CN114019296A - Distribution line ground fault identification method based on BP neural network - Google Patents
Distribution line ground fault identification method based on BP neural network Download PDFInfo
- Publication number
- CN114019296A CN114019296A CN202111130863.7A CN202111130863A CN114019296A CN 114019296 A CN114019296 A CN 114019296A CN 202111130863 A CN202111130863 A CN 202111130863A CN 114019296 A CN114019296 A CN 114019296A
- Authority
- CN
- China
- Prior art keywords
- neural network
- training
- distribution line
- ground fault
- error
- 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
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 75
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000012549 training Methods 0.000 claims abstract description 60
- 238000012360 testing method Methods 0.000 claims abstract description 25
- 238000004364 calculation method Methods 0.000 claims abstract description 19
- 230000005540 biological transmission Effects 0.000 claims abstract description 9
- 210000002569 neuron Anatomy 0.000 claims abstract description 8
- 230000005284 excitation Effects 0.000 claims description 6
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 238000012886 linear function Methods 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 description 9
- 230000008569 process Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 208000034699 Vitreous floaters Diseases 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 239000012212 insulator Substances 0.000 description 1
- 230000016507 interphase Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000001629 suppression Effects 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
- G01R31/083—Locating faults in cables, transmission lines, or networks according to type of conductors in cables, e.g. underground
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)
Abstract
The invention provides a distribution line ground fault identification method based on a BP neural network, which comprises the following steps: acquiring waveform data of a certain distribution line in normal operation and when a ground fault occurs as sample data, and calculating to obtain an identification characteristic value of each waveform data; constructing a BP neural network, training the BP neural network by using a training sample, and inspecting the BP neural network by using a test sample; acquiring waveform data of a distribution line in actual operation, calculating an identification characteristic value of the waveform data, inputting the identification characteristic value into a BP (back propagation) neural network, and performing forward transmission calculation by using the BP neural network to obtain a result of whether a ground fault occurs. The invention utilizes the BP neural network, wherein the weight parameters of the neurons are obtained by learning and training samples, and the weight parameters are not required to be set manually and are simple and easy to operate. The ground fault identification is carried out on the distribution line in a targeted manner, and the identification accuracy is high.
Description
Technical Field
The invention relates to the technical field of distribution line ground fault identification of a power grid, in particular to a distribution line ground fault identification method based on a BP neural network.
Background
In the field of power grids, the operating environment of a distribution line is complex and severe, and the distribution line is easily affected by foreign objects such as tree barriers, floaters and the like to generate ground faults. And because distribution lines insulation grade is lower, therefore equipment such as insulator, arrester receive equipment quality, environmental pollution influence easily, the risk that takes place earth fault further increases. The existing power distribution network grounding mode is mostly ungrounded or grounded through an arc suppression coil, when single-phase grounding is caused to occur, the sensitivity of a substation line selection device is insufficient, the line selection accuracy is low, unnecessary tripping of a normal line can be caused, the operation of a power distribution line with a grounding fault can be caused, if electric arcs are generated, the fault is easily caused to be enlarged to an interphase short circuit, even, the public potential safety hazard of electric interference is caused, and the personal safety is seriously threatened.
Disclosure of Invention
The invention aims to provide a distribution line ground fault identification method based on a BP neural network, which can solve the problem caused by inaccurate ground fault detection in the prior art.
The purpose of the invention is realized by the following technical scheme:
the distribution line ground fault identification method based on the BP neural network comprises the following steps:
step S1, waveform data of a certain distribution line in normal operation and when a ground fault occurs are obtained as sample data, and an identification characteristic value of each waveform data is obtained through calculation;
step S2, constructing a BP neural network, training the BP neural network by using a training sample, and inspecting the BP neural network by using a test sample;
and step S3, acquiring waveform data of the distribution line in actual operation, calculating an identification characteristic value of the waveform data, inputting the identification characteristic value into a BP neural network, and performing forward transmission calculation by using the BP neural network to obtain a result of whether the ground fault occurs.
Further, the step S2 includes:
step S201, forward transfer calculation is carried out on the BP neural network by using the ith training sample;
step S202, calculating an error E;
step S203, error back propagation is carried out on the BP neural network by adopting a gradient descent method, and the weight omega is correctedijAnd a threshold value of bjSo that the error E (ω, b) is reduced;
and step S204, repeating the steps S201 to S203 until all training samples are trained.
And S205, inputting the test sample into the BP neural network, calculating the error E of the test sample, judging whether the error E of the test sample is lower than a set error threshold value, if not, starting a new round of sample training until the error E of the test sample is lower than the error threshold value, and finishing the training.
Further, the step S2 further includes: and setting a training time threshold, and finishing the training of the BP neural network when the training time is greater than the set training time threshold.
Further, the calculation formula of the forward transfer calculation is as follows:
f is the excitation function, ωijRepresenting the weight between the BP neural network node i and the node j; bjA threshold representing node j; x is the number ofjRepresents the output value of node j; m represents the number of neurons of the node i.
Further, the error E is calculated by the following formula:
djrepresenting all results, y, of the BP neural network output layerjFor the expected output of the training samples, n represents the number of output layer data results.
Further, the excitation function adopts a Sigmoid function or a linear function.
According to the distribution line ground fault identification method based on the BP neural network, the BP neural network can be used for pertinently identifying the ground fault of a certain distribution line. Because the judgment logic of the BP neural network depends on the training sample, the operation characteristics of the distribution line can be adapted only by using the sample collected by the distribution line as the training sample, the ground fault identification is carried out on the distribution line in a targeted manner, and the identification accuracy is high.
The BP neural network focuses on the relation between input quantity and output quantity, wherein weight parameters of neurons are obtained by learning and training samples, manual setting is not needed, and the operation is simple and easy. And the BP neural network can also be maintained and updated subsequently, and the BP neural network can be ensured to adapt to the latest operation condition only by acquiring a new sample and training the new sample.
The BP neural network has generalization capability, and can correctly judge other untrained scenes as long as scene characteristics are similar while ensuring correct judgment of the trained scenes.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flow chart of a distribution line ground fault identification method based on a BP neural network according to the present invention;
fig. 2 is a schematic diagram of a training process of the BP neural network of the present invention.
Detailed Description
The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.
The embodiments of the present disclosure are described below with specific examples, and other advantages and effects of the present disclosure will be readily apparent to those skilled in the art from the disclosure in the specification. It is to be understood that the described embodiments are merely illustrative of some, and not restrictive, of the embodiments of the disclosure. The disclosure may be embodied or carried out in various other specific embodiments, and various modifications and changes may be made in the details within the description without departing from the spirit of the disclosure. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present disclosure.
The invention provides a distribution line ground fault identification method based on a Back Propagation (BP) neural network, wherein the BP neural network only needs to give a certain amount of training samples because of self-learning, self-adaption and generalization capabilities, and the BP neural network can self-learn the rules and memorize the rules into neurons for use in the next calculation. Moreover, the BP neural network has the capability of correctly judging other untrained objects while ensuring that the objects of the required category are correctly judged. The method comprises the steps of taking an identification characteristic value of the distribution line ground fault as an input quantity, taking whether the ground fault occurs as an output quantity, constructing a large number of normal and fault samples, training a BP neural network to obtain the BP neural network with the function of identifying the ground fault, and calculating the ground fault identification characteristic value obtained by calculating the distribution line in real time by using the BP neural network so as to judge whether the distribution line has the ground fault.
Specifically, the distribution line ground fault identification method based on the BP neural network, as shown in fig. 1, includes the following steps:
and step S1, acquiring waveform data of a certain distribution line during normal operation and when the ground fault occurs as sample data, and calculating to obtain an identification characteristic value of each waveform data.
The waveform data includes a secondary voltage waveform and a current waveform. For training the BP neural network, a large number of secondary voltage and current waveforms of the distribution line which normally operate and have ground faults are obtained as training samples, and a certain proportion of samples are extracted from the training samples to serve as test samples.
Furthermore, the related identification characteristic value can be obtained through wavelet transformation, Fourier transformation, time domain calculation and other modes. Identifying the characteristic value is required to distinguish two operating states of normal operation and ground fault occurrence of the distribution line. And for the characteristic values of different operating conditions, a label of whether the ground fault occurs needs to be marked, and the label is used as expected output of a training sample so as to carry out supervised training on the BP neural network.
And step S2, constructing the BP neural network, training the BP neural network by using the training sample, and inspecting the BP neural network by using the testing sample.
Dividing sample data into training samples and test samples, constructing a BP neural network structure, determining the number of neurons of an input layer, an intermediate layer and an output layer, initializing weight parameters, starting forward transmission by using the training samples, performing error backward propagation after errors are obtained, updating neuron weights and thresholds, adopting a test sample inspection effect after all samples are trained once, obtaining a test sample error E after forward transmission calculation, judging whether the test sample error E is lower than a set threshold, finishing training if the test sample error E is lower than the threshold, judging whether a set training time threshold is reached if the test sample error E is higher than the threshold, finishing training if the test sample error E is reached, preventing the condition that the training cannot be finished due to divergence of the BP neural network training, and starting a new round of sample training if the test sample error E is not reached.
Specifically, the training process of step S2 is shown in fig. 2, and includes:
step S201, forward transmission calculation is carried out on the BP neural network by using the ith training sample.
Suppose the weight between the BP neural network node i and the node j is omegaijAnd m represents the number of neurons of the node i. The threshold of node j is bjThe output value of each node is xjThe output of each node is the output of all nodes in the upper layer, the cost of the nodeThe weight of all nodes on the upper layer and the threshold of the nodes are determined by an excitation function, namely:
wherein f is an excitation function, and generally adopts an S-shaped function (such as Sigmoid function) or a linear function. The above equation is the calculation process of the forward pass of the BP neural network.
Step S202, error E is calculated.
Assume all results of the output layer are djThen the error function is:
wherein y isjFor the expected output of the training samples, n represents the number of output layer data results.
Step S203, error back propagation is carried out on the BP neural network by adopting a gradient descent method, and the weight omega is correctedijAnd a threshold value of bjSo that the error E (ω, b) is reduced.
And step S204, repeating the steps S201 to S203 until all training samples are trained.
And S205, inputting the test sample into the BP neural network, calculating the error E of the test sample, judging whether the error E of the test sample is lower than a set error threshold value, if not, starting a new round of sample training until the error E of the test sample is lower than the error threshold value, and finishing the training.
Preferably, in order to prevent the situation that the training cannot be terminated due to the divergence of the training of the BP neural network, the invention sets a training time threshold, and ends the training of the BP neural network when the training time is greater than the set training time threshold.
And step S3, acquiring waveform data of the distribution line in actual operation, calculating an identification characteristic value of the waveform data, inputting the identification characteristic value into a BP neural network, and performing forward transmission calculation by using the BP neural network to obtain a result of whether the ground fault occurs.
The trained BP neural network is used as a tool for judging whether the distribution line has the ground fault, and the judgment on whether the ground fault occurs or not can be judged only by calculating the current ground fault identification characteristic value of the distribution line, wherein the calculation method of the characteristic value is consistent with the characteristic value adopted by the training sample and using the characteristic value as the input quantity of the BP neural network for forward transmission calculation.
The above description is for the purpose of illustrating embodiments of the invention and is not intended to limit the invention, and it will be apparent to those skilled in the art that any modification, equivalent replacement, or improvement made without departing from the spirit and principle of the invention shall fall within the protection scope of the invention.
Claims (6)
1. The distribution line ground fault identification method based on the BP neural network is characterized by comprising the following steps of:
step S1, waveform data of a certain distribution line in normal operation and when a ground fault occurs are obtained as sample data, and an identification characteristic value of each waveform data is obtained through calculation;
step S2, constructing a BP neural network, training the BP neural network by using a training sample, and inspecting the BP neural network by using a test sample;
and step S3, acquiring waveform data of the distribution line in actual operation, calculating an identification characteristic value of the waveform data, inputting the identification characteristic value into a BP neural network, and performing forward transmission calculation by using the BP neural network to obtain a result of whether the ground fault occurs.
2. The distribution line ground fault identification method based on the BP neural network according to claim 1, wherein the step S2 comprises:
step S201, forward transfer calculation is carried out on the BP neural network by using the ith training sample;
step S202, calculating an error E;
step S203, error back propagation is carried out on the BP neural network by adopting a gradient descent method, and the weight omega is correctedijAnd a threshold value of bjSo that the error E (ω, b) is reduced;
and step S204, repeating the steps S201 to S203 until all training samples are trained.
And S205, inputting the test sample into the BP neural network, calculating the error E of the test sample, judging whether the error E of the test sample is lower than a set error threshold value, if not, starting a new round of sample training until the error E of the test sample is lower than the error threshold value, and finishing the training.
3. The distribution line ground fault identification method based on the BP neural network according to claim 2, wherein the step S2 further comprises: and setting a training time threshold, and finishing the training of the BP neural network when the training time is greater than the set training time threshold.
4. The distribution line ground fault identification method based on the BP neural network as claimed in claim 2, wherein the calculation formula of the forward transmission calculation is:
f is the excitation function, ωijRepresenting the weight between the BP neural network node i and the node j; bjA threshold representing node j; x is the number ofjRepresents the output value of node j; m represents the number of neurons of the node i.
5. The distribution line ground fault identification method based on the BP neural network according to claim 2, wherein the error E is calculated by the following formula:
djrepresenting all results, y, of the BP neural network output layerjFor expecting the input of training samplesAnd n represents the number of output layer data results.
6. The distribution line ground fault identification method based on the BP neural network as claimed in claim 4, wherein the excitation function is a Sigmoid function or a linear function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111130863.7A CN114019296A (en) | 2021-09-26 | 2021-09-26 | Distribution line ground fault identification method based on BP neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111130863.7A CN114019296A (en) | 2021-09-26 | 2021-09-26 | Distribution line ground fault identification method based on BP neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114019296A true CN114019296A (en) | 2022-02-08 |
Family
ID=80054950
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111130863.7A Pending CN114019296A (en) | 2021-09-26 | 2021-09-26 | Distribution line ground fault identification method based on BP neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114019296A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116087692A (en) * | 2023-04-12 | 2023-05-09 | 国网四川省电力公司电力科学研究院 | Distribution network tree line discharge fault identification method, system, terminal and medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103884966A (en) * | 2014-04-15 | 2014-06-25 | 河海大学常州校区 | Power distribution network low-current single-phase earth fault positioning method based on neural network |
| CN104459373A (en) * | 2014-11-11 | 2015-03-25 | 广东电网有限责任公司东莞供电局 | Node voltage sag amplitude calculation method based on BP neural network |
| CN105759167A (en) * | 2016-01-28 | 2016-07-13 | 江苏省电力公司南京供电公司 | Wavelet neural network-based distribution network single-phase short circuit line selection method |
| CN108491404A (en) * | 2018-01-22 | 2018-09-04 | 国电南瑞科技股份有限公司 | A kind of state estimation bad data recognition method based on BP neural network |
| CN109871812A (en) * | 2019-02-22 | 2019-06-11 | 苏州工业园区测绘地理信息有限公司 | A kind of multi-temporal remote sensing image urban vegetation extracting method neural network based |
| WO2020015277A1 (en) * | 2018-07-20 | 2020-01-23 | 国电南瑞科技股份有限公司 | Arc light fault identifying device and method based on panoramic information |
| CN111062569A (en) * | 2019-11-15 | 2020-04-24 | 南京天能科创信息科技有限公司 | Low-current fault discrimination method based on BP neural network |
-
2021
- 2021-09-26 CN CN202111130863.7A patent/CN114019296A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103884966A (en) * | 2014-04-15 | 2014-06-25 | 河海大学常州校区 | Power distribution network low-current single-phase earth fault positioning method based on neural network |
| CN104459373A (en) * | 2014-11-11 | 2015-03-25 | 广东电网有限责任公司东莞供电局 | Node voltage sag amplitude calculation method based on BP neural network |
| CN105759167A (en) * | 2016-01-28 | 2016-07-13 | 江苏省电力公司南京供电公司 | Wavelet neural network-based distribution network single-phase short circuit line selection method |
| CN108491404A (en) * | 2018-01-22 | 2018-09-04 | 国电南瑞科技股份有限公司 | A kind of state estimation bad data recognition method based on BP neural network |
| WO2020015277A1 (en) * | 2018-07-20 | 2020-01-23 | 国电南瑞科技股份有限公司 | Arc light fault identifying device and method based on panoramic information |
| CN109871812A (en) * | 2019-02-22 | 2019-06-11 | 苏州工业园区测绘地理信息有限公司 | A kind of multi-temporal remote sensing image urban vegetation extracting method neural network based |
| CN111062569A (en) * | 2019-11-15 | 2020-04-24 | 南京天能科创信息科技有限公司 | Low-current fault discrimination method based on BP neural network |
Non-Patent Citations (1)
| Title |
|---|
| 姜晓东;任力;刘铭;李彦;陈平;: "基于BP神经网络的小电流接地故障选线方法", 山东理工大学学报(自然科学版), no. 01 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116087692A (en) * | 2023-04-12 | 2023-05-09 | 国网四川省电力公司电力科学研究院 | Distribution network tree line discharge fault identification method, system, terminal and medium |
| CN116087692B (en) * | 2023-04-12 | 2023-06-23 | 国网四川省电力公司电力科学研究院 | Distribution network tree line discharge fault identification method, system, terminal and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gholami et al. | Detecting the location of short-circuit faults in active distribution network using PMU-based state estimation | |
| Tong et al. | Detection and classification of transmission line transient faults based on graph convolutional neural network | |
| CN106908674B (en) | Transformer state evaluation method based on multi-state quantity prediction | |
| CN114414942B (en) | Power transmission line fault identification classifier, identification method and system based on transient waveform image identification | |
| Ferreira et al. | Probabilistic transmission line fault diagnosis using autonomous neural models | |
| CN107561407B (en) | Distributed transformer substation grounding grid detection and evaluation method | |
| Shaban et al. | A cascade of artificial neural networks to predict transformers oil parameters | |
| CN110543921A (en) | cable early fault identification method based on waveform learning | |
| Jamali et al. | A Fast and accurate fault location method for distribution networks with DG using genetic algorithms | |
| KR101025058B1 (en) | A method for tracking signal detection of low voltage lines using fuzzy RBF neural network algorithm | |
| CN113447766A (en) | Method, device, equipment and storage medium for detecting high-resistance ground fault | |
| CN114019296A (en) | Distribution line ground fault identification method based on BP neural network | |
| CN116089882A (en) | Cable fault prediction processing method and device and electronic equipment | |
| CN110265906B (en) | Transformer substation grounding grid state evaluation method and computer system | |
| Janarthanam et al. | Electrical faults-detection and classification using machine learning | |
| CN111062569A (en) | Low-current fault discrimination method based on BP neural network | |
| CN114236332A (en) | Power cable insulation state judgment method and system | |
| CN111695543B (en) | Method for identifying hidden danger discharge type of power transmission line based on traveling wave time-frequency characteristics | |
| CN117590147A (en) | Cable fault detection method based on multilayer feedforward neural network and genetic algorithm | |
| CN112710923A (en) | Data-driven single-phase earth fault line selection method based on post-fault steady-state information | |
| CN112613191A (en) | Cable health state evaluation method and device, computer equipment and storage medium | |
| CN103942416B (en) | Voltage estimation method based on weighted node spanning tree | |
| CN113392589B (en) | High-voltage direct-current converter station fault analysis method and system based on convolutional neural network | |
| CN114184766B (en) | Grounding grid soil hazard characteristic evaluation platform and method based on corrosive ions | |
| CN115902557A (en) | Switch cabinet fault diagnosis processing method and device and nonvolatile storage medium |
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 |