CN105372528B - A kind of state maintenance method of Power Transformer Internal Faults and New Transformer - Google Patents
A kind of state maintenance method of Power Transformer Internal Faults and New Transformer Download PDFInfo
- Publication number
- CN105372528B CN105372528B CN201510821187.6A CN201510821187A CN105372528B CN 105372528 B CN105372528 B CN 105372528B CN 201510821187 A CN201510821187 A CN 201510821187A CN 105372528 B CN105372528 B CN 105372528B
- Authority
- CN
- China
- Prior art keywords
- transformer
- state
- signal
- power transformer
- symbol
- 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
Links
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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a kind of state maintenance methods of Power Transformer Internal Faults and New Transformer, the internal fault model of transformer is emulated, set up the assessment models of the power transformer operation conditions based on symbolic dynamics, using power transformer interior fault electric current as main study subject, a kind of novel Power Transformer Condition of research overhauls model.The model is using the inside transformer electric current that monitors as input, according to the comparison of the real time data of acquisition and historical data, take transformer factory data into consideration, historical test data, assessment is made to the operation of transformer current state, judge transformer whether in failure early stage state and the type of failure and the degree of failure.Target is that the Condition Maintenance Method of Transformer model is applied to the process of virtual condition maintenance, cost-effective to achieve the purpose that, or the repair based on condition of component model is used for integrated system, reaches the target of optimization final result.
Description
Technical field
The invention belongs to computer software technical field, more particularly to a kind of electric system core equipment internal fault identification
Software application technology, in particular to a kind of state maintenance method of Power Transformer Internal Faults and New Transformer.
Background technology
Transformer as electric system pick-up and distribution electric energy core equipment, safety and stability to close weight
It wants.Therefore, it is essential to carry out maintenance to transformer.Traditional maintenance mode is periodic inspection, as Repair of Transformer
Usual way, preventive trial and periodic inspection bear the character of much blindness with it is mandatory, cause " to cross and repair, owe to repair, blindly tie up
Repair, the wasting of resources " a series of problems, such as, this also just highlights the necessity of repair based on condition of component.Repair based on condition of component is the fortune to transformer
Market condition carries out assessment in real time and obtains state index, and a kind of maintenance side overhauled to transformer according to state index
Formula;State index refers to being detected the operating parameter of transformer with various methods in the prior art, such as:Transformer temperature, oil
Middle gas ratio, degree of discharge etc..The foundation of Power Transformer Internal Faults and New Transformer model is by being closed to inside transformer magnetic linkage
The research of system builds the simulation model of power transformer using MATLA softwares, realizes substantially at present.
Invention content
It is an object of the present invention to which the method for distinguishing sequence signal is studied to judge whether signal fault is deposited
And failure degree, to overcome the deficiencies in the prior art.
The invention is realized by the following technical scheme, by the Power Transformer Internal Faults and New Transformer model of foundation, in emulation
Portion's electric current is main study subject, is compared to electric current and running current with symbolic dynamics, it is established that signal difference
Different mechanism is online monitored transformer and is diagnosed to be the operating status of transformer at this time, is specifically divided into health status, mistake
State, malfunction;To study a kind of novel Power Transformer Condition maintenance model based on state.
Including establishing Power Transformer Internal Faults and New Transformer model, the current or voltage data run under failure are then obtained.It should
Transformer Model and unconventional equivalent operation circuit model, but the abnormal operating state that be directed to transformer is established
The fault model of transient state, which can be emulated for different degrees of malfunction, in order to obtain higher emulation
The efficiency of model.
Symbolic dynamics method is applied to signal processing, with the state of the sequence description nonlinear system after symbolism
Thought is applied in the description to signal, and symbolic dynamics are as a kind of description to dynamical system coarse, to consecutive hours
Between signal carry out phase space division, phase space is indicated with limited symbol.
The signal phase space divides, and the phase space for studying signal waveform divides and makes to obtain with reference to maximum entropy partitioning
Serializing signal possess the maximum information content of original data signal.
The principle based on symbolic dynamics, using advantage of the symbolic dynamics in terms of signal processing, further
Sequence is analyzed, and corresponding Markov state transfer matrix is established by symbol sebolic addressing, then uses matrix Euclidean distance
Signal difference is characterized, thus the fast and effeciently presence of failure judgement.
The state maintenance method of this Power Transformer Internal Faults and New Transformer provided by the invention can quickly and effectively identify change
The presence of depressor internal fault reduces simultaneously so that service personnel can more rapidly and accurately find out failure weak element
Introduce the extra cost of the equipment installation maintenances such as related sensor maintenance.
Description of the drawings
Fig. 1 is principle flow chart.
Fig. 2 is maximum entropy partitioning schematic diagram.
Fig. 3 is to improve maximum entropy algorithm flow chart.
Fig. 4 is markoff process schematic diagram.
Fig. 5 is Markov matrix schematic diagram.
Specific implementation mode
1 to 5 embodiments of the present invention is further illustrated below in conjunction with the accompanying drawings, and Power Transformer Internal Faults and New Transformer model is built
Vertical is that the simulation model of power transformer is built using MATLA softwares by the research to inside transformer magnetic linkage relationship.It is imitative
Repair based on condition of component in true mode is to carry out assessment in real time to the operating condition of transformer to obtain state index, refers to that electric current is believed
Number processing after obtained matrix distance size;A series of emulation examination is carried out to the Power Transformer Faults model set up
Test, obtain with the relevant fault test index of transformer and data, the index and data refer to fault degree size and
Current data;Play an assessment models by these Index Establishments again, assessment models refer to method used in the present invention and
Value range, and then the online operation conditions for judging transformer.
The present inventor's created symbol dynamic method, the target component of monitoring can only be the internal operating currents of transformer
Or voltage, additional setting sensor is not needed, while also having abundant historical data availability test and use.Using
Symbolic dynamics method can quickly and effectively judge the presence of power transformer interior fault, therefore can explore a kind of more existing side
The more acurrate more efficient Condition Maintenance Method of Transformer mechanism of method.
The present invention includes carrying out repair based on condition of component to the Power Transformer Internal Faults and New Transformer model of foundation, in the fault model
Repair based on condition of component is to carry out assessment in real time to the operating condition of transformer to show that state index, the state index refer to electric current letter
Number processing after obtained matrix distance size;
L-G simulation test is carried out to the Power Transformer Internal Faults and New Transformer model set up, is obtained relevant fault with transformer
Test index and data, the index and data refer to fault degree size and current data;
Assessment models are set up by index and data again, the assessment models refer to that with model internal current be main
Research object forms signal by improved maximum entropy algorithm, and symbolic dynamics is applied to signal processing, passes through symbol sequence
Row establish corresponding Markov state transfer matrix, then matrix Euclidean distance are used to characterize signal difference, pass through signal difference
Compare, and the assessment models of power transformer operation conditions is combined to carry out the value range that state is drafted, and then judges transformation online
The operation conditions of device, thus the fast and effeciently presence of failure judgement.
The present invention includes the following steps:
Step 1, according to Power Transformer Internal Faults and New Transformer model, the internal fault for simulating transformer in transient state is electric
Stream, gets the current data signal under different degrees of failure;
Step 2, using model internal current as main study subject, signal is formed by improved maximum entropy algorithm,
The improved maximum entropy algorithm includes maximum entropy partitioning and first-order difference partitioning;
Step 3, symbolic dynamics are applied to signal processing, electric current is realized by the sequence that improved maximum entropy algorithm obtains
It is compared with the electric current of normal work;
Step 4, corresponding Markov state transfer matrix is established by symbol sebolic addressing, then uses matrix Euclidean distance table
Signal difference is levied, is compared by signal difference, and the assessment models of power transformer operation conditions is combined to carry out what state was drafted
Value range, and then the online operation conditions for judging transformer, thus the fast and effeciently presence of failure judgement;
The maximum entropy algorithm of the step 2 is, if being with the size of alphabet | A |, signal length N is then based on maximum
The algorithmic procedure of the partitioning of entropy is as follows,
I initiation parameters, if k=1, and a threshold parameter ε is chosen, 0 ε≤1 <, threshold parameter determines algorithm
Termination condition;
II is by the signal that ascending order spread length is N;
III is by each continuous length in signalA signal constitutes an independent letter member of the secondary division
Element;
The division and alphabet preliminary symbol original signal that IV is obtained in being walked according to III, fall according to the value of signal
Section in III step, or represented equal to the minimum value in the section, and by the corresponding letter in the value section of signal;
V calculates the Probability p that symbol occursi, i=1,2 ... k;
VI calculates corresponding H (k) at this time and h (k)=H (k)-H (k-1);
If the h (k) that VII is calculated is less than the thresholding ε of setting, terminate algorithm, is otherwise incremented by 1 for K, and jump to
III walks.
The narration of maximum entropy division principle can see from above, divided by maximum entropy, the variation fierceness of signal
Region has used more symbolic formulations, but does not still reflect the front and back relationship changed of signal.Therefore, base herein
On plinth, in conjunction with first difference method, using calculus of finite differences can before and after reaction signal variation relation advantage, further improve this stroke
Point-score.
The improved maximum entropy algorithm of the step 2, that is, include the division of maximum entropy partitioning and first-order difference partitioning
Rule is as follows:
X3,X2,X1Represent the minimum value in each section divided by III step;
By the combination of above-mentioned two groups of formula, division is made to substantially section using maximum entropy division, then inside section
It is further divided according to signal context using first difference method, can be obtained by symbol in this way as 0 to 7 based on most
The signal code method of big entropy and first-order difference.
It is described corresponding Markov state transfer matrix is established by symbol sebolic addressing to include:
If symbol sebolic addressing { St, symbols alphabet size is k, a length of D of sliding window, from first symbol of symbol sebolic addressing one by one
Number letter starts, and with a length of D of every window to right translation, often slips over a symbol letter, retains the rear (D- of preceding state successively
1) a symbol is alphabetical and adds new symbol letter one new state of generation;
I sets symbol sebolic addressing, and the symbol of next appearance is related to D moment symbol before, then this process is referred to as D ranks
Markoff process;
I.e.:P(Si|Si-1...Si- D...)=P (Si|Si-1...Si- D),
Setting alphabet is A, and size is | A |, then all shapes being likely to occur of markoff process that this sequence obtains
State quantity should be:|A|D;
II defines piAnd pjEquivalent state respectively in system, i-th of state in state set Q are transferred to j-th of shape
Probability of state is:
πij=P (S ∈ A | qj∈Q,(qi,S)→qj),
Transition probabilities of the matrix π for stateful.
Transition probability can use a matrix description, i.e. Markov square between obtaining all states by this sequence
Battle array.If the alphabet A of symbol includes only two letters { 0,1 }, next state and two before in markoff process
Symbol is related, and simplest markoff process can be expressed as shown in Figure 5
And Markov matrix should have:
∑jP (i, j)=1
It is 1 that i.e. matrix, which meets the sum of all elements of arbitrary a line,.
The signal difference compares,
It is described to be compared by signal difference, and the assessment models of power transformer operation conditions is combined to carry out what state was drafted
Value range, and then judge that the operation conditions of transformer includes online:
The assessment models that set carry out value range that state is drafted as Nmin and Nmax,
Matrix Euclidean distance:
d<Nmin, then indication transformer be in health status;
Nmax>d>Nmin, then indication transformer be in error condition;
d>Nmax, then indication transformer be in malfunction.
Euclidean distance refers between matrix:The set that matrix is regarded as to n vector, what is obtained is between n vectorial
Euclidean distance.Then
The assessment models of the power transformer operation conditions carry out state and draft, and are taken based on symbolic dynamics judgement
It is obtaining as a result, in conjunction with actual transformer working condition and historical data come to result carry out different conditions classification;Including
The mode of different symbolism sequences is obtained based on the serializing that phase space divides.The signal phase space divides, research letter
The phase space of number waveform divide and the serializing signal that makes with reference to maximum entropy partitioning to possess original data signal maximum
Information content.
In the experiment of emulation, find the failure of transformer with the increase of decaying, the difference of fault-signal and original signal
Exponentially is increased, is then tended towards stability again, this illustrates that the change procedure of failure is mutated faster there are one the relatively narrow rate in section
Process, drafting for state will can be respectively drafted before, during and after section as different states to take by finding the section
Different maintenance policies.Specific embodiment is expressed as:
1) if Markov state transfer matrix Euclidean distance is less than 1.5, then it represents that transformer is in health status, is not necessarily to
Transformer is overhauled.
2) if Markov state transfer matrix Euclidean distance is more than 1.5 and is less than 2.5, then it represents that transformer is in mistake
State need to be overhauled according to plan.
3) if Markov state transfer matrix Euclidean distance is more than 2.5, then it represents that transformer is in malfunction, need to stand
Quarter is overhauled.
One embodiment of the present invention above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
Cannot the limitation to the scope of the claims of the present invention therefore be interpreted as.It should be pointed out that for those of ordinary skill in the art
For, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to the guarantor of the present invention
Protect range.Therefore, the protection domain of patent of the present invention should be determined by the appended claims.
Claims (6)
1. a kind of state maintenance method of Power Transformer Internal Faults and New Transformer,
Include the Power Transformer Internal Faults and New Transformer model progress repair based on condition of component to foundation, it is characterized in that be,
Repair based on condition of component in the fault model is to carry out assessment in real time to the operating condition of transformer to obtain state index, institute
It refers to the obtained matrix distance size after current signal processing to state state index;
L-G simulation test is carried out to the Power Transformer Internal Faults and New Transformer model set up, is obtained and the relevant fault experiment of transformer
Index and data, the index and data refer to fault degree size and current data;
Assessment models are set up by index and data again, the assessment models refer to that with model internal current be main research
Object forms signal by improved maximum entropy algorithm, and symbolic dynamics is applied to signal processing, is built by symbol sebolic addressing
Corresponding Markov state transfer matrix is found, then matrix Euclidean distance is used to characterize signal difference, is compared by signal difference,
And the assessment models of power transformer operation conditions is combined to carry out the value range that state is drafted, and then the online fortune for judging transformer
Row situation, thus the fast and effeciently presence of failure judgement.
2. a kind of state maintenance method of Power Transformer Internal Faults and New Transformer according to claim 1, it is characterised in that including
Following steps:
Step 1, according to Power Transformer Internal Faults and New Transformer model, simulate transformer in the case that transient state internal fault current,
Get the current data signal under different degrees of failure;
Step 2, using model internal current as main study subject, signal is formed by improved maximum entropy algorithm,
The improved maximum entropy algorithm includes maximum entropy partitioning and first-order difference partitioning;
Step 3, symbolic dynamics are applied to signal processing, the sequence obtained by improved maximum entropy algorithm realize electric current with just
The electric current often to work is compared;
Step 4, corresponding Markov state transfer matrix is established by symbol sebolic addressing, then uses matrix Euclidean distance characterization letter
Number difference, is compared by signal difference, and the assessment models of power transformer operation conditions is combined to carry out the range that state is drafted
Value, and then the online operation conditions for judging transformer, thus the fast and effeciently presence of failure judgement.
3. a kind of state maintenance method of Power Transformer Internal Faults and New Transformer according to claim 2, which is characterized in that
The maximum entropy algorithm of the step 2 is, if being with the size of alphabet | A |, signal length N, then based on maximum entropy
The algorithmic procedure of partitioning is as follows,
I initiation parameters, if k=1, and a threshold parameter ε is chosen, 0 ε≤1 <, threshold parameter determines the end of algorithm
Condition;
II is by the signal that ascending order spread length is N;
III is by each continuous length in signalA signal constitutes an independent alphabetical element of the secondary division;
Obtained division and alphabet preliminary symbol original signal during IV is walked according to III are fallen according to the value of signal the
Section in III step, or represented equal to the minimum value in the section, and by the corresponding letter in the value section of signal;
V calculates the Probability p that symbol occursi, i=1,2 ... k;
VI calculates corresponding H (k) at this time and h (k)=H (k)-H (k-1);
If the h (k) that VII is calculated is less than the thresholding ε of setting, terminate algorithm, is otherwise incremented by 1 for K, and jump to III
Step;
The improved maximum entropy algorithm of the step 2, that is, include the division rule of maximum entropy partitioning and first-order difference partitioning
It is as follows:
X3,X2,X1Represent the minimum value in each section divided by III step;
By the combination of above-mentioned two groups of formula, division is made to substantially section using maximum entropy division, then used inside section
First difference method is further divided according to signal context, can be obtained by symbol in this way as 0 to 7 based on maximum entropy
With the signal code method of first-order difference.
4. a kind of state maintenance method of Power Transformer Internal Faults and New Transformer according to claim 2, which is characterized in that described
Establishing corresponding Markov state transfer matrix by symbol sebolic addressing includes:
If symbol sebolic addressing { St, symbols alphabet size is k, a length of D of sliding window, from first symbol word of symbol sebolic addressing one by one
Mother starts, and with a length of D of every window to right translation, often slips over a symbol letter, and rear (D-1) for retaining preceding state successively is a
Symbol letter simultaneously adds new symbol letter one new state of generation;
I sets symbol sebolic addressing, and the symbol of next appearance is related to D moment symbol before, then this process is referred to as D ranks Ma Er
Section's husband's process;
I.e.:P(Si|Si-1...Si- D...)=P (Si|Si-1...Si- D),
Setting alphabet is A, and size is | A |, then all status numbers being likely to occur of markoff process that this sequence obtains
Measuring be:|A|D;
II defines piAnd pjEquivalent state respectively in system, i-th of state in state set Q are transferred to j-th state
Probability is:
πij=P (S ∈ A | qj∈Q,(qi,S)→qj),
Transition probabilities of the matrix π for stateful.
5. a kind of state maintenance method of Power Transformer Internal Faults and New Transformer according to claim 2, which is characterized in that
It is described to be compared by signal difference, and the assessment models of power transformer operation conditions is combined to carry out the range that state is drafted
Value, and then judge that the operation conditions of transformer includes online:
The assessment models that set carry out value range that state is drafted as Nmin and Nmax,
According to matrix Euclidean distance:
Then d<Nmin, then indication transformer be in health status;
Then Nmax>d>Nmin, then indication transformer be in error condition;
Then d>Nmax, then indication transformer be in malfunction.
6. a kind of state maintenance method of Power Transformer Internal Faults and New Transformer according to claim 1, which is characterized in that described
Power transformer operation conditions assessment models carry out state draft, be taken based on it is that symbolic dynamics judge as a result,
To carry out result the classification of different conditions in conjunction with actual transformer working condition and historical data;It is drawn including being based on phase space
The serializing divided obtains the mode of different symbolism sequences.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510821187.6A CN105372528B (en) | 2015-11-24 | 2015-11-24 | A kind of state maintenance method of Power Transformer Internal Faults and New Transformer |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510821187.6A CN105372528B (en) | 2015-11-24 | 2015-11-24 | A kind of state maintenance method of Power Transformer Internal Faults and New Transformer |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105372528A CN105372528A (en) | 2016-03-02 |
CN105372528B true CN105372528B (en) | 2018-10-09 |
Family
ID=55374890
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510821187.6A Expired - Fee Related CN105372528B (en) | 2015-11-24 | 2015-11-24 | A kind of state maintenance method of Power Transformer Internal Faults and New Transformer |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105372528B (en) |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106569052B (en) * | 2016-10-11 | 2017-12-15 | 国网湖北省电力公司 | Consider the power transformer reliability estimation method of real time health state |
CN107063331A (en) * | 2016-11-09 | 2017-08-18 | 贵州电网有限责任公司凯里供电局 | A kind of power transformer interior fault detecting system based on microrobot |
CN107292512B (en) * | 2017-06-20 | 2020-09-15 | 中国电力科学研究院 | Power equipment space-time multi-dimensional safety assessment method based on symbolic dynamics and hidden Markov model |
CN108037378B (en) * | 2017-10-26 | 2020-08-07 | 上海交通大学 | Transformer operation state prediction method and system based on long-time and short-time memory network |
CN107991097A (en) * | 2017-11-16 | 2018-05-04 | 西北工业大学 | A kind of Method for Bearing Fault Diagnosis based on multiple dimensioned symbolic dynamics entropy |
CN111738286A (en) * | 2020-03-17 | 2020-10-02 | 北京京东乾石科技有限公司 | Fault determination and model training method, device, equipment and storage medium thereof |
CN111308336A (en) * | 2020-03-24 | 2020-06-19 | 广西电网有限责任公司电力科学研究院 | High-voltage circuit breaker fast overhaul method and device based on big data |
CN112327084B (en) * | 2020-11-03 | 2022-01-21 | 华北电力大学 | Method and system for detecting vibration and sound of running state of transformer by utilizing equidistant transformation |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20100029338A (en) * | 2008-09-08 | 2010-03-17 | 하주영 | System for diagnostication of transformer using ultrasonic wave |
CN102779230A (en) * | 2012-06-14 | 2012-11-14 | 华南理工大学 | State analysis and maintenance decision judging method of power transformer system |
CN103091603A (en) * | 2013-01-14 | 2013-05-08 | 华北电力大学 | Breakdown intelligent classification and positioning method of electric transmission line |
CN103245911A (en) * | 2013-05-03 | 2013-08-14 | 云南电力试验研究院(集团)有限公司电力研究院 | Breaker fault diagnosis method based on Bayesian network |
WO2014072444A1 (en) * | 2012-11-09 | 2014-05-15 | Bombardier Transportation Gmbh | Method and device for monitoring a transformer state |
CN104007343A (en) * | 2014-05-23 | 2014-08-27 | 清华大学 | Dynamic comprehensive transformer fault diagnosis method based on Bayesian network |
CN104036131A (en) * | 2014-06-06 | 2014-09-10 | 清华大学 | Transformer aging fault rate estimation method |
CN104077231A (en) * | 2014-07-16 | 2014-10-01 | 国家电网公司 | Transformer maintenance optimization method based on symbol dynamics and LS-SVM |
CN104101795A (en) * | 2014-02-19 | 2014-10-15 | 江苏倍尔科技发展有限公司 | Transformer fault control method |
-
2015
- 2015-11-24 CN CN201510821187.6A patent/CN105372528B/en not_active Expired - Fee Related
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20100029338A (en) * | 2008-09-08 | 2010-03-17 | 하주영 | System for diagnostication of transformer using ultrasonic wave |
CN102779230A (en) * | 2012-06-14 | 2012-11-14 | 华南理工大学 | State analysis and maintenance decision judging method of power transformer system |
WO2014072444A1 (en) * | 2012-11-09 | 2014-05-15 | Bombardier Transportation Gmbh | Method and device for monitoring a transformer state |
CN103091603A (en) * | 2013-01-14 | 2013-05-08 | 华北电力大学 | Breakdown intelligent classification and positioning method of electric transmission line |
CN103245911A (en) * | 2013-05-03 | 2013-08-14 | 云南电力试验研究院(集团)有限公司电力研究院 | Breaker fault diagnosis method based on Bayesian network |
CN104101795A (en) * | 2014-02-19 | 2014-10-15 | 江苏倍尔科技发展有限公司 | Transformer fault control method |
CN104007343A (en) * | 2014-05-23 | 2014-08-27 | 清华大学 | Dynamic comprehensive transformer fault diagnosis method based on Bayesian network |
CN104036131A (en) * | 2014-06-06 | 2014-09-10 | 清华大学 | Transformer aging fault rate estimation method |
CN104077231A (en) * | 2014-07-16 | 2014-10-01 | 国家电网公司 | Transformer maintenance optimization method based on symbol dynamics and LS-SVM |
Non-Patent Citations (3)
Title |
---|
单相双绕组变压器匝间短路故障诊断;谭喜堂 等;《电力***及其自动化》;20150131;第37卷(第1期);第73-74、107页 * |
基于灰云模型的电力变压器故障诊断;蔡红梅 等;《电力***保护与控制》;20120616;第40卷(第12期);第151-155页 * |
基于符号动力学信息熵与改进神经网络的风机故障诊断研究;王松岭 等;《华北电力大学学报》;20130731;第40卷(第4期);第51-58页 * |
Also Published As
Publication number | Publication date |
---|---|
CN105372528A (en) | 2016-03-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105372528B (en) | A kind of state maintenance method of Power Transformer Internal Faults and New Transformer | |
CN109842373A (en) | Diagnosing failure of photovoltaic array method and device based on spatial and temporal distributions characteristic | |
CN105677791B (en) | For analyzing the method and system of the operation data of wind power generating set | |
CN110879917A (en) | Electric power system transient stability self-adaptive evaluation method based on transfer learning | |
CN109297689B (en) | Large-scale hydraulic machinery intelligent diagnosis method introducing weight factors | |
CN106647650B (en) | Distributing Industrial Process Monitoring method based on variable weighting pca model | |
CN110703077B (en) | HPSO-TSVM-based high-voltage circuit breaker fault diagnosis method | |
CN110689069A (en) | Transformer fault type diagnosis method based on semi-supervised BP network | |
CN110689068B (en) | Transformer fault type diagnosis method based on semi-supervised SVM | |
CN110766313A (en) | Cable tunnel comprehensive state evaluation method based on operation and maintenance system | |
CN116401532B (en) | Method and system for recognizing frequency instability of power system after disturbance | |
CN108776017A (en) | A kind of rolling bearing method for predicting residual useful life improving CHSMM | |
CN105954616B (en) | Photovoltaic module method for diagnosing faults based on external characteristics electric parameter | |
Yau et al. | Chaotic eye‐based fault forecasting method for wind power systems | |
CN111999591B (en) | Method for identifying abnormal state of primary equipment of power distribution network | |
CN107102223A (en) | NPC photovoltaic DC-to-AC converter method for diagnosing faults based on improved hidden Markov model GHMM | |
CN114169249B (en) | Artificial intelligent identification method for high-resistance ground fault of power distribution network | |
CN115879048A (en) | Series arc fault identification method and system based on WRFMDA model | |
CN113572771B (en) | Power grid CPS network attack identification method and system | |
Aziz et al. | Critical comparison of power-based wind turbine fault-detection methods using a realistic framework for SCADA data simulation | |
Jing et al. | Adjustable piecewise regression strategy based wind turbine power forecasting for probabilistic condition monitoring | |
CN107147143B (en) | Method for establishing early warning model of fan interlocking off-line fault | |
CN105116323B (en) | A kind of electrical fault detection method based on RBF | |
CN111953657B (en) | Sequence-data joint driven CPS network attack identification method for power distribution network | |
CN105741184A (en) | Transformer state evaluation method and apparatus |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into 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: 20181009 Termination date: 20211124 |
|
CF01 | Termination of patent right due to non-payment of annual fee |