CN109917228A - A kind of traveling wave method distribution net work earthing fault localization method based on RBF neural - Google Patents
A kind of traveling wave method distribution net work earthing fault localization method based on RBF neural Download PDFInfo
- Publication number
- CN109917228A CN109917228A CN201910241239.0A CN201910241239A CN109917228A CN 109917228 A CN109917228 A CN 109917228A CN 201910241239 A CN201910241239 A CN 201910241239A CN 109917228 A CN109917228 A CN 109917228A
- Authority
- CN
- China
- Prior art keywords
- fault
- point
- rbf neural
- distance
- distribution 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
Landscapes
- Locating Faults (AREA)
Abstract
The invention discloses a kind of traveling wave method distribution net work earthing fault localization method based on RBF neural, fault point position length at a distance from measurement point is obtained by traveling wave method, but due to distribution net work structure complexity, the reasons such as lines branch is more only can not determine specific abort situation by distance.Therefore, after obtaining fault distance, the branched lines being likely to occur where fault point all under the distance is determined, the fault waveform feature of different branches is then extracted by RBF neural method, accurate judgement is carried out in which branch to fault point.By combining c-type traveling wave method and RBF neural, to achieve the purpose that fault point is accurately positioned in essence rapidly.
Description
Technical field
The present invention relates to a kind of traveling wave method power distribution network ground connection based on RBF neural for distribution network automated field
Fault Locating Method.
Background technique
Power distribution network still has very big problem when area is built other than city at present.Some areas are due to big face
Product afforestation, trees fast-growth are unable to amicable settlement tree line contradiction, and furthermore birds influence, insulation rate is low, disastrous day
The odjective causes such as gas all easily cause short circuit or the ground connection of conducting wire.After distribution is broken down, shone by geographical conditions, night
The influence of the factors such as bright deficiency, bad weather affects electric power confession so that staff's lookup ground fault position time is long
It answers.And prolonged ground fault also has certain influence to the insulation of equipment itself, once ground fault develops into multipoint fault,
It is easy to cause trip accident.
When current singlephase earth fault occurs by investigation, even if part substation has been mounted with line selection device for low current,
It is general still to operate a switch power failure by outlet one by one to judge faulty line but since precision is lower, seriously affect the reliable of power supply
Property.Failure line selection traditional at present using zero-sequence current method is compared, realize in protection for feed line, only goes out to substation by inconvenience
Line carries out failure line selection, can not positioning failure section.And the distribution for 10 KV distribution routes based on radial, power supply become
Diameter is longer, and distribution network line branch is more, and when ground fault occurs for route, service personnel needs to open from substation's station outlet
Begin, along faulty line, find next stage branch point, specific faulty line is then determined by the bar knife mode for manually trying to draw
Section.The trouble shoot method, human input is larger, and the malfunction elimination time is longer, brings hidden danger to the safe operation of system.
Summary of the invention
The purpose of the invention is to overcome the deficiencies of the prior art and provide a kind of traveling wave method based on RBF neural
Distribution net work earthing fault localization method, it can assist in the efficient positioning that staff carries out multiple-limb distribution net work earthing fault.
Realizing a kind of technical solution of above-mentioned purpose is: a kind of traveling wave method power distribution network ground connection event based on RBF neural
Hinder localization method, includes the following steps:
Step 1, measuring process carries out the positioning of ground fault distance and position using c-type traveling wave method, includes the following steps；
Step 1.1, in the normal mutually with failure mutually in time point t of the head end of multiple terminals power distribution network1A row is inputted respectively
Wave signal；
Step 1.2, the waveform measurement that the travelling wave signal of reflection is proceeded through in the head end of multiple terminals power distribution network obtains normal
Phase travelling wave signal waveform and failure phase travelling wave signal waveform；
Step 1.3, more normal phase travelling wave signal waveform and failure phase travelling wave signal waveform are found between the two first
Difference time point t2；
Step 1.3, calculate fault point between measurement point at a distance from, formula is as follows:
In formula, XLFor earth fault to the distance between measurement point, V is travelling wave signal velocity of wave；
Step 2, grounding point simulated database is established, and with ATP software analogue ground point current-voltage waveform, forms route
Database file under ground state, includes the following steps；
Step 2.1, substitutional connection dry run figure is established by the actual track method of operation in ATP software；
Step 2.2, in ATP software, incoming line parameter (including become on electric current, voltage, frequency, line length, route
Depressor equivalent capacity), by obtaining electric current, voltage change data value in the case where ground state occurs for different circuit branch roads, is formed and closed
Key electrical quantity (electric current, voltage) database file；
Step 3, RBF neural grounding point fault location training step, the database file that ATP software is formed as
Training sample is stored in P_train file, the different branches of multiple terminals power distribution network is set as different fault type, is stored in
In T_train file；The position of fault point and the corresponding relationship of incoming line parameter state are carried out using RBF neural method
Training, obtains RBF neural grounding point fault location model；
Step 4, fault current, the input of voltage waveform value trained RBF neural grounding point event will be actually occurred
Hinder location model, exports the branch range of multiple terminals power distribution network；
Step 5, positioning step, according to fault point between measurement point distance XLWith the branch range two of multiple terminals power distribution network
A condition, the specific specific ground fault position for positioning multiple terminals power distribution network.
A kind of traveling wave method distribution net work earthing fault localization method based on RBF neural of the invention, by using row
The method of wave method positions power outlet to the distance between earth fault, due to the complexity of distribution network structure, the mostly spy of branch
Property, a possibility that there are multiple nodes under same distance within the scope of distribution, be grounded with ATP Software Create route each position therefore
Critical electrical amount changing value under barrier, is trained these data by RBF neural network algorithm, critical electrical amount is changed
Value input RBF neural network algorithm model, just can determine that ground connection when finding that these critical datas change in operational process
The branch range of failure.The distribution branch range information of fault point distance information and fault point is combined, it is accurate to can be obtained
Position of failure point information.
Detailed description of the invention
Fig. 1 is a kind of ground connection of traveling wave method distribution net work earthing fault localization method based on RBF neural of the invention
The schematic diagram of range measurement between fault point F to measurement point M；
Fig. 2 is a kind of RBF mind of traveling wave method distribution net work earthing fault localization method based on RBF neural of the invention
Through network function schematic diagram；
A kind of RBF mind of the position Fig. 3 traveling wave method distribution net work earthing fault localization method based on RBF neural of the invention
Through network from input neuron X to hidden layer input value r schematic diagram of the function.
Specific embodiment
In order to preferably understand technical solution of the present invention, below by specifically embodiment and in conjunction with attached drawing
It is described in detail:
A kind of traveling wave method distribution net work earthing fault localization method based on RBF neural of the invention, including walk as follows
It is rapid:
Step 1, measuring process carries out the positioning of ground fault distance and position using c-type traveling wave method, includes the following steps；
Step 1.1, in the normal mutually with failure mutually in time point t of the head end of multiple terminals power distribution network1A row is inputted respectively
Wave signal；
Step 1.2, the waveform measurement that the travelling wave signal of reflection is proceeded through in the head end of multiple terminals power distribution network obtains normal
Phase travelling wave signal waveform and failure phase travelling wave signal waveform；
Step 1.3, more normal phase travelling wave signal waveform and failure phase travelling wave signal waveform are found between the two first
Difference time point t2；
Step 1.3, calculate fault point between measurement point at a distance from, formula is as follows:
In formula, XLFor earth fault to the distance between measurement point, V is travelling wave signal velocity of wave；
Step 2, grounding point simulated database is established, and with ATP software analogue ground point current-voltage waveform, forms route
Database file under ground state, includes the following steps；
Step 2.1, substitutional connection dry run figure is established by the actual track method of operation in ATP software；
Step 2.2, in ATP software, incoming line parameter (including become on electric current, voltage, frequency, line length, route
Depressor equivalent capacity), by obtaining electric current, voltage change data value in the case where ground state occurs for different circuit branch roads, form number
According to library file；
Step 3, RBF neural grounding point fault location training step, the database file that ATP software is formed as
Training sample is stored in P_train file, the different branches of multiple terminals power distribution network is set as different fault type, is stored in
In T_train file；The position of fault point and the corresponding relationship of incoming line parameter state are carried out using RBF neural method
Training, obtains RBF neural grounding point fault location model；
Step 4, fault current, the input of voltage waveform value trained RBF neural grounding point event will be actually occurred
Hinder location model, exports the branch range of multiple terminals power distribution network；
Step 5, positioning step, according to fault point between measurement point distance XLWith the branch range two of multiple terminals power distribution network
A condition, the specific specific ground fault position for positioning multiple terminals power distribution network.
It determines transmitting and receives travelling wave signal time t1、t2Key be determining fault point back wave.By in normal phase
A manual signal is mutually injected separately into failure, the two travelling wave signals waveform before reaching fault point is all consistent, until
Travelling wave signal where failure phase arrives at the reflection that can occur when failure is pointed out different from travelling wave signal where normal phase, by normal phase
It is compared with the travelling wave signal of failure phase, the first discrepancy of two waveforms is exactly the back wave from fault point, the reflection
Waveform is exactly the back wave of fault point.
Referring to Fig. 2, the RBF neural includes input layer, hidden layer and output layer, the hidden layer is using radial
Basic function is usually Gaussian function as excitation function, the radial basis function.Referring to Fig. 2, each nerve of the input layer
Weight vector W1 and the distance between input vector X of member multiplied by radial basis function threshold value b1 as their own hidden layer it is defeated
Enter, the input k of hidden layer can be obtained.
For having from i-th training to the RBF neural of jth time training
And i-th training after hidden layer output are as follows:
The input of output layer is the weighted sum of each hidden layer neuron output.Since excitation function is purely linear function,
Export y are as follows:
Wherein, w2iFor the weight vector of hidden layer and output interlayer after i training.
The training process of RBF network is divided into two steps: the first step learns for no teacher's formula, determines training input layer and hidden layer
Between weight w 1；Second step is to have the study of teacher's formula, determines training hidden layer and exports the weight w 2 of interlayer.Before training,
Need to provide input vector X, corresponding target vector and radial basis function threshold value b1.Trained purpose is seek two layers final
Weight w 1, w2 and threshold value b1.
It in RBF network training process, is trained since 0 neuron, by checking that output error increases network automatically
Add neuron.It is recycled every time, input vector corresponding to the worst error for generating network is as weight vector w1i, produce
A raw new hidden layer neuron, then checks the error of new network, repeats this process until reaching error requirements or maximum
Until hidden layer neuron number.It can be seen that radial primary function network have structure adaptive determining, output and initial weight without
The features such as pass.
In the method, input vector X is failure phase travelling wave signal waveform (the Current Voltage wave of power distribution network difference branch
Shape), input quantity is the specific branch of power distribution network.Distribution net work earthing fault positioning each time, which is calculated by RBF neural, draws
Determine distribution network failure branch range, and combines earth fault to the distance X between measurement pointLDetermine specific fault point；In reality
After determining specific fault point, then it is anti-to RBF neural progress by branch where fault point and failure phase travelling wave signal waveform
Feedback training, to further increase its precision.
Those of ordinary skill in the art it should be appreciated that more than embodiment be intended merely to illustrate the present invention,
And be not used as limitation of the invention, as long as the change in spirit of the invention, to embodiment described above
Change, modification will all be fallen within the scope of claims of the present invention.
Claims (1)
1. a kind of traveling wave method distribution net work earthing fault localization method based on RBF neural, which is characterized in that including walking as follows
It is rapid:
Step 1, measuring process carries out the positioning of ground fault distance and position using c-type traveling wave method, includes the following steps；
Step 1.1, in the normal mutually with failure mutually in time point t of the head end of multiple terminals power distribution network1A traveling wave letter is inputted respectively
Number；
Step 1.2, the waveform measurement that the travelling wave signal of reflection is proceeded through in the head end of multiple terminals power distribution network obtains normal mutually row
Wave signal waveform and failure phase travelling wave signal waveform；
Step 1.3, more normal phase travelling wave signal waveform and failure phase travelling wave signal waveform, find first difference between the two
Time point t2；
Step 1.3, calculate fault point between measurement point at a distance from, formula is as follows:
In formula, XLFor earth fault to the distance between measurement point, V is travelling wave signal velocity of wave；
Step 2, grounding point simulated database is established, and with ATP software analogue ground point current-voltage waveform, forms failure phase line
Crucial electrical quantity (electric current, voltage) database file, includes the following steps under the ground state of road；
Step 2.1, substitutional connection dry run figure is established by the actual track method of operation in ATP software；
Step 2.2, in ATP software, incoming line parameter (including transformer on electric current, voltage, frequency, line length, route
Equivalent capacity), by obtaining electric current, voltage change data value in the case where ground state occurs for different circuit branch roads, form database
File；
Step 3, RBF neural grounding point fault location training step, the database file that ATP software is formed is as training
Sample is stored in P_train file, and the different branches of multiple terminals power distribution network are set as different fault type, are stored in T_
In train file；The position of fault point and the corresponding relationship of incoming line parameter state are instructed using RBF neural method
Practice, obtains RBF neural grounding point fault location model；
Step 4, actually occurring fault current, the input of voltage waveform value, trained RBF neural ground connection point failure is determined
Bit model exports the branch range of multiple terminals power distribution network；
Step 5, positioning step, according to fault point between measurement point distance XLWith two items of branch range of multiple terminals power distribution network
Part, the specific specific ground fault position for positioning multiple terminals power distribution network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910241239.0A CN109917228A (en) | 2019-03-28 | 2019-03-28 | A kind of traveling wave method distribution net work earthing fault localization method based on RBF neural |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910241239.0A CN109917228A (en) | 2019-03-28 | 2019-03-28 | A kind of traveling wave method distribution net work earthing fault localization method based on RBF neural |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109917228A true CN109917228A (en) | 2019-06-21 |
Family
ID=66967343
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910241239.0A Pending CN109917228A (en) | 2019-03-28 | 2019-03-28 | A kind of traveling wave method distribution net work earthing fault localization method based on RBF neural |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109917228A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112782532A (en) * | 2020-10-19 | 2021-05-11 | 国网辽宁省电力有限公司 | Power distribution network fault location method based on traveling wave signal generated by circuit breaker closing |
| CN113447764A (en) * | 2021-08-09 | 2021-09-28 | 安徽恒凯电力保护设备有限公司 | Intelligent monitoring and fault control method applied to power grid |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5640103A (en) * | 1994-06-30 | 1997-06-17 | Siemens Corporate Research, Inc. | Radial basis function neural network autoassociator and method for induction motor monitoring |
| JP2000088925A (en) * | 1998-09-14 | 2000-03-31 | Toshiba Corp | Method and apparatus for specifying fault position of semiconductor device |
| CN103091603A (en) * | 2013-01-14 | 2013-05-08 | 华北电力大学 | Breakdown intelligent classification and positioning method of electric transmission line |
| CN103278747A (en) * | 2013-06-03 | 2013-09-04 | 东南大学 | High-tension transmission line single-ended traveling wave fault distance detection method combined with time-frequency characteristics |
| CN103745229A (en) * | 2013-12-31 | 2014-04-23 | 北京泰乐德信息技术有限公司 | Method and system of fault diagnosis of rail transit based on SVM (Support Vector Machine) |
| CN104360227A (en) * | 2014-10-31 | 2015-02-18 | 国家电网公司 | Substation cable outlet fault monitoring method based on traveling wave method and transient basic frequency method |
| CN105137293A (en) * | 2015-09-24 | 2015-12-09 | 国网技术学院 | Positioning method of fault points in power distribution network mixed circuits |
| CN107153735A (en) * | 2017-04-28 | 2017-09-12 | 无锡开放大学 | Motor driven systems PWM inverter method for diagnosing faults |
| US9864004B1 (en) * | 2016-03-17 | 2018-01-09 | Cadence Design Systems, Inc. | System and method for diagnosing failure locations in electronic circuits |
| CN108008247A (en) * | 2017-11-24 | 2018-05-08 | 国网北京市电力公司 | Distribution net work earthing fault localization method and device |
-
2019
- 2019-03-28 CN CN201910241239.0A patent/CN109917228A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5640103A (en) * | 1994-06-30 | 1997-06-17 | Siemens Corporate Research, Inc. | Radial basis function neural network autoassociator and method for induction motor monitoring |
| JP2000088925A (en) * | 1998-09-14 | 2000-03-31 | Toshiba Corp | Method and apparatus for specifying fault position of semiconductor device |
| CN103091603A (en) * | 2013-01-14 | 2013-05-08 | 华北电力大学 | Breakdown intelligent classification and positioning method of electric transmission line |
| CN103278747A (en) * | 2013-06-03 | 2013-09-04 | 东南大学 | High-tension transmission line single-ended traveling wave fault distance detection method combined with time-frequency characteristics |
| CN103745229A (en) * | 2013-12-31 | 2014-04-23 | 北京泰乐德信息技术有限公司 | Method and system of fault diagnosis of rail transit based on SVM (Support Vector Machine) |
| CN104360227A (en) * | 2014-10-31 | 2015-02-18 | 国家电网公司 | Substation cable outlet fault monitoring method based on traveling wave method and transient basic frequency method |
| CN105137293A (en) * | 2015-09-24 | 2015-12-09 | 国网技术学院 | Positioning method of fault points in power distribution network mixed circuits |
| US9864004B1 (en) * | 2016-03-17 | 2018-01-09 | Cadence Design Systems, Inc. | System and method for diagnosing failure locations in electronic circuits |
| CN107153735A (en) * | 2017-04-28 | 2017-09-12 | 无锡开放大学 | Motor driven systems PWM inverter method for diagnosing faults |
| CN108008247A (en) * | 2017-11-24 | 2018-05-08 | 国网北京市电力公司 | Distribution net work earthing fault localization method and device |
Non-Patent Citations (7)
| Title |
|---|
| HANG CUI等: "HVDC transmission line fault localization base on RBF neural network with wavelet packet decomposition", 《 2015 12TH INTERNATIONAL CONFERENCE ON SERVICE SYSTEMS AND SERVICE MANAGEMENT (ICSSSM)》 * |
| R.N. MAHANTY等: "Application of RBF neural network to fault classification and location in transmission lines", 《IEE PROCEEDINGS - GENERATION, TRANSMISSION AND DISTRIBUTION》 * |
| 严凤等: "基于C型行波与SVM的配电线路故障定位", 《电力系统及其自动化学报》 * |
| 严凤等: "基于径向基神经网络的配电网单相接地故障研究", 《电测与仪表》 * |
| 严凤等: "基于神经网络的配电线路故障定位方法", 《电力系统及其自动化学报》 * |
| 严凤等: "新型10KV配电线路综合故障定位方法", 《电力系统及其自动化学报》 * |
| 王育飞等: "基于混沌-RBF神经网络的光伏发电功率超短期预测模型", 《电网技术》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112782532A (en) * | 2020-10-19 | 2021-05-11 | 国网辽宁省电力有限公司 | Power distribution network fault location method based on traveling wave signal generated by circuit breaker closing |
| CN113447764A (en) * | 2021-08-09 | 2021-09-28 | 安徽恒凯电力保护设备有限公司 | Intelligent monitoring and fault control method applied to power grid |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109270407B (en) | Extra-high voltage direct current transmission line fault reason identification method based on multi-source information fusion | |
| Xie et al. | Predicting lightning outages of transmission lines using generalized regression neural network | |
| CN105353268B (en) | One kind is used for the judgement of transmission line of electricity distribution traveling wave fault and localization method | |
| CN107219440B (en) | The localization method of single-ended radiation type distribution network singlephase earth fault | |
| CN102074937B (en) | Ranging method for adaptive transmission line fault of model | |
| CN104898021B (en) | A kind of distribution network fault line selection method based on k means cluster analyses | |
| Aziz et al. | Applications of ANFIS in high impedance faults detection and classification in distribution networks | |
| CN103869171B (en) | Zero-sequence parameter measuring method for ultrahigh-voltage transmission line with four-circuit alternating current on one tower and double-circuit double-electrode direct current | |
| CN108120903A (en) | A kind of low-current single-phase earth fault line selection method based on pulse nerve membranous system | |
| CN109917228A (en) | A kind of traveling wave method distribution net work earthing fault localization method based on RBF neural | |
| CN104599193A (en) | Rule base based single-phase ground fault positioning method of distribution network | |
| CN105785232A (en) | Method for synthetically selecting lines of single-phase ground fault of small current ground system | |
| CN108256234A (en) | A kind of method and system for being used to assess transformer DC magnetic bias influence | |
| CN114236288B (en) | Fault positioning method based on power transmission line | |
| CN107271842A (en) | A kind of parallel erected on same tower double-circuit line Fault Locating Method based on positive-sequence component in the same direction | |
| CN107728006A (en) | Traction substation traction current is monitored on-line with ground state and fault diagnosis system | |
| CN109188193A (en) | A kind of distribution network fault line selection method based on feature band convergence Min Shi distance | |
| CN113985194A (en) | Power distribution network fault positioning method based on stack self-encoder | |
| CN104849620A (en) | Grounding grid fault diagnosis method based on BP neural network | |
| Xu | Fault location and incipient fault detection in distribution cables | |
| CN110082634A (en) | The one-phase earthing failure in electric distribution network localization method of wide area current time sequence | |
| CN106597161A (en) | Shunting coefficient obtaining method of short circuit current of overhead line ground wire | |
| Naidu et al. | Hybrid fault location method for two terminal mixed lines using one ended measurements | |
| CN108169612A (en) | The localization method of overhead line DC transmission system DC short trouble | |
| CN110736900A (en) | distribution line single-phase earth fault line selection method based on directional traveling waves |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190621 |
|
| WD01 | Invention patent application deemed withdrawn after publication |