CN107132443A - A kind of three-level STATCOM intelligent failure diagnosis method - Google Patents

A kind of three-level STATCOM intelligent failure diagnosis method Download PDF

Info

Publication number
CN107132443A
CN107132443A CN201710119870.4A CN201710119870A CN107132443A CN 107132443 A CN107132443 A CN 107132443A CN 201710119870 A CN201710119870 A CN 201710119870A CN 107132443 A CN107132443 A CN 107132443A
Authority
CN
China
Prior art keywords
fault
feature vector
statcom
energy
support vector
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.)
Granted
Application number
CN201710119870.4A
Other languages
Chinese (zh)
Other versions
CN107132443B (en
Inventor
张晓华
李浩洋
郭源博
李林
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.)
Dalian University of Technology
Original Assignee
Dalian University of Technology
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 Dalian University of Technology filed Critical Dalian University of Technology
Priority to CN201710119870.4A priority Critical patent/CN107132443B/en
Publication of CN107132443A publication Critical patent/CN107132443A/en
Application granted granted Critical
Publication of CN107132443B publication Critical patent/CN107132443B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/50Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/24Classification techniques
    • G06F18/241Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08Feature extraction

Abstract

The invention discloses a kind of three-level STATCOM intelligent failure diagnosis method, it includes:Running status to three-level STATCOM main circuit power device is divided;Set up the corresponding three-level STATCOM simulation model of every kind of fault type and obtain sample;Feature extraction is carried out to each sample collected using based on Wavelet Packet Energy Spectrum and average current compound characteristics extractive technique, and obtains the fault feature vector corresponding to each fault type;Multi-category support vector machines model is constructed, and is trained based on genetic algorithm and carries out parameter optimization;Open fault diagnosis is carried out to three-level STATCOM power device using the multi-category support vector machines model trained.The present invention is used as fault diagnosis signal using output current, it is not necessary to increase extra voltage sensor, and accurately extracts fault feature vector while improving fault diagnosis accuracy.

Description

A kind of three-level STATCOM intelligent failure diagnosis method
Technical field
The present invention relates to power electronics fault diagnosis field, particularly relate to a kind of to the main electricity of three-level STATCOM The method that road power device open circuit carries out fault diagnosis.
Background technology
At present, neutral-point-clamped formula (NPC) three-level STATCOM is because its high pressure, capacity are big, output harmonic wave is small and structure Relatively easy the advantages of, it is widely used.Because the quantity of three-level STATCOM power switch pipe is that two level are opened up Flutter structure two times, therefore the possibility that power switch pipe breaks down also greatly increases.With the increase of STATCOM capacity, Its reliability is directly connected to the safe operation of power network, once open fault occurs for its power switch pipe, STATCOM can be to electricity The a large amount of harmonic waves of net injection, and can not also continue compensating power;This will easily make the increase of mains by harmonics content, voltage flicker, And the damage of region electrical equipment, cause heavy losses.Therefore, three-level STATCOM needs fast and accurately fault diagnosis Technology, to improve the reliability of its system.
In general, STATCOM main circuits power device occurs after open fault, and its output voltage and electric current can occur Distortion, by analyzing voltage x current, it is possible to judge the position of power device broken down.Such as in middle promulgated by the State Council A kind of level NPC fault of converter diagnostic methods of diode three are disclosed in the bright A of application for patent CN 105095566, it is this Method by the phase voltage signal of failure with it is normal when phase voltage signal subtract each other, obtained error voltage signal is used into small echo Conversion obtains the energy of each frequency band as fault feature vector, in this, as data sample, utilizes MATLAB softwares and LIBSVM Multi-class Classifier is set up in tool box, so as to realize that diode NPC three-level inverters intersect the fault diagnosis of bidirectional bridge.But It should be noted that carrying out fault diagnosis, it is necessary to add extra voltage using three-phase output voltage signal difference by this method Sensor, adds equipment cost;And this method does not account for the situation of load change, when load changes, easily Generation wrong diagnosis.
Another existing document also has the method for diagnosing faults for proposing to analyze in real time based on modulation voltage waveform, and it is according to tune Circuit, is divided into four operation intervals by the direction of voltage processed and correspondence phase current;When device is opened a way in different operating interval, Detailed real-time analysis is carried out to bridge arm voltage waveform, and proposes three breakdown judge foundations;Design specific fault diagnosis electricity Road, and the automatic diagnosis of device open fault is realized in level according to bridge arm voltage waveform and its change of duration.Experiment As a result show, this method can quickly distinguish the fault mode of proposition, and be pin-pointed to defective device.But this method is same Need to increase extra voltage sensor, add equipment cost.
The content of the invention
In view of the defect that prior art is present, the invention aims to provide a kind of compound event based on output current Hinder the three-level STATCOM power device open fault diagnostic method of feature extraction and SVMs, this method is using output Electric current is used as fault diagnosis signal, it is not necessary to increases extra voltage sensor, and accurately extracts fault feature vector simultaneously Improve fault diagnosis accuracy.
To achieve these goals, technical scheme:
A kind of three-level STATCOM intelligent failure diagnosis method, it is characterised in that comprise the following steps:
Step 1, based on single tube failure situation and normal operation, to three-level STATCOM main circuit power device Running status is divided, and sets corresponding label respectively to every kind of fault type;
Step 2, the corresponding three-level STATCOM simulation model of every kind of fault type is set up, respectively to each emulation mould Type is emulated and to the net side three-phase current i of every kind of fault typela、ilb、ilcSampled, to obtain every kind of fault type Net side three-phase current sample;
Step 3, using each sampling based on Wavelet Packet Energy Spectrum and average current compound characteristics extractive technique to collecting The fault signature of sample extraction frequency domain and time domain, and time domain and frequency domain fault signature by obtaining constituted after fault feature vector Dimension-reduction treatment is carried out to fault feature vector using principle component analysis, to obtain the fault signature corresponding to each fault type Vector;
Step 4, line label is entered to each sample and multi-category support vector machines model is constructed, and based on genetic algorithm The multi-category support vector machines model constructed is trained and carries out parameter optimization, optimal punishment parameter C and core is obtained Function σ;
Step 5, using the multi-category support vector machines model trained three-level STATCOM power device is opened Road fault diagnosis.
Further, as the preferred scheme of the present invention
The step 3 comprises the following steps:
Carry out three layers of WAVELET PACKET DECOMPOSITION to each sample for collecting, and extract each layer from low to high 8 respectively The coefficient sequence of individual frequency range is wavelet coefficient;
Each wavelet coefficient is reconstructed and calculates each after reconstruct by the formula of the coefficient quadratic sum shown in formula (1) Energy corresponding to frequency range, and each energy is constituted into the fault feature vector as shown in formula (2) as element
In above formula:For the gross energy corresponding to each frequency range; For the discrete point amplitude of reconstruction signal;Ej,nIt is then the energy of jth layer n node WAVELET PACKET DECOMPOSITION Number Sequences;
Unified dimension, each fault feature vector of normalized, even
Vector after normalization
The average current model as shown in formula (5) is used to extract phase to phase fault current signal each sample collected Average value as index,
Wherein ix(n) it is signal sampling value, N is sampling number.
In order to overcome fault signature to the sensitiveness of load change, it is normalized, obtains normalized flat Equal current value ux,
Wherein IxFor x phase current virtual values.
Compound spy of the band energy value that wavelet-packet energy spectrometry is extracted with normalization average current value jointly constructs newly Vector is levied, is obtained
T'h=[ua,ub,Ej,0/E,Ej,1/E,…,Ej,15/E] (8)
Fundamental factor adjustment is carried out to fault feature vector and each fault feature vector is entered successively using principle component analysis Row dimension-reduction treatment, to obtain the fault feature vector corresponding to each fault type.
Further, as the preferred scheme of the present invention
The multi-category support vector machines model uses one-to-one multi-category support vector machines model, i.e., to the event of every two class Barrier type sets up a C- SVMs.
Further, as the preferred scheme of the present invention
The multi-category support vector machines model selection Gaussian radial basis function RBF is as kernel function, i.e., from formula (9) function shown in,
Compared with prior art, beneficial effects of the present invention:
The present invention can be with real-time diagnosis three-level STATCOM power device open fault, and it is by using SVMs Method for diagnosing faults cause it need not be analyzed the fault signature of system accurately to recognize without having to system model Know, you can failure diagnostic process is completed, so as to improve the Generalization Ability of the fault diagnosis algorithm;And the method for diagnosing faults It is trained using machine learning method, the larger accuracy rate for improving fault diagnosis.
Brief description of the drawings
Fig. 1 is the corresponding flow chart of steps of the method for the invention;
Fig. 2 is three-level STATCOM circuit topological structure figure of the present invention;
Fig. 3 is the output current that a phases during open fault occur for example of the present invention;
Fig. 4 is three layers of WAVELET PACKET DECOMPOSITION schematic diagram;
Fig. 5 is the corresponding genetic algorithm fitness curve of example of the present invention.
Embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached in the embodiment of the present invention Figure, technical scheme is clearly and completely described, it is clear that described embodiment is that a part of the invention is real Apply example, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making wound The every other embodiment obtained under the premise of the property made work, belongs to the scope of protection of the invention.
The present invention proposes three level of a kind of combined failure feature extraction based on output current and SVMs STATCOM power device open fault diagnostic methods, compared with existing method for diagnosing faults, this method is only with output electricity Stream is used as fault diagnosis signal, it is not necessary to increase extra voltage sensor;In addition fault signature is extracted using based on wavelet packet The compound characteristics extractive technique of energy spectrum and average current, can more accurately extract fault feature vector;Simultaneously using heredity Algorithm carries out optimizing to SVMs parameter, further increases fault diagnosis accuracy.
Specifically, as shown in figure 1, the method for the invention comprises the following steps:
Step 1, based on single tube open fault condition and normal operation, to three-level STATCOM main circuit power The running status of device is divided, and sets corresponding label respectively to every kind of fault type;Specifically, such as Fig. 2 institutes Show, S is expressed as because three-level STATCOM main circuit power device circuit includes 12 power tube IGBT1、 S2、……S12, it is considered to the situation of single tube open fault, 12 kinds of fault types can be divided into altogether, while by normal operation As the special fault type of a class, then STATCOM running status is divided into 13 kinds of situations, as shown in table 1 respectively to every One class carries out fault type label.
Fault type Fault-free S1Open circuit S2Open circuit S3Open circuit S4Open circuit S5Open circuit S6Open circuit
Label 0 1 2 3 4 5 6
Fault type S7Open circuit S8Open circuit S9Open circuit S10Open circuit S11Open circuit S12Open circuit
Label 7 8 9 10 11 12
Table 1
Step 2, in MATLAB three corresponding level of every kind of fault type are set up respectively to above-mentioned 13 kinds of fault types STATCOM simulation models, are emulated to each simulation model and to the net side three-phase current i of every kind of fault type respectivelyla、 ilb、ilcSampled, to obtain the net side three-phase current sample of every kind of fault type;Specifically, real as shown in Figure 3 Example, with Sa1Exemplified by the output current of three-phase during generation open fault, its outlet side three-phase electricity is gathered by current sensor Stream, and using sample frequency as 1kHz, 20ms obtains data as collecting sample by a cycle.Further, load is distinguished It is set as 20 kinds of situations:RlExcursion is 4 Ω~10 Ω, LlExcursion is 0.03H~0.06H;Every kind of fault type difference 20 samples are gathered, every kind of loading condition is used as a sample;And cause 10 conducts in above-mentioned 20 samples Training sample, 10 are used as test sample.
Step 3, using each sampling based on Wavelet Packet Energy Spectrum and average current compound characteristics extractive technique to collecting The fault signature of sample extraction frequency domain and time domain, and time domain and frequency domain fault signature by obtaining constituted after fault feature vector Dimension-reduction treatment is carried out to fault feature vector using principle component analysis, to obtain the fault signature corresponding to each fault type Vector;Specifically, the step 3 includes:
As shown in figure 4, carry out three layers of WAVELET PACKET DECOMPOSITION to each sample for collecting, and extract respectively each layer from The coefficient sequence of low frequency to 8 frequency ranges of high frequency is wavelet coefficient;
Resulting each wavelet coefficient is reconstructed and weight is calculated by the formula of the coefficient quadratic sum shown in formula (1) Each frequency range or the energy corresponding to frequency band after structure, and each energy is constituted into the failure spy as shown in formula (2) as element Levy vector
In above formula:For the gross energy corresponding to each frequency range; For the discrete point amplitude of reconstruction signal;Ej,nIt is then the energy of jth layer n node WAVELET PACKET DECOMPOSITION Number Sequences;
Unified dimension, each fault feature vector of normalized, even
Vector after normalization
The average current model as shown in formula (5) is used to extract phase to phase fault current signal each sample collected Average value as index,
Wherein ix(n) it is signal sampling value, N is sampling number.
In order to overcome fault signature to the sensitiveness of load change, it is normalized, obtains normalized flat Equal current value ux,
Wherein IxFor x phase current virtual values.
Compound spy of the band energy value that wavelet-packet energy spectrometry is extracted with normalization average current value jointly constructs newly Vector is levied, is obtained
T'h=[ua,ub,Ej,0/E,Ej,1/E,…,Ej,15/E] (8)
Then the dimension of the characteristic vector of each fault type is 18, wherein load is Rl=7 Ω, LlDuring=0.04H, Sa1The characteristic vector of open fault is as shown in table 2.
Characteristic vector 1 2 3 4 5 6 7 8 9
-0.2006 0.0239 99.9815 0.0103 0.0024 0.0039 0.0001 0.0001 0.0013
Characteristic vector 10 11 12 13 14 15 16 17 18
0.0004 99.9892 0.0062 0.0015 0.0021 0.0001 0.0004 0.0006 0.0002
Table 2
And dimension-reduction treatment is carried out to each fault feature vector using principle component analysis successively, to obtain each fault type Corresponding fault feature vector.The specific variance contribution ratio that defines is the S after 0.99 dimensionality reduction1Open fault characteristic vector For:
TS1=[- 0.0127 0.0014]
Other fault feature vectors are subjected to dimension-reduction treatment after the same method, the characteristic vector sample of each failure is obtained This.
Step 4, line label is entered to each sample and multi-category support vector machines model is constructed, and based on genetic algorithm The multi-category support vector machines model constructed is trained and carries out parameter optimization, due to punishment parameter C and kernel function σ It is to influence the principal element of SVMs performance, it is therefore desirable to obtain optimal punishment parameter C and kernel function σ;Specifically may be used Using one-to-one multi-category support vector machines, i.e., a C- SVMs is set up to every two classes fault type;It is (high from RBF This RBF) as kernel function, its as shown in formula (9),
Wherein x2For kernel function center, σ is the width parameter of function, the radial effect scope for control function.Specifically , as shown in figure 5, obtaining the optimized parameter of SVMs using genetic algorithm:It is 200, population quantity to set and terminate algebraically Pop is 20, and it obtains optimal parameter c=6.2769, σ=42.5435, and the accuracy rate of 5 folding cross validations and is 98.4615%.
Step 5, using the multi-category support vector machines model trained three-level STATCOM power device is opened Road fault diagnosis.Specifically, experiment confirms to test the SVM classifier set up using 130 groups of test samples, classify Accuracy rate be 96.1538%.
The foregoing is only a preferred embodiment of the present invention, but protection scope of the present invention be not limited to This, any one skilled in the art the invention discloses technical scope in, technique according to the invention scheme And its inventive concept is subject to equivalent substitution or change, it should all be included within the scope of the present invention.

Claims (4)

1. a kind of three-level STATCOM intelligent failure diagnosis method, it is characterised in that comprise the following steps:
Step 1, based on single tube failure situation and normal operation, the operation to three-level STATCOM main circuit power device State is divided, and sets corresponding label respectively to every kind of fault type;
Step 2, the corresponding three-level STATCOM simulation model of every kind of fault type is set up, each simulation model is entered respectively Row is emulated and to the net side three-phase current i of every kind of fault typela、ilb、ilcSampled, to obtain the net of every kind of fault type Side three-phase current sample;
Step 3, using based on Wavelet Packet Energy Spectrum and average current compound characteristics extractive technique to each sample for collecting Extract the fault signature of frequency domain and time domain, and the time domain and frequency domain fault signature by obtaining are constituted after fault feature vector using master Meta analysis method carries out dimension-reduction treatment to fault feature vector, to obtain the fault feature vector corresponding to each fault type;
Step 4, line label is entered to each sample and multi-category support vector machines model is constructed, and based on genetic algorithm to institute's structure The multi-category support vector machines model made, which is trained, carries out parameter optimization, obtains optimal punishment parameter C and kernel function σ;
Step 5, using the multi-category support vector machines model trained to three-level STATCOM power device carry out open fault Diagnosis.
2. according to the method described in claim 1, it is characterised in that:
The step 3 comprises the following steps:
Three layers of WAVELET PACKET DECOMPOSITION are carried out to each sample collected, and extract each layer 8 frequency from low to high respectively The coefficient sequence of section is wavelet coefficient;
Each wavelet coefficient is reconstructed and each frequency range institute after reconstruct is calculated by the formula of the coefficient quadratic sum shown in formula (1) Corresponding energy, and each energy is constituted into the fault feature vector as shown in formula (2) as element
In above formula:For the gross energy corresponding to each frequency range;(n=1,2 ..., 2j-1;K=1,2 ..., N) believe for reconstruct Number discrete point amplitude;Ej,nIt is then the energy of jth layer n node WAVELET PACKET DECOMPOSITION Number Sequences;
Unified dimension, each fault feature vector of normalized, even
Vector after normalization
The average current model as shown in formula (5) is used to extract being averaged for phase to phase fault current signal each sample collected It is worth as index,
Wherein ix(n) it is signal sampling value, N is sampling number.
In order to overcome fault signature to the sensitiveness of load change, it is normalized, normalized average electricity is obtained Flow valuve ux,
Wherein IxFor x phase current virtual values.
Band energy value that wavelet-packet energy spectrometry the is extracted compound characteristics new with normalization average current value jointly constructs to Amount, is obtained
T′h=[ua,ub,Ej,0/E,Ej,1/E,…,Ej,15/E] (8)
Fundamental factor adjustment is carried out to fault feature vector and each fault feature vector is dropped successively using principle component analysis Dimension processing, to obtain the fault feature vector corresponding to each fault type.
3. according to the method described in claim 1, it is characterised in that:
The multi-category support vector machines model uses one-to-one multi-category support vector machines model, i.e., to every two classes fault type Set up a C- SVMs.
4. according to the method described in claim 1, it is characterised in that:
The multi-category support vector machines model selection Gaussian radial basis function RBF is as kernel function, i.e., from formula (9) Suo Shi Function,
CN201710119870.4A 2017-03-01 2017-03-01 A kind of three-level STATCOM intelligent failure diagnosis method Expired - Fee Related CN107132443B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710119870.4A CN107132443B (en) 2017-03-01 2017-03-01 A kind of three-level STATCOM intelligent failure diagnosis method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710119870.4A CN107132443B (en) 2017-03-01 2017-03-01 A kind of three-level STATCOM intelligent failure diagnosis method

Publications (2)

Publication Number Publication Date
CN107132443A true CN107132443A (en) 2017-09-05
CN107132443B CN107132443B (en) 2019-07-12

Family

ID=59721669

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710119870.4A Expired - Fee Related CN107132443B (en) 2017-03-01 2017-03-01 A kind of three-level STATCOM intelligent failure diagnosis method

Country Status (1)

Country Link
CN (1) CN107132443B (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107992665A (en) * 2017-11-27 2018-05-04 国家电网公司 A kind of ultra-high voltage converter station alternating current filter on-line fault diagnosis analysis method
CN108628185A (en) * 2018-06-26 2018-10-09 上海海事大学 Five-electrical level inverter fault diagnosis and fault-tolerant control method and semi-physical emulation platform
CN109541498A (en) * 2018-11-30 2019-03-29 苏州数言信息技术有限公司 A kind of general lamp failure intelligent detecting method and system
CN110618394A (en) * 2019-10-31 2019-12-27 武汉大学 Fault diagnosis method for photovoltaic microgrid direct current and alternating current converter power supply based on current average value
CN111291783A (en) * 2020-01-15 2020-06-16 北京市燃气集团有限责任公司 Intelligent fault diagnosis method, system, terminal and storage medium for gas pressure regulating equipment
CN111948487A (en) * 2020-07-17 2020-11-17 国网上海市电力公司 High-voltage power equipment fault diagnosis method and system based on artificial intelligence
CN112560328A (en) * 2020-11-18 2021-03-26 电子科技大学 IGBT bonding lead fault diagnosis method based on surface micro-strain signal
CN115169405B (en) * 2022-07-14 2024-02-02 北京威控科技股份有限公司 Hotel guest room equipment fault diagnosis method and system based on support vector machine

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104267296A (en) * 2014-10-21 2015-01-07 国家电网公司 STATCOM fault diagnosis method based on MMC
CN204442019U (en) * 2015-03-18 2015-07-01 威凡智能电气高科技有限公司 A kind of long distance control system based on high pressure STATCOM
CN105515023A (en) * 2014-10-20 2016-04-20 南京工业大学 Novel cascade STATCOM converter unit fault-tolerant method
CN105891671A (en) * 2016-04-01 2016-08-24 昆明理工大学 STATCOM-contained line single-end traveling wave distance measurement method based on fault traveling wave line distribution characteristics
CN106019117A (en) * 2016-08-03 2016-10-12 盐城工学院 Open-circuit fault diagnosis method of cascaded STATCOM system IGBT
CN106370968A (en) * 2016-10-10 2017-02-01 中国矿业大学 IGBT open circuit fault positioning method for three-phase four-wire system three-leg three-level SVG

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105515023A (en) * 2014-10-20 2016-04-20 南京工业大学 Novel cascade STATCOM converter unit fault-tolerant method
CN104267296A (en) * 2014-10-21 2015-01-07 国家电网公司 STATCOM fault diagnosis method based on MMC
CN204442019U (en) * 2015-03-18 2015-07-01 威凡智能电气高科技有限公司 A kind of long distance control system based on high pressure STATCOM
CN105891671A (en) * 2016-04-01 2016-08-24 昆明理工大学 STATCOM-contained line single-end traveling wave distance measurement method based on fault traveling wave line distribution characteristics
CN106019117A (en) * 2016-08-03 2016-10-12 盐城工学院 Open-circuit fault diagnosis method of cascaded STATCOM system IGBT
CN106370968A (en) * 2016-10-10 2017-02-01 中国矿业大学 IGBT open circuit fault positioning method for three-phase four-wire system three-leg three-level SVG

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
杨晓冬 等: "静止同步补偿器IGBT开路故障诊断方法", 《电工技术学报》 *
谢小荣 等: "± 20 Mvar STATCOM的运行监测与故障诊断系统", 《电力系统自动化》 *
陈雷 等: "基于电流检测的STATCOM变流器开路故障诊断", 《电力电容器与无功补偿》 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107992665A (en) * 2017-11-27 2018-05-04 国家电网公司 A kind of ultra-high voltage converter station alternating current filter on-line fault diagnosis analysis method
CN108628185A (en) * 2018-06-26 2018-10-09 上海海事大学 Five-electrical level inverter fault diagnosis and fault-tolerant control method and semi-physical emulation platform
CN109541498A (en) * 2018-11-30 2019-03-29 苏州数言信息技术有限公司 A kind of general lamp failure intelligent detecting method and system
CN110618394A (en) * 2019-10-31 2019-12-27 武汉大学 Fault diagnosis method for photovoltaic microgrid direct current and alternating current converter power supply based on current average value
CN111291783A (en) * 2020-01-15 2020-06-16 北京市燃气集团有限责任公司 Intelligent fault diagnosis method, system, terminal and storage medium for gas pressure regulating equipment
CN111948487A (en) * 2020-07-17 2020-11-17 国网上海市电力公司 High-voltage power equipment fault diagnosis method and system based on artificial intelligence
CN112560328A (en) * 2020-11-18 2021-03-26 电子科技大学 IGBT bonding lead fault diagnosis method based on surface micro-strain signal
CN112560328B (en) * 2020-11-18 2022-04-19 电子科技大学 IGBT bonding lead fault diagnosis method based on surface micro-strain signal
CN115169405B (en) * 2022-07-14 2024-02-02 北京威控科技股份有限公司 Hotel guest room equipment fault diagnosis method and system based on support vector machine

Also Published As

Publication number Publication date
CN107132443B (en) 2019-07-12

Similar Documents

Publication Publication Date Title
CN107132443A (en) A kind of three-level STATCOM intelligent failure diagnosis method
CN109307824B (en) Clustering-based power distribution network single-phase earth fault section positioning method
CN104977502B (en) A kind of extra high voltage direct current transmission line internal fault external fault recognition methods
CN104698397B (en) A kind of multi-electrical level inverter method for diagnosing faults
CN109977920A (en) Fault Diagnosis of Hydro-generator Set method based on time-frequency spectrum and convolutional neural networks
CN103761372A (en) Multilevel inverter fault diagnosis strategy based on principal component analysis and multi-classification related vector machine(PCA-mRVM)
CN107167702A (en) A kind of distribution feeder fault type recognition method and device
CN102095996B (en) Malfunction phase selecting method of high-voltage transmission circuit based on natural frequency of travelling wave
CN108663602A (en) Flexible direct current power distribution network monopole failure line selection and Section Location and system
CN113219300B (en) Power distribution network single-phase earth fault sensing method based on phase current transient state steady state
CN109782091A (en) Multi-electrical level inverter method for diagnosing faults based on depth small echo extreme learning machine
CN112803377B (en) Single-ended electric quantity protection method suitable for hybrid bipolar direct current transmission line
CN103018632B (en) Small current grounding system single-phase ground fault line selection method based on fisher information
CN106093678B (en) A method of quick and precisely diagnosing flexible HVDC transmission system converter fault
CN112462193B (en) Automatic reclosing judgment method for power distribution network based on real-time fault filtering data
Song et al. A negative selection algorithm-based identification framework for distribution network faults with high resistance
CN111308260B (en) Electric energy quality monitoring and electric appliance fault analysis system based on wavelet neural network and working method thereof
CN104967097B (en) Excitation flow recognition method based on support vector classification
CN106019173A (en) Real-time fault detection method applied to voltage source converter
CN115329816A (en) Multi-level inverter fault diagnosis method based on simulation data migration learning
CN110068776A (en) Three-level inverter open-circuit fault diagnostic method based on Support Vector Machines Optimized
CN112255495B (en) Micro-grid high-resistance fault detection method
CN106291239A (en) A kind of direct current transmission line fault recognition methods using filter branches electric current and principal component analytical method
CN115128400A (en) Distribution network fault type identification and fault route selection comprehensive research and judgment method and system
CN112363009B (en) Single-ended fault location method and system for same-tower line ground fault

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

Granted publication date: 20190712

Termination date: 20210301

CF01 Termination of patent right due to non-payment of annual fee