CN109613402B - Power distribution network high-resistance grounding fault detection method based on wavelet transformation and neural network - Google Patents

Power distribution network high-resistance grounding fault detection method based on wavelet transformation and neural network Download PDF

Info

Publication number
CN109613402B
CN109613402B CN201910114060.9A CN201910114060A CN109613402B CN 109613402 B CN109613402 B CN 109613402B CN 201910114060 A CN201910114060 A CN 201910114060A CN 109613402 B CN109613402 B CN 109613402B
Authority
CN
China
Prior art keywords
neural network
neuron
power distribution
distribution network
detection method
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.)
Expired - Fee Related
Application number
CN201910114060.9A
Other languages
Chinese (zh)
Other versions
CN109613402A (en
Inventor
苏文聪
朱星宇
金涛
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fuzhou University
Original Assignee
Fuzhou University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Fuzhou University filed Critical Fuzhou University
Priority to CN201910114060.9A priority Critical patent/CN109613402B/en
Publication of CN109613402A publication Critical patent/CN109613402A/en
Application granted granted Critical
Publication of CN109613402B publication Critical patent/CN109613402B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/08—Locating faults in cables, transmission lines, or networks
    • G01R31/081—Locating faults in cables, transmission lines, or networks according to type of conductors
    • G01R31/086—Locating faults in cables, transmission lines, or networks according to type of conductors in power transmission or distribution networks, i.e. with interconnected conductors
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/08—Locating faults in cables, transmission lines, or networks
    • G01R31/088—Aspects of digital computing
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
    • Y04S10/52—Outage or fault management, e.g. fault detection or location

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Theoretical Computer Science (AREA)
  • Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)
  • Locating Faults (AREA)

Abstract

The invention relates to a high-resistance earth fault detection method for a power distribution network based on wavelet transformation and a neural network. The invention uses the evolved neural network to improve the traditional detection method. The evolved neural network is an intelligent system based on a dynamic connection structure, and the topological structure of the system can be adjusted through incremental learning so as to incorporate new information. The invention utilizes discrete wavelet transform to process fault signals, and inputs the fault signals into an evolved neural network so as to detect the high-resistance grounding fault of the power distribution network.

Description

Power distribution network high-resistance grounding fault detection method based on wavelet transformation and neural network
Technical Field
The invention relates to power distribution network ground fault detection, in particular to a power distribution network high-resistance ground fault detection method based on wavelet transformation and a neural network.
Background
High resistance ground fault (HIF) is a ground fault occurring when a power line passes through a conductive medium such as a road, soil, tree branches, or a cement building, and may occur in each voltage class to affect normal operation of a power distribution network. Due to the high impedance characteristic of the nonmetal conducting medium, when a high-impedance grounding fault occurs in a power distribution network, the fault current is very small and is often accompanied by an electric arc, and the common zero-sequence current protection is difficult to detect. Among the high resistance earth faults of low current earth systems, especially resonant earth systems, the arc earth fault is a large part. The grounding impedance of the arc grounding is greatly changed due to air ionization, so that the conventional protection is repeatedly started and recovered, the protection of adjacent lines and equipment is tripped out in an override mode, the whole system overvoltage is caused, the electrical equipment is further damaged, the accident is enlarged, and the power supply reliability of a power grid is reduced. The long-time fault operation of the line may cause the temperature of the fault point to be too high, thereby causing fire hazard and causing permanent damage to electrical equipment, and the step voltage around the grounding point can reach several kilovolts, thereby seriously threatening the stable operation and personal safety of the power system. The method can effectively detect the high-impedance grounding fault and provide a criterion for the subsequent line selection and positioning, so that the high-impedance grounding fault detection of the power distribution network is very important.
Disclosure of Invention
In view of this, the present invention provides a method for detecting a high-impedance ground fault of a power distribution network based on wavelet transformation and a neural network, which can effectively detect the high-impedance ground fault of the power distribution network.
In order to achieve the purpose, the invention adopts the following technical scheme:
a high-resistance grounding fault detection method for a power distribution network based on wavelet transformation and a neural network comprises the following steps:
step S1, collecting the current signals of the power distribution network line to be detected;
s2, adopting db4 wavelet as mother wavelet, carrying out discrete wavelet transform on the linear current signal and obtaining a reconstructed waveform;
step S3, setting the parameter of the evolved neural network;
step S4, taking the reconstructed waveform as the input quantity of the evolved neural network, and calculating the Manhattan distance between the input quantity and the k-th neuron of the evolution layer, the activation degree of the k-th neuron and the error between the output quantity and an expected value;
step S5, if the activation degree of the kth neuron is smaller than a preset threshold or the error between the output quantity and the expected value exceeds the preset threshold, placing a new neuron at (k +1) positions, otherwise, updating the weight;
step S6, calculating the Manhattan distance between the o-th neuron and the p-th neuron
Figure BDA0001969465130000021
And
Figure BDA0001969465130000022
in step S7, if yes
Figure BDA0001969465130000023
And
Figure BDA0001969465130000024
are all less than a threshold value DthrThen the two neurons are aggregated;
and S8, repeating the steps S4-S7 until all input quantities are processed, obtaining an evolved neural network output result, and judging whether the distribution network to be detected has a high-resistance grounding fault according to the output.
Further, the discrete wavelet transform has a sampling frequency of 10kHz and a sampling time of 0.5 s.
Further, the evolved neural network parameter includes a neuron activation level threshold athrError threshold value EthrManhattan distance threshold D of two neuronsthrLearning rate α1And alpha2。
Further, the parameter calculation in step S4 is specifically:
j (th) input quantity IjThe Manhattan distance from the kth neuron of the evolution layer is
Figure BDA0001969465130000031
Activation degree of kth neuron Ak＝1-Djk，WikIs the weight;
output quantity O1To the expected value
Figure BDA0001969465130000032
Error of (2)
Figure BDA0001969465130000033
Further, the step S5 is specifically: if AkIs less than AthrOr E1Over EthrPlacing a new neuron in (k +1) positions, otherwise updating the weight Wik(t+1)＝Wik(t)+α1(Ii(t+1)-Wik(t)) And Wk1(t+1)＝Wk1(t)+α2(AkE1)。
Further, the
Figure BDA0001969465130000034
And
Figure BDA0001969465130000035
the method specifically comprises the following steps:
Figure BDA0001969465130000036
Figure BDA0001969465130000037
wherein Wio、Wip、Wo1And Wp1Is a weight value.
Compared with the prior art, the invention has the following beneficial effects:
1. the neural network evolved by the method has no fixed topological structure, and can be evolved into a new structure on line after data processing.
2. The neural network evolved by the method can be rapidly learned through incremental training, and has good generalization capability and the capability of recombining a network model into a constantly changing environment, wherein the structure and the parameters are adapted simultaneously.
3. The neural network evolved by the method avoids catastrophic forgetting by adding and deleting the neurons on line and adjusting the connection rights of the neurons, and the reliability and the efficiency of detecting the high-resistance grounding fault of the power distribution network are improved.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
fig. 2 is a schematic structural diagram of an evolved neural network in an embodiment of the present invention.
Detailed Description
The invention is further explained below with reference to the drawings and the embodiments.
Referring to fig. 1, the present invention provides a method for detecting a high impedance ground fault of a power distribution network based on wavelet transformation and a neural network, comprising the following steps:
step S1, collecting the current signals of the power distribution network line to be detected;
s2, adopting db4 wavelet as mother wavelet, carrying out discrete wavelet transform on the linear current signal and obtaining a reconstructed waveform;
step S3, setting the parameter of the evolved neural network;
step S4, taking the reconstructed waveform as the input quantity of the evolved neural network, and calculating the Manhattan distance between the input quantity and the k-th neuron of the evolution layer, the activation degree of the k-th neuron and the error between the output quantity and an expected value;
step S5, if the activation degree of the kth neuron is smaller than a preset threshold or the error between the output quantity and the expected value exceeds the preset threshold, placing a new neuron at (k +1) positions, otherwise, updating the weight;
step S6, calculating the Manhattan distance between the o-th neuron and the p-th neuron
Figure BDA0001969465130000051
And
Figure BDA0001969465130000052
in step S7, if yes
Figure BDA0001969465130000053
And
Figure BDA0001969465130000054
are all less than a threshold value DthrThen the two neurons are aggregated;
and S8, repeating the steps S4-S7 until all input quantities are processed, obtaining an evolved neural network output result, and judging whether the distribution network to be detected has a high-resistance grounding fault according to the output.
In this embodiment, the discrete wavelet transform has a sampling frequency of 10kHz and a sampling time of 0.5 s.
In this embodiment, the evolved neural network parameter includes a neuron activation level threshold AthrError threshold value EthrManhattan distance threshold D of two neuronsthrLearning rate α1And alpha2。
Further, the parameter calculation in step S4 is specifically:
j (th) input quantity IjThe Manhattan distance from the kth neuron of the evolution layer is
Figure BDA0001969465130000055
Activation degree of kth neuron Ak＝1-Djk，WikIs the weight;
output quantity O1To the expected value
Figure BDA0001969465130000056
Error of (2)
Figure BDA0001969465130000057
In this embodiment, the step S5 specifically includes: if AkIs less than AthrOr E1Over EthrPlacing a new neuron in (k +1) positions, otherwise updating the weight Wik(t+1)＝Wik(t)+α1(Ii(t+1)-Wik(t)) And Wk1(t+1)＝Wk1(t)+α2(AkE1)。
In the present embodiment, the
Figure BDA0001969465130000061
And
Figure BDA0001969465130000062
the method specifically comprises the following steps:
Figure BDA0001969465130000063
Figure BDA0001969465130000064
wherein Wio、Wip、Wo1And Wp1Is a weight value.
The above description is only a preferred embodiment of the present invention, and all equivalent changes and modifications made in accordance with the claims of the present invention should be covered by the present invention.

Claims (6)

1. A high-resistance grounding fault detection method for a power distribution network based on wavelet transformation and a neural network is characterized by comprising the following steps:
step S1, collecting the current signals of the power distribution network line to be detected;
s2, adopting db4 wavelet as mother wavelet, carrying out discrete wavelet transform on the linear current signal and obtaining a reconstructed waveform;
step S3, setting the parameter of the evolved neural network;
step S4, taking the reconstructed waveform as the input quantity of the evolved neural network, and calculating the Manhattan distance between the input quantity and the k-th neuron of the evolution layer, the activation degree of the k-th neuron and the error between the output quantity and an expected value;
step S5, if the activation degree of the kth neuron is smaller than a preset threshold or the error between the output quantity and the expected value exceeds the preset threshold, placing a new neuron at (k +1) positions, otherwise, updating the weight;
step S6, calculating the Manhattan distance between the o-th neuron and the p-th neuron
Figure FDA0002629251250000011
And
Figure FDA0002629251250000012
in step S7, if yes
Figure FDA0002629251250000013
And
Figure FDA0002629251250000014
are all less than a threshold value DthrThen the two neurons are aggregated;
and S8, repeating the steps S4-S7 until all input quantities are processed, obtaining an evolved neural network output result, and judging whether the distribution network to be detected has a high-resistance grounding fault according to the output.
2. The method for detecting the high-resistance grounding fault of the power distribution network based on the wavelet transform and the neural network according to claim 1 is characterized in that: the discrete wavelet transform has the sampling frequency of 10kHz and the sampling time of 0.5 s.
3. The wavelet transform and neural network-based high-resistance ground fault detection method for the power distribution network according to claim 1, wherein: the evolved neural network parameters include a neuron activation degree threshold AthrError threshold value EthrManhattan distance threshold D of two neuronsthrLearning rate α1And alpha2。
4. The wavelet transform and neural network-based high-resistance ground fault detection method for the power distribution network according to claim 3, wherein: the step S4 is specifically performed by:
j (th) input quantity IjThe Manhattan distance from the kth neuron of the evolution layer is
Figure FDA0002629251250000021
Activation degree of kth neuron Ak＝1-Djk,WikIs the weight;
output quantity O1To the expected value
Figure FDA0002629251250000022
Error of (2)
Figure FDA0002629251250000023
5. The wavelet transform and neural network-based high-resistance ground fault detection method for the power distribution network according to claim 4, wherein: the step S5 specifically includes: if AkIs less than AthrOr E1Over EthrPlacing a new neuron in (k +1) positions, otherwise updating the weight Wik(t+1)＝Wik(t)+α1(Ii(t+1)-Wik(t)) And Wk1(t+1)＝Wk1(t)+α2(AkE1)。
6. The wavelet transform and neural network-based high-resistance ground fault detection method for the power distribution network according to claim 4, wherein: the above-mentioned
Figure FDA0002629251250000024
And
Figure FDA0002629251250000025
the method specifically comprises the following steps:
Figure FDA0002629251250000026
Figure FDA0002629251250000027
wherein Wio、Wip、Wo1And Wp1Is a weight value.
CN201910114060.9A 2019-02-14 2019-02-14 Power distribution network high-resistance grounding fault detection method based on wavelet transformation and neural network Expired - Fee Related CN109613402B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910114060.9A CN109613402B (en) 2019-02-14 2019-02-14 Power distribution network high-resistance grounding fault detection method based on wavelet transformation and neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910114060.9A CN109613402B (en) 2019-02-14 2019-02-14 Power distribution network high-resistance grounding fault detection method based on wavelet transformation and neural network

Publications (2)

Publication Number Publication Date
CN109613402A CN109613402A (en) 2019-04-12
CN109613402B true CN109613402B (en) 2020-11-27

Family

ID=66019693

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910114060.9A Expired - Fee Related CN109613402B (en) 2019-02-14 2019-02-14 Power distribution network high-resistance grounding fault detection method based on wavelet transformation and neural network

Country Status (1)

Country Link
CN (1) CN109613402B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113129161A (en) * 2019-12-31 2021-07-16 国网山东省电力公司威海供电公司 Wavelet decomposition and neural network-based animal fault analysis method in overhead power distribution system
CN113325269A (en) * 2021-05-28 2021-08-31 西安交通大学 Distribution network high-resistance fault monitoring method, system, equipment and storage medium
CN114089218A (en) * 2021-10-19 2022-02-25 广东电网有限责任公司东莞供电局 Power distribution network high-resistance grounding fault identification method, device, terminal and medium
CN115494350B (en) * 2022-11-21 2023-03-24 昆明理工大学 Alternating current transmission line lightning stroke fault recognition method and system

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
ATE536655T1 (en) * 2004-03-16 2011-12-15 Abb Technology Ag DIGITAL SIGNAL PROCESSOR IMPLEMENTATION OF HIGH-IMPEDANCE ERROR ALGORITHMS
CN101975910B (en) * 2010-09-07 2012-07-04 昆明理工大学 Intelligent fault classification and location method for ultra-high voltage direct current transmission line
TWI425226B (en) * 2011-07-06 2014-02-01 Univ Nat Taiwan Method and system for fault detection, identification and location in high-voltage power transmission networks
CN102611140B (en) * 2012-03-23 2014-04-30 合肥工业大学 Grid-connected inverter island detection method based on wavelet transform and neural network
CN103257304A (en) * 2013-04-10 2013-08-21 昆明理工大学 ANN fault line selection method through CWT coefficient RMS in zero-sequence current feature band
CN103728535B (en) * 2013-10-28 2016-09-14 昆明理工大学 A kind of extra-high-voltage direct-current transmission line fault location based on wavelet transformation transient state energy spectrum
CN103729687A (en) * 2013-12-18 2014-04-16 国网山西省电力公司晋中供电公司 Electricity price forecasting method based on wavelet transform and neural network
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
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
CN108279364B (en) * 2018-01-30 2020-01-14 福州大学 Power distribution network single-phase earth fault line selection method based on convolutional neural network

Also Published As

Publication number Publication date
CN109613402A (en) 2019-04-12

Similar Documents

Publication Publication Date Title
CN109613402B (en) Power distribution network high-resistance grounding fault detection method based on wavelet transformation and neural network
CN110082640B (en) Distribution network single-phase earth fault identification method based on long-time memory network
CN105425118A (en) Multi-information fusion fault arc detection method and device
CN111308263B (en) High-resistance grounding fault detection method for power distribution network
CN112147462A (en) Power transmission line fault identification method based on deep learning
CN110687395A (en) Fault line selection method for power distribution network with distributed power supply based on deep belief network
CN107561407B (en) Distributed transformer substation grounding grid detection and evaluation method
CN115932484B (en) Power transmission line fault identification and fault location method and device and electronic equipment
CN110865266B (en) Lightning-resistant horizontal test method for power transmission line of cross-shaped grounding device
CN105809574A (en) Method for judging cloud-to-ground lightning high-risk section of line corridor in combination with topographic parameters
Cong et al. Root-cause identification of single line-to-ground fault in urban small current grounding systems based on correlation dimension and average resistance
CN112557808A (en) Single-phase earth fault positioning method for power distribution network
Hashim et al. Investigation of features extraction condition for impedance-based fault location in transmission lines
CN116167205A (en) High-voltage transmission line hidden defect detection method and early warning system
CN112444758B (en) Intelligent power distribution network line fault diagnosis and classification evaluation method
Wu et al. Transmission line fault cause identification method based on transient waveform image and MCNN-LSTM
CN117668471A (en) Tree line discharge fault identification method based on fault traveling wave current characteristics
CN109375041B (en) Single-phase grounding fault judgment method for small-current grounding system
Wang et al. Transmission line fault diagnosis based on wavelet packet analysis and convolutional neural network
CN112255493A (en) Multi-criterion comprehensive voting power distribution network fault line selection method
da Silva et al. High impedance fault location-case study using wavelet transform and artificial neural networks
Cheong et al. A novel fault location technique based on current signals only for thyristor controlled series compensated transmission lines using wavelet analysis and self organising map neural networks
Jamil et al. Fault classification of three-phase transmission network using genetic algorithm
Zhang et al. A fast fault detection method for controlled fault interruption based on wavelet singularities and BP neural network
Huang et al. New recognition method of lightning interference in long distance VSC-HVDC transmission system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20201127

Termination date: 20220214