording device in the power transmission line automatically records and stores three-phase voltage signals and three-phase current signal data before and after the fault occurs and records the three-phase voltage signals and the three-phase current signal data as fault recording data;
taking fault recording data as a verification set;
s3: processing the training set, the test set and the verification set by utilizing a decimal sampling rate conversion algorithm, and unifying the sampling rates of the three-phase voltage signals and the three-phase current signals in the training set, the test set and the verification set;
s4: constructing a neural network and training the network by utilizing a training set and a test set;
s5: and inputting the verification set into the trained neural network, and judging the relation between the position of the fault recording device and the fault position, so that the fault position can be positioned according to the circuit topological graph.
2. The single-phase earth fault positioning method according to claim 1, wherein S1 specifically is:
s1-1: creating a Simulink line topology simulation model according to the medium-voltage transmission line topology domain;
s1-2: simulating single-phase earth fault waveform data under different parameters by using a Simulink simulation model;
the parameters include: low-voltage side load, transition resistance, fault position, fault occurrence phase, fault duration and sampling frequency;
the fault waveform data comprises three-phase voltage signals and three-phase current signal data;
s1-3: and randomly dividing the obtained fault waveform data into a training set and a testing set according to the proportion of 3.
3. The single-phase ground fault positioning method according to claim 1, wherein the fractional sampling rate conversion algorithm in S3 is specifically:
s3-1: the Fourier interpolation method is utilized to improve the sampling rate of three-phase voltage and current signals in the training set, the testing set and the verification set;
s3-2: and reducing and unifying the sampling rate of the three-phase voltage and current signals in the training set, the testing set and the verification set by using an integral multiple extraction method.
4. The single-phase earth fault location method of claim 1, wherein the neural network in S4 comprises four parts:
the first part is a three-phase data characteristic synthesis module which is an independent double-branch structure, each branch independently processes three-phase voltage and current signal data, and each branch comprises two layers of convolution layers; the first layer of the convolutional layer takes three-phase voltage and current signal input data as 3-channel one-dimensional data, the three-phase voltage and current signal data are expanded to 8 channels by using convolutional kernel convolution operation, and the output characteristic length of the first layer of the convolutional layer is consistent with the length of the input data; the input of the second layer of convolutional layer is the output of the first layer of convolutional layer, the second layer of convolutional layer compresses the input characteristic to 1 channel by using convolutional kernel convolution operation, and the output characteristic length of the second layer of convolutional layer is consistent with the input characteristic length; splicing the output characteristics of the double branches together to serve as subsequent input;
the second part is a voltage-current characteristic comparison module which is formed by stacking a plurality of convolution layers, each convolution layer further compares the output characteristics of the convolution layer on the upper layer, and the final output characteristic has the channel number of 1 and the length of one fourth of the input length of the module;
the third part is a characteristic self-adaptive selection module which consists of a multilayer perceptron; a segment of features is divided into a plurality of segments of features after being input into the module; the multilayer perceptron calculates the importance degree of each segment of characteristics, and then synthesizes the multiple segments of characteristics into the whole segment of characteristics according to the calculation result, namely the output of the characteristic self-adaptive selection module, and the process is expressed by the following formula:
in the above formula, MLP is a multilayer perceptron; fe _ in i Inputting features for the ith segment; fe _ num is the number of input feature segments; fe _ out is the output of the characteristic self-adaptive selection module;
the fourth part is a classifier module which consists of 4 layers of full connection layers, wherein the output of the first three layers adopts a LeakyRelu activation function, and the output of the last layer adopts a Softmax activation function; specifically, the method comprises the following steps:
in is the input value of the first three full-connection layers, z i And the Class _ num is the output value of the ith node of the last layer of the full connection layer, and is the number of the nodes.
5. The single-phase earth fault location method according to claim 1, wherein in S4, the constructed neural network is trained by using an Adam optimizer, a learning rate is set to 0.0001, and a cross entropy function is used as a loss function;
when the neural network training process reaches 1000 rounds or the accuracy of the test set is improved by less than 0.1% in any continuous 20 training rounds, the training is stopped and the parameter model with the highest accuracy of the test set is stored.
6. The single-phase ground fault location method according to claim 1, wherein the S5 specifically is:
s5-1, establishing a network description matrix D for the power distribution network system according to the topological relation; the assignment principle of the elements in the matrix is as follows:
if there is a line connection between node i and node j and node number i<j, then d when the direction of the node i pointing to the node j is the same as the reference direction ij =1; the node is a fault recording device;
if there is a line connection between node i and node j and node number i<j, then d when the direction of the node i pointing to the node j is different from the reference direction ij =0；
Otherwise let d ij =0；
S5-2, based on the verification set, generating a single-phase earth fault information sequence E = [ E ] by using the trained neural network 1 ，e 2 ，…，e n ](ii) a Wherein n represents the number of nodes, and if the neural network judges that the node i is positioned in the fault section, e i =1, otherwise e i =0；
The fault information sequence E = [ E ] 1 ，e 2 ，…，e n ]Substituting diagonal elements of the matrix D to obtain a fault discrimination matrix P;
s5-3, judging and distinguishing P in the matrix P if the fault occurs ii =1 and for all p in the same row element ij J corresponding to =1 all have p jj And =1, it is determined that a single-phase ground fault occurs between the node i and the node j, that is, the line between the node i and the node j is a fault section.
7. The single-phase ground fault locating method according to claim 6, wherein i in S5-3 is not equal to j.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310156920.1A CN115980515A (en) | 2023-02-23 | 2023-02-23 | Single-phase earth fault positioning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310156920.1A CN115980515A (en) | 2023-02-23 | 2023-02-23 | Single-phase earth fault positioning method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115980515A true CN115980515A (en) | 2023-04-18 |
Family
ID=85961675
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310156920.1A Pending CN115980515A (en) | 2023-02-23 | 2023-02-23 | Single-phase earth fault positioning method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115980515A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117826019A (en) * | 2024-03-06 | 2024-04-05 | 国网吉林省电力有限公司长春供电公司 | Line single-phase grounding fault area and type detection method of neutral point ungrounded system |
-
2023
- 2023-02-23 CN CN202310156920.1A patent/CN115980515A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117826019A (en) * | 2024-03-06 | 2024-04-05 | 国网吉林省电力有限公司长春供电公司 | Line single-phase grounding fault area and type detection method of neutral point ungrounded system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108279364B (en) | Power distribution network single-phase earth fault line selection method based on convolutional neural network | |
| Ferreira et al. | A survey on intelligent system application to fault diagnosis in electric power system transmission lines | |
| Deng et al. | Fault location in loop distribution network using SVM technology | |
| Chen et al. | Intelligent transient overvoltages location in distribution systems using wavelet packet decomposition and general regression neural networks | |
| CN112149554B (en) | Training and fault detection methods of fault classification model and related devices | |
| Duan et al. | Method of inter‐turn fault detection for next‐generation smart transformers based on deep learning algorithm | |
| CN105322519A (en) | Big data fusion analysis and running state monitoring method for intelligent power distribution network | |
| Jana et al. | A novel zone division approach for power system fault detection using ANN-based pattern recognition technique | |
| CN115980515A (en) | Single-phase earth fault positioning method | |
| Moravej et al. | A new approach based on S-transform for discrimination and classification of inrush current from internal fault currents using probabilistic neural network | |
| Shiddieqy et al. | Power line transmission fault modeling and dataset generation for ai based automatic detection | |
| Grando et al. | Fault classification in power distribution systems using PMU data and machine learning | |
| Yang et al. | Intelligent fault detection and location scheme for modular multi‐level converter multi‐terminal high‐voltage direct current | |
| Yuan et al. | Faulty feeder detection based on image recognition of voltage-current waveforms in non-effectively grounded distribution networks | |
| Mehinović et al. | Application of artificial intelligence methods for determination of transients in the power system | |
| CN109782126B (en) | Power distribution network early fault detection method based on humanoid concept learning | |
| CN113610119B (en) | Method for identifying power transmission line development faults based on convolutional neural network | |
| Bessam et al. | A novel method for induction motors stator inter-turn short circuit fault diagnosis based on wavelet energy and neural network | |
| Jamali et al. | Adaptive single-pole auto-reclosure for transmission lines using sound phases currents and wavelet packet transform | |
| CN112710923A (en) | Data-driven single-phase earth fault line selection method based on post-fault steady-state information | |
| Kou et al. | Transmission line fault identification based on BP neural network | |
| de Souza et al. | High impedance fault detection in distribution systems: An approach based on fourier transform and artificial neural networks | |
| CN112710924B (en) | Single-phase earth fault line selection method based on steady-state variable quantity before and after fault | |
| CN114062832A (en) | Method and system for identifying short-circuit fault type of power distribution network | |
| Li et al. | Fault detection in power distribution systems based on gated recurrent attention network |
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 |