CN114814501A - On-line diagnosis method for capacitor breakdown fault of capacitor voltage transformer - Google Patents

On-line diagnosis method for capacitor breakdown fault of capacitor voltage transformer Download PDF

Info

Publication number
CN114814501A
CN114814501A CN202210758583.9A CN202210758583A CN114814501A CN 114814501 A CN114814501 A CN 114814501A CN 202210758583 A CN202210758583 A CN 202210758583A CN 114814501 A CN114814501 A CN 114814501A
Authority
CN
China
Prior art keywords
capacitor
cvt
breakdown
samples
voltage
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
CN202210758583.9A
Other languages
Chinese (zh)
Other versions
CN114814501B (en
Inventor
魏伟
叶利
方毅
彭涛
明东岳
汪应春
王雅兰
杨丽华
谢玮
李云峰
徐子雅
周卓鹏
刘义
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.)
Metering Center of State Grid Hubei Electric Power Co Ltd
Original Assignee
Metering Center of State Grid Hubei Electric Power Co Ltd
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 Metering Center of State Grid Hubei Electric Power Co Ltd filed Critical Metering Center of State Grid Hubei Electric Power Co Ltd
Priority to CN202210758583.9A priority Critical patent/CN114814501B/en
Publication of CN114814501A publication Critical patent/CN114814501A/en
Application granted granted Critical
Publication of CN114814501B publication Critical patent/CN114814501B/en
Active 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
    • G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R35/00—Testing or calibrating of apparatus covered by the other groups of this subclass
    • G01R35/02—Testing or calibrating of apparatus covered by the other groups of this subclass of auxiliary devices, e.g. of instrument transformers according to prescribed transformation ratio, phase angle, or wattage rating
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
    • G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Power Engineering (AREA)
  • Testing Electric Properties And Detecting Electric Faults (AREA)

Abstract

The application relates to an on-line diagnosis method for a capacitor breakdown fault of a capacitor voltage transformer, which comprises the following specific steps: s1, collecting secondary voltage amplitude data of CVTs of the same phase and the same model on the same bus by using a CVT online monitoring system, constructing a data set and attaching corresponding labels; s2, constructing a feature vector set by using the inter-group amplitude ratio parameters, and performing feature extraction on the preprocessed data set; s3, training a KNN model by adopting a KNN machine learning mode, adjusting corresponding hyper-parameters, and outputting an optimal KNN model; and S4, inputting secondary voltage amplitude data in the CVT on-line monitoring system into the optimal KNN model, and diagnosing the capacitor breakdown state of the CVT to be detected in real time. The invention carries out online diagnosis on breakdown faults of the CVT capacitor and feeds back the state of the capacitor in the CVT in real time, thereby ensuring the stability and the safety performance of the operation of a power grid.

Description

On-line diagnosis method for capacitor breakdown fault of capacitor voltage transformer
Technical Field
The application relates to the field of capacitor breakdown fault online diagnosis, in particular to a capacitor breakdown fault online diagnosis method for a capacitor voltage transformer.
Background
Compared with the traditional electromagnetic transformer, the Capacitor Voltage Transformer (CVT) has the characteristics of strong anti-ferromagnetic resonance capability, low manufacturing cost, small volume, light weight and the like. However, because the CVT has a complex internal structure, the operating state is susceptible to environmental factors, and abnormal states such as error over-tolerance and insulation performance degradation are likely to occur in the long-term operation process, thereby affecting the accuracy of electric energy metering and the safety of the power system. Among them, as one of the main factors affecting the state of the CVT, the state of the insulation performance is mainly determined by the state of the capacitor and the insulating medium in the insulating structure inside the CVT. In contrast to insulating media, where there are many well-established methods for characterizing the state, there is currently no accurate method for characterizing the state of capacitance on-line. And the capacitor state is monitored on line in real time, so that the state of the CVT can be judged in time, and related operation and maintenance personnel can conveniently overhaul and maintain the work. If the capacitor state is not found to be abnormal in time, the state of the transformer is abnormal, so that the operation of a power grid is influenced, and even the personnel safety problem caused by explosion is caused. In order to avoid inaccuracy of a secondary information system information source, reduce loss of electric energy measurement and ensure normal operation of a measurement and control protection device, the on-line diagnosis of the breakdown fault of the capacitor of the CVT is a technical problem.
Disclosure of Invention
The embodiment of the application aims to provide an online diagnosis method for capacitor breakdown faults of a capacitor voltage transformer, which is used for online diagnosis of the capacitor breakdown faults of a CVT (constant voltage transformer), and feeding back the state of a capacitor in the CVT in real time, so that the running stability and safety performance of a power grid are ensured.
In order to achieve the above purpose, the present application provides the following technical solutions:
the embodiment of the application provides an on-line diagnosis method for a capacitor breakdown fault of a capacitor voltage transformer, which comprises the following specific steps:
s1, collecting secondary voltage amplitude data of CVTs of the same phase and the same model on the same bus by using a CVT online monitoring system, constructing a data set and attaching corresponding labels;
s2, constructing a feature vector set by using the inter-group amplitude ratio parameters, and performing feature extraction on the preprocessed data set;
s3, training a KNN model by adopting a KNN machine learning mode, adjusting corresponding hyper-parameters, and outputting an optimal KNN model;
and S4, inputting secondary voltage amplitude data in the CVT on-line monitoring system into the optimal KNN model, and diagnosing the capacitor breakdown state of the CVT to be detected in real time.
The specific method for constructing the data set comprises the steps of screening historical data of secondary voltage amplitude values and metering errors of CVTs (continuously variable Transmission) which are of the same type and have the same phase and adopt the same wiring mode on one bus from an online monitoring system, manually classifying sample data at each moment according to the relation between capacitance breakdown and CVT transformation ratio, wherein the category of the capacitance breakdown state is
Figure 193357DEST_PATH_IMAGE001
Wherein
Figure 214009DEST_PATH_IMAGE002
Indicating that the capacitance has not broken down,
Figure 208510DEST_PATH_IMAGE003
indicating a breakdown of a single high-voltage capacitor,
Figure 959428DEST_PATH_IMAGE004
indicating that the two high-voltage capacitors are broken down,
Figure 722985DEST_PATH_IMAGE005
indicating a single medium voltage capacitance breakdown and,
Figure 416003DEST_PATH_IMAGE006
and representing two medium-voltage capacitor breakdowns, taking one CVT as a CVT to be tested for capacitor breakdown fault diagnosis, and selecting a sample with only the CVT to be tested in an abnormal capacitance state or all the CVTs in a normal state from the selected data set as a data set D for subsequent online capacitor breakdown diagnosis.
The construction of the feature vector set by using the inter-group amplitude ratio parameters is to extract the features of the preprocessed data set,
splitting the data set D into a training set and a test set;
carrying out normalization processing on samples in the training set;
and constructing a characteristic vector of the CVT capacitance breakdown online diagnosis according to the inter-group amplitude ratio parameter so as to eliminate the influence of primary voltage on the capacitance breakdown online diagnosis result.
The specific method for splitting the data set D into the training set and the test set is to set the sample proportion of the training set and the test set as 7: if the number of samples in the data set D is
Figure 632221DEST_PATH_IMAGE007
Then the number of samples in the training set is
Figure 921251DEST_PATH_IMAGE008
The number of samples in the test set is
Figure 804893DEST_PATH_IMAGE009
Randomly non-replaced drawn samples from the data set D to the training sample set
Figure 904698DEST_PATH_IMAGE010
In, co-extraction
Figure 873792DEST_PATH_IMAGE011
Next, the remaining samples in data set D constitute the test set
Figure 700933DEST_PATH_IMAGE012
。
The normalization process for the samples in the training set specifically includes,
training set with z-score normalization
Figure 704661DEST_PATH_IMAGE010
Normalized by z-score of the formula
Figure 473903DEST_PATH_IMAGE013
Wherein
Figure 930292DEST_PATH_IMAGE014
Is the mean value of the features of the dimension,
Figure 561125DEST_PATH_IMAGE015
for the standard deviation of the dimension feature, record the correspondence
Figure 419359DEST_PATH_IMAGE014
，
Figure 338598DEST_PATH_IMAGE015
Apply it to test set
Figure 547862DEST_PATH_IMAGE012
And (4) normalization processing.
Parameter of amplitude ratio between groups
Figure 716807DEST_PATH_IMAGE016
Refers to the ratio of the secondary voltage amplitudes of two CVTs of the same model representing the same phase, in different groups located on the same bus,
Figure 429548DEST_PATH_IMAGE017
wherein
Figure 540592DEST_PATH_IMAGE018
Respectively representing the secondary voltage amplitudes of two different CVTs according to a metering error ratio difference formula of the CVT
Figure 705994DEST_PATH_IMAGE019
Wherein
Figure 678629DEST_PATH_IMAGE020
Is the ratio of the rated variation to the rated variation,
Figure 511456DEST_PATH_IMAGE021
is the voltage of the secondary battery, and is,
Figure 294867DEST_PATH_IMAGE022
is a primary voltage, and the form conversion is carried out to obtain:
Figure 947565DEST_PATH_IMAGE023
substituting the form-converted ratio equation into the interclass amplitude ratio parameter
Figure 723891DEST_PATH_IMAGE024
Because two CVTs are connected to the same bus, have the same model and represent the same phase, the primary voltage and the rated transformation ratio of the CVT are the same, namely the formula can be simplified into
Figure 411224DEST_PATH_IMAGE025
That is, it is approximately considered that the secondary voltage value of the CVT between the two sets is related to the ratio difference of the two CVTs, thereby eliminating the influence of the primary voltage value, in the same bus, assuming that there are n sets of CVTs of the same model,
Figure 864071DEST_PATH_IMAGE026
for the CVTs located in the same phase in the n sets of CVTs, selecting one of the CVTs to perform capacitance breakdown online diagnosis, and marking the CVT as G, the set of the sets of the CVTs located in the same phase on the same bus can be marked as G
Figure 4066DEST_PATH_IMAGE027
The feature vector set of the inter-group amplitude ratio parameters of the individual samples set accordingly is
Figure 318503DEST_PATH_IMAGE028
Taking the characteristic vector set as a characteristic vector set for on-line diagnosis of the capacitor breakdown, and after the characteristic vector set extraction is carried out on the training set after the normalization operation is completed, taking the sample of the training set as
Figure 125922DEST_PATH_IMAGE029
The corresponding category is
Figure 248206DEST_PATH_IMAGE030
The samples of the test set corresponding to the training set using normalization
Figure 282021DEST_PATH_IMAGE014
And
Figure 524783DEST_PATH_IMAGE015
corresponding normalization is carried out to obtain a training set
Figure 45763DEST_PATH_IMAGE031
And corresponding categories
Figure 715779DEST_PATH_IMAGE032
。
The method for training the KNN model by adopting the KNN machine learning mode, adjusting the corresponding hyper-parameters and outputting the optimal KNN model specifically comprises the following steps of,
s31, selecting the hyperparameter in KNN through knowledge and experience in the field
Figure 705732DEST_PATH_IMAGE033
，
Figure 486606DEST_PATH_IMAGE034
Is a positive integer;
s32. selecting a sample from the test set, calculating its distance to all training samples, wherein the distance metric used is Min's distance
Figure 363558DEST_PATH_IMAGE035
Wherein
Figure 204475DEST_PATH_IMAGE036
，
Figure 540778DEST_PATH_IMAGE037
Respectively represent two
Figure 709DEST_PATH_IMAGE038
The dimension samples are then processed to obtain a dimensional sample,
Figure 371648DEST_PATH_IMAGE039
indicates that is the second of the sample
Figure 242521DEST_PATH_IMAGE040
Dimensional features, s is a constant;
s33, selecting the nearest distance in all the distances obtained by the last step of calculation
Figure 331699DEST_PATH_IMAGE034
Training sample corresponding to each distance
Figure 188797DEST_PATH_IMAGE041
；
S34, the selected training samples correspond to the classes respectively
Figure 289608DEST_PATH_IMAGE042
Counting W according to L, namely calculating the number of the various types in L appearing in W
Figure 737907DEST_PATH_IMAGE043
Selecting the most numerous classes
Figure 406392DEST_PATH_IMAGE044
The corresponding class is used as the class of the test sample;
s35, repeating the steps from S32 to S34, and classifying all samples in the test set to obtain a corresponding class set
Figure 332760DEST_PATH_IMAGE045
The result of the classification
Figure 412712DEST_PATH_IMAGE046
And the actual results
Figure 641699DEST_PATH_IMAGE047
Comparing to obtain the precision of the scheme;
Figure 174311DEST_PATH_IMAGE048
s36. selected according to step S31
Figure 29004DEST_PATH_IMAGE034
Value, selection
Figure 963462DEST_PATH_IMAGE049
Repeating the steps S31 to S35 for a plurality of right and left data to output a plurality of different accuracies, and selecting the most accurate data as the optimal parameters of KNN.
The step S4 is to collect the amplitude of the CVT on-line monitoring system at the current time, and input the amplitude with the best hyper-parameter after data preprocessing
Figure 487984DEST_PATH_IMAGE049
And outputting the on-line diagnosis result of the capacitor breakdown of the CVT to be detected at the current moment in the KNN model.
Compared with the prior art, the invention has the beneficial effects that:
a scheme for on-line diagnosis of the breakdown fault of the capacitor voltage transformer based on the KNN is provided. And constructing a characteristic vector according to the inter-group amplitude ratio parameters, and constructing a capacitor breakdown online diagnosis model of the capacitor voltage transformer by using KNN. And inputting real-time data of the CVT on-line monitoring system into the trained KNN model, and diagnosing the breakdown fault state of the current CVT capacitor in real time. And performing online diagnosis on breakdown faults of the capacitor of the CVT, and feeding back the state of the capacitor in the CVT in real time, thereby ensuring the stability and safety performance of the operation of a power grid.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and that those skilled in the art can also obtain other related drawings based on the drawings without inventive efforts.
FIG. 1 is a schematic flow chart of a method according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
In a first aspect of the present invention, an online diagnosis method for a capacitor breakdown fault of a capacitor voltage transformer is provided, as shown in fig. 1,
step 1: acquisition of a data set
And acquiring secondary voltage amplitude information through a CVT online monitoring system. The specific mode is that historical data of secondary voltage amplitude and metering error of CVTs of the same type and the same phase adopting the same wiring mode on one bus are screened from an online monitoring system, sample data at each moment can be manually classified according to the relation between capacitance breakdown and CVT transformation ratio, and the category of the capacitance breakdown state is
Figure 648838DEST_PATH_IMAGE050
Wherein
Figure 651429DEST_PATH_IMAGE051
Indicating that the capacitance has not broken down,
Figure 440393DEST_PATH_IMAGE052
indicating a single high-voltage capacitance breakdown and,
Figure 27495DEST_PATH_IMAGE053
indicating that the two high-voltage capacitors are broken down,
Figure 800279DEST_PATH_IMAGE054
indicating a single medium voltage capacitance breakdown and,
Figure 216348DEST_PATH_IMAGE055
and representing two medium-voltage capacitor breakdowns, taking one CVT as a CVT to be tested for capacitor breakdown fault diagnosis, and selecting a sample with only the CVT to be tested in an abnormal capacitance state or all the CVTs in a normal state from the selected data set as a data set D for subsequent online capacitor breakdown diagnosis.
Step2 data preprocessing
Step2.1 dataset splitting
Gathering CVT historical data collected by step1
Figure 390977DEST_PATH_IMAGE056
And splitting the training set and the test set. The specific method is that the sample proportion of the training set and the test set is set as 7: if the number of samples in the data set D is
Figure 381936DEST_PATH_IMAGE057
Then the number of samples in the training set is
Figure 642016DEST_PATH_IMAGE058
The number of samples in the test set is
Figure 127355DEST_PATH_IMAGE059
Randomly non-replaced drawn samples from the data set D to the training sample set
Figure 890912DEST_PATH_IMAGE060
In, co-extraction
Figure 193717DEST_PATH_IMAGE061
Next, the remaining samples in data set D constitute the test set
Figure 298683DEST_PATH_IMAGE062
。
Step2.2 normalization treatment
For the training set
Figure 712347DEST_PATH_IMAGE060
Normalizing the sample in (1)And (6) chemical treatment. Training set with z-score normalization
Figure 205776DEST_PATH_IMAGE060
Normalized by z-score of the formula
Figure 679483DEST_PATH_IMAGE063
Wherein
Figure 773209DEST_PATH_IMAGE064
Is the mean value of the features of the dimension,
Figure 724985DEST_PATH_IMAGE065
for the standard deviation of the dimension feature, record the correspondence
Figure 463134DEST_PATH_IMAGE064
，
Figure 983108DEST_PATH_IMAGE065
Apply it to test set
Figure 439497DEST_PATH_IMAGE062
And (4) normalization processing.
Construction of Step2.3 feature vectors
And constructing a characteristic vector of the CVT capacitance breakdown online diagnosis according to the inter-group amplitude ratio parameter so as to eliminate the influence of primary voltage on the capacitance breakdown online diagnosis result. Parameter of amplitude ratio between groups
Figure 821062DEST_PATH_IMAGE066
Refers to the ratio of the secondary voltage amplitudes of two CVTs of the same model representing the same phase, in different groups located on the same bus,
Figure 413717DEST_PATH_IMAGE067
wherein
Figure 839014DEST_PATH_IMAGE068
Respectively representing the secondary voltage amplitudes of two different CVTs according to a metering error ratio difference formula of the CVT
Figure 782699DEST_PATH_IMAGE069
Wherein
Figure 466490DEST_PATH_IMAGE070
Is the ratio of the rated variation to the rated variation,
Figure 913652DEST_PATH_IMAGE071
is the voltage of the secondary battery, and is,
Figure 634483DEST_PATH_IMAGE072
is a primary voltage, and the form conversion is carried out to obtain:
Figure 940831DEST_PATH_IMAGE073
substituting the form-converted ratio equation into the interclass amplitude ratio parameter
Figure 38100DEST_PATH_IMAGE074
Because two CVTs are connected to the same bus, have the same model and represent the same phase, the primary voltage and the rated transformation ratio of the CVT are the same, namely the formula can be simplified into
Figure 505814DEST_PATH_IMAGE075
That is, it is approximately considered that the secondary voltage value of the CVT between the two sets is related to the ratio difference of the two CVTs, thereby eliminating the influence of the primary voltage value, in the same bus, assuming that there are n sets of CVTs of the same model,
Figure 397547DEST_PATH_IMAGE076
for n sets of CVT neutralFor the CVT with the same phase, one CVT is selected for capacitance breakdown online diagnosis, the CVT is marked as G, and the set of the groups of CVTs with the same phase under the same bus can be marked as G
Figure 191191DEST_PATH_IMAGE077
The feature vector set of the inter-group amplitude ratio parameters of the individual samples set accordingly is
Figure 92150DEST_PATH_IMAGE078
Taking the characteristic vector set as a characteristic vector set for on-line diagnosis of the capacitor breakdown, and after the characteristic vector set extraction is carried out on the training set after the normalization operation is completed, taking the sample of the training set as
Figure 638538DEST_PATH_IMAGE079
The corresponding category is
Figure 966751DEST_PATH_IMAGE080
The samples of the test set corresponding to the training set using normalization
Figure 247691DEST_PATH_IMAGE064
And
Figure 686763DEST_PATH_IMAGE065
corresponding normalization is carried out to obtain a training set
Figure 854701DEST_PATH_IMAGE081
And corresponding categories
Figure 353816DEST_PATH_IMAGE082
。
Step3 classification of CVT for on-line diagnosis of capacitive breakdown using KNN
Step3.1 selection of hyper-parameters in KNN by domain knowledge and experience
Figure 856472DEST_PATH_IMAGE083
（
Figure 99235DEST_PATH_IMAGE083
Is a positive integer).
Step3.2 select one sample from the test set and calculate its distance to all training samples. Wherein the distance measure used is a Min-style distance
Figure 620215DEST_PATH_IMAGE084
Wherein
Figure 290231DEST_PATH_IMAGE085
，
Figure 14604DEST_PATH_IMAGE086
Respectively represent two
Figure 795478DEST_PATH_IMAGE087
The dimension samples are then processed to obtain a dimensional sample,
Figure 935079DEST_PATH_IMAGE088
indicates that is the second of the sample
Figure 448100DEST_PATH_IMAGE089
Dimensional features, s is a constant;
step3.3 selects the nearest of all distances calculated by step3.2
Figure 49983DEST_PATH_IMAGE083
Training sample corresponding to each distance
Figure 493603DEST_PATH_IMAGE090
。
The classes corresponding to the training samples selected by Step3.4 are respectively
Figure 864541DEST_PATH_IMAGE091
Counting W according to L, namely calculating the number of the various types in L appearing in W
Figure 610780DEST_PATH_IMAGE092
Selecting the most numerous classes
Figure 309746DEST_PATH_IMAGE093
The corresponding class is taken as the class of the test sample;
step3.5 repeating step3.2-step3.4, classifying all samples in the test set to obtain a corresponding category set
Figure 698002DEST_PATH_IMAGE094
The result of the classification
Figure 283966DEST_PATH_IMAGE095
And the actual results
Figure 466686DEST_PATH_IMAGE096
Comparing to obtain the precision of the scheme;
Figure 918527DEST_PATH_IMAGE097
step3.6 selected according to step3.1
Figure 579315DEST_PATH_IMAGE083
Value, selection
Figure 659267DEST_PATH_IMAGE083
And repeating the steps of Step3.1-Step3.5 on the left and the right to output a plurality of different accuracies, and selecting the optimal parameter with the highest accuracy as KNN.
Step4 classification of CVT for on-line diagnosis of capacitive breakdown using KNN
Collecting the amplitude value in the CVT on-line monitoring system at the current moment, preprocessing the data, and inputting the data with the optimal hyper-parameter
Figure 403101DEST_PATH_IMAGE083
And outputting the on-line diagnosis result of the capacitor breakdown of the CVT to be detected at the current moment in the KNN model.
Specific application examples of the present application:
firstly, a simulation circuit experiment platform is constructed, 6 groups of three-phase four-wire CVTs of the same type are connected to a 110kV bus, an A phase is selected, the secondary voltage amplitude of each group of A phase CVT is recorded according to seconds, and 15000 samples are obtained in total.
Each sample was labeled as follows:
label 1: the corresponding ratio difference change is about 0 when the capacitance is not broken down
And 2, label: when a single high-voltage capacitor breaks down, the corresponding ratio difference change is about 0.4%;
and (3) labeling: the corresponding ratio difference change of the two high-voltage capacitors is about 0.8% when the two high-voltage capacitors are broken down;
and (4) labeling: the corresponding ratio difference change is about-0.4% when a single medium-voltage capacitor breaks down;
and (5) labeling: the corresponding change in the ratio between the two medium voltage capacitors in breakdown is about-0.8%.
I.e. each sample taken contains six secondary voltage magnitudes of the CVT and corresponding labels. Collected data set analysis is shown in the following table
Figure 935713DEST_PATH_IMAGE098
The collected sample data set comprises 600 normal samples, 2400 samples with No. 1 CVT failure (600 samples in four abnormal states), 2400 samples with No. 2 CVT failure (600 samples in four abnormal states), 2400 samples with No. 3 CVT failure (600 samples in four abnormal states), 2400 samples with No. 4 CVT failure (600 samples in four abnormal states), 2400 samples with No. 5 CVT failure (600 samples in four abnormal states), and 2400 samples with No. 6 CVT failure (600 samples in four abnormal states).
And secondly, performing online monitoring analysis on the failed CVT, namely constructing six data sets,
Figure 541138DEST_PATH_IMAGE099
data set
Figure 475596DEST_PATH_IMAGE100
The number of samples containing normal samples and CVT No. 1 failed is 3000. Data set
Figure 888867DEST_PATH_IMAGE101
The number of samples containing normal samples and CVT No. 2 failed is 3000. Data set
Figure 908775DEST_PATH_IMAGE102
The samples containing normal samples and CVT No. 3 failed, total 3000. Data set
Figure 176945DEST_PATH_IMAGE103
The samples containing normal samples and CVT No. 4 failed, total 3000. Data set
Figure 841276DEST_PATH_IMAGE104
The samples containing normal samples and CVT No. 5 failed, total 3000. Data set
Figure 802279DEST_PATH_IMAGE105
The number of samples containing normal samples and CVT No. 6 failed is 3000. Each data set was calculated as 7: 3 into training set and test set. And carrying out normalization operation on each data set, and then extracting a data feature set based on the voltage ratio difference between the groups.
Based on experience, will
Figure 434117DEST_PATH_IMAGE083
The nearest neighbor number is set to 5, and the settings are different after the experiment
Figure 240399DEST_PATH_IMAGE083
The method accuracy of the value mapping is shown below
Figure 415029DEST_PATH_IMAGE106
As shown in the table, the highest precision squareCases 5, 10, 15 were obtained because of empirical settings
Figure 891141DEST_PATH_IMAGE083
The accuracy of the KNN model corresponding to the value is highest, so that the KNN model is selected
Figure 151221DEST_PATH_IMAGE083
The KNN method at =5 is an optimum method for this method.
The embodiment of the application provides a KNN-based capacitive voltage transformer capacitance breakdown fault on-line diagnosis scheme. And constructing a characteristic vector according to the inter-group amplitude ratio parameters, and constructing a capacitor breakdown online diagnosis model of the capacitor voltage transformer by using KNN. And inputting real-time data of the CVT on-line monitoring system into the trained KNN model, and diagnosing the breakdown fault state of the current CVT capacitor in real time.
The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims (8)

1. The on-line diagnosis method for the capacitor breakdown fault of the capacitor voltage transformer is characterized by comprising the following specific steps of:
s1, collecting secondary voltage amplitude data of CVTs of the same phase and the same model on the same bus by using a CVT online monitoring system, constructing a data set and attaching corresponding labels;
s2, constructing a feature vector set by using the inter-group amplitude ratio parameters, and performing feature extraction on the preprocessed data set;
s3, training a KNN model by adopting a KNN machine learning mode, adjusting corresponding hyper-parameters, and outputting an optimal KNN model;
and S4, inputting secondary voltage amplitude data in the CVT on-line monitoring system into the optimal KNN model, and diagnosing the capacitor breakdown state of the CVT to be detected in real time.
2. The method according to claim 1, wherein the specific method for constructing the data set is to screen historical data of secondary voltage amplitude and metering error of CVT (constant-voltage transformer) of the same type and the same phase in the same wiring way on a bus from an online monitoring system, and manually classify sample data at each moment according to the relation between the capacitor breakdown and the CVT transformation ratio, wherein the capacitor breakdown state is classified into
Figure 149963DEST_PATH_IMAGE001
In which
Figure 298048DEST_PATH_IMAGE002
Indicating that the capacitance has not broken down,
Figure 181690DEST_PATH_IMAGE003
indicating a single high-voltage capacitance breakdown and,
Figure 248872DEST_PATH_IMAGE004
indicating that the two high-voltage capacitors are broken down,
Figure 217965DEST_PATH_IMAGE005
indicating a single breakdown of the medium voltage capacitor,
Figure 310686DEST_PATH_IMAGE006
and representing two medium-voltage capacitor breakdowns, taking one CVT as a CVT to be tested for capacitor breakdown fault diagnosis, and selecting a sample with only the CVT to be tested in an abnormal capacitance state or all the CVTs in a normal state from the selected data set as a data set D for subsequent online capacitor breakdown diagnosis.
3. The on-line diagnosis method for the capacitor breakdown fault of the capacitor voltage transformer as claimed in claim 2, wherein the feature vector set is constructed by using the inter-group amplitude ratio parameter, and the feature extraction is specifically performed on the preprocessed data set,
splitting the data set D into a training set and a test set;
carrying out normalization processing on samples in the training set;
and constructing a characteristic vector of the CVT capacitance breakdown online diagnosis according to the inter-group amplitude ratio parameter so as to eliminate the influence of primary voltage on the capacitance breakdown online diagnosis result.
4. The on-line diagnosis method for the capacitor breakdown fault of the capacitor voltage transformer as claimed in claim 3, wherein the data set D is divided into the training set and the test set by setting the sample ratio of the training set to 7: if the number of samples in the data set D is
Figure 48835DEST_PATH_IMAGE007
Then the number of samples in the training set is
Figure 319541DEST_PATH_IMAGE008
The number of samples in the test set is
Figure 510351DEST_PATH_IMAGE009
Randomly non-replaced drawn samples from the data set D to the training sample set
Figure 328135DEST_PATH_IMAGE010
In, co-extraction
Figure 809538DEST_PATH_IMAGE011
Next, the remaining samples in data set D constitute the test set
Figure 765993DEST_PATH_IMAGE012
。
5. The on-line diagnosis method for the capacitor breakdown fault of the capacitor voltage transformer as claimed in claim 4, wherein the normalization process for the samples in the training set is specifically,
training set with z-score normalization
Figure 444099DEST_PATH_IMAGE010
Normalized by z-score of the formula
Figure 127890DEST_PATH_IMAGE013
Wherein
Figure 840631DEST_PATH_IMAGE014
Is the mean value of the features of the dimension,
Figure 702408DEST_PATH_IMAGE015
for the standard deviation of the dimension feature, record the correspondence
Figure 133389DEST_PATH_IMAGE014
，
Figure 856757DEST_PATH_IMAGE015
Apply it to test set
Figure 424004DEST_PATH_IMAGE012
The normalization processing of (3).
6. The on-line diagnosis method for the capacitor breakdown fault of the capacitor voltage transformer as claimed in claim 5, wherein the inter-group amplitude ratio parameter
Figure 456682DEST_PATH_IMAGE016
Refers to the ratio of the secondary voltage amplitudes of two CVTs of the same model representing the same phase, in different groups located on the same bus,
Figure 109381DEST_PATH_IMAGE017
wherein
Figure 134974DEST_PATH_IMAGE018
Respectively representing the secondary voltage amplitudes of two different CVTs according to a metering error ratio difference formula of the CVT
Figure 822308DEST_PATH_IMAGE019
Wherein
Figure 884941DEST_PATH_IMAGE020
Is the ratio of the rated variation to the rated variation,
Figure 165881DEST_PATH_IMAGE021
is the voltage of the secondary battery, and is,
Figure 339374DEST_PATH_IMAGE022
is a primary voltage, and the form conversion is carried out to obtain:
Figure 516101DEST_PATH_IMAGE023
substituting the ratio difference formula after the form conversion into the intergroup amplitude ratio parameter
Figure 15216DEST_PATH_IMAGE024
Because two CVTs are connected to the same bus, have the same model and represent the same phase, the primary voltage and the rated transformation ratio of the CVT are the same, namely the formula can be simplified into
Figure 517872DEST_PATH_IMAGE025
Namely, the secondary voltage value of the CVT between the two groups is approximately considered to be related to the ratio difference of the two CVTsThereby eliminating the influence of the primary voltage value, in the same bus, assuming that there are n sets of CVTs of the same model,
Figure 150848DEST_PATH_IMAGE026
for the CVTs located in the same phase in the n sets of CVTs, selecting one of the CVTs to perform capacitance breakdown online diagnosis, and marking the CVT as G, the set of the sets of the CVTs located in the same phase on the same bus can be marked as G
Figure 812773DEST_PATH_IMAGE027
The feature vector set of the inter-group amplitude ratio parameters of the individual samples set accordingly is
Figure 358155DEST_PATH_IMAGE028
Taking the characteristic vector set as a characteristic vector set for on-line diagnosis of capacitor breakdown, and after extracting the characteristic vector set from the training set after normalization operation, taking the sample of the training set as
Figure 207162DEST_PATH_IMAGE029
The corresponding category is
Figure 879714DEST_PATH_IMAGE030
The samples of the test set corresponding to the training set using normalization
Figure 661726DEST_PATH_IMAGE014
And
Figure 378009DEST_PATH_IMAGE015
corresponding normalization is carried out to obtain a training set
Figure 714312DEST_PATH_IMAGE031
And corresponding categories
Figure 689090DEST_PATH_IMAGE032
。
7. The on-line diagnosis method for the capacitor breakdown fault of the capacitor voltage transformer as claimed in claim 5, wherein the KNN model is trained by KNN machine learning, the corresponding hyper-parameters are adjusted, and the optimal KNN model is outputted,
s31, selecting the hyperparameter in KNN through knowledge and experience in the field
Figure 200974DEST_PATH_IMAGE033
，
Figure 212793DEST_PATH_IMAGE034
Is a positive integer;
s32. selecting a sample from the test set, calculating its distance to all training samples, wherein the distance metric used is Min's distance
Figure 190720DEST_PATH_IMAGE035
Wherein
Figure 47817DEST_PATH_IMAGE036
，
Figure 414208DEST_PATH_IMAGE037
Respectively represent two
Figure 596927DEST_PATH_IMAGE038
The dimension samples are then processed to obtain a dimensional sample,
Figure 32457DEST_PATH_IMAGE039
indicates that is the second of the sample
Figure 958824DEST_PATH_IMAGE040
Dimensional features, s is a constant;
s33, selecting the obtained result obtained by the previous step of calculationWith the nearest in distance
Figure 179721DEST_PATH_IMAGE034
Training sample corresponding to each distance
Figure 267763DEST_PATH_IMAGE041
；
S34, the selected training samples correspond to the classes respectively
Figure 957633DEST_PATH_IMAGE042
Counting W according to L, namely calculating the number of the various types in L appearing in W
Figure 422112DEST_PATH_IMAGE043
Selecting the most numerous classes
Figure 497515DEST_PATH_IMAGE044
The corresponding class is used as the class of the test sample;
s35, repeating the steps from S32 to S34, and classifying all samples in the test set to obtain a corresponding class set
Figure 22038DEST_PATH_IMAGE045
The result of the classification
Figure 432159DEST_PATH_IMAGE046
And the actual results
Figure 589078DEST_PATH_IMAGE047
Comparing to obtain the precision of the scheme;
Figure 518988DEST_PATH_IMAGE048
s36. selected according to step S31
Figure 214411DEST_PATH_IMAGE034
Value, selection
Figure 377408DEST_PATH_IMAGE049
Repeating the steps S31 to S35 for a plurality of right and left data to output a plurality of different accuracies, and selecting the most accurate data as the optimal parameters of KNN.
8. The on-line diagnosis method for capacitor breakdown fault of capacitor voltage transformer of claim 7, wherein the step S4 is to collect the amplitude value of the CVT on-line monitoring system at the current moment, and after data preprocessing, input the amplitude value with the best hyper-parameter
Figure 449269DEST_PATH_IMAGE049
And outputting the on-line diagnosis result of the capacitor breakdown of the CVT to be detected at the current moment in the KNN model.
CN202210758583.9A 2022-06-30 2022-06-30 On-line diagnosis method for capacitor breakdown fault of capacitor voltage transformer Active CN114814501B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210758583.9A CN114814501B (en) 2022-06-30 2022-06-30 On-line diagnosis method for capacitor breakdown fault of capacitor voltage transformer

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210758583.9A CN114814501B (en) 2022-06-30 2022-06-30 On-line diagnosis method for capacitor breakdown fault of capacitor voltage transformer

Publications (2)

Publication Number Publication Date
CN114814501A true CN114814501A (en) 2022-07-29
CN114814501B CN114814501B (en) 2022-09-16

Family

ID=82523552

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210758583.9A Active CN114814501B (en) 2022-06-30 2022-06-30 On-line diagnosis method for capacitor breakdown fault of capacitor voltage transformer

Country Status (1)

Country Link
CN (1) CN114814501B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115267641A (en) * 2022-09-22 2022-11-01 武汉格蓝若智能技术有限公司 Method and system for identifying error abnormity of current transformer in same-tower double-circuit power transmission line
CN117169669A (en) * 2023-11-02 2023-12-05 国网江西省电力有限公司供电服务管理中心 Breakdown capacitance monitoring method and device, electronic equipment and readable storage medium

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106096076A (en) * 2016-05-26 2016-11-09 国网江苏省电力公司检修分公司 Capacitance type potential transformer operation troubles analogy method based on PSCAD
US20200309829A1 (en) * 2019-04-01 2020-10-01 Dake He On-line monitoring system for the performance of the measurement equipment in the entire power grid based on wide-area synchronous measurement
CN113447783A (en) * 2021-08-30 2021-09-28 武汉格蓝若智能技术有限公司 Voltage transformer insulation fault identification model construction method and device
CN114492675A (en) * 2022-04-01 2022-05-13 武汉格蓝若智能技术有限公司 Intelligent fault cause diagnosis method for capacitor voltage transformer
CN114548188A (en) * 2022-04-21 2022-05-27 武汉格蓝若智能技术有限公司 Method for identifying operation mode of capacitive voltage transformer in transformer substation
CN114757305A (en) * 2022-06-13 2022-07-15 华中科技大学 Voltage transformer insulation fault identification method and system based on ensemble learning
CN114779029A (en) * 2022-06-16 2022-07-22 武汉格蓝若智能技术有限公司 CVT internal insulation online evaluation method fusing group redundancy association and structural parameters

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106096076A (en) * 2016-05-26 2016-11-09 国网江苏省电力公司检修分公司 Capacitance type potential transformer operation troubles analogy method based on PSCAD
US20200309829A1 (en) * 2019-04-01 2020-10-01 Dake He On-line monitoring system for the performance of the measurement equipment in the entire power grid based on wide-area synchronous measurement
CN113447783A (en) * 2021-08-30 2021-09-28 武汉格蓝若智能技术有限公司 Voltage transformer insulation fault identification model construction method and device
CN114492675A (en) * 2022-04-01 2022-05-13 武汉格蓝若智能技术有限公司 Intelligent fault cause diagnosis method for capacitor voltage transformer
CN114548188A (en) * 2022-04-21 2022-05-27 武汉格蓝若智能技术有限公司 Method for identifying operation mode of capacitive voltage transformer in transformer substation
CN114757305A (en) * 2022-06-13 2022-07-15 华中科技大学 Voltage transformer insulation fault identification method and system based on ensemble learning
CN114779029A (en) * 2022-06-16 2022-07-22 武汉格蓝若智能技术有限公司 CVT internal insulation online evaluation method fusing group redundancy association and structural parameters

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
席佳伟 等: "330KV电容式电压互感器分压电容击穿故障原因分析及预控措施", 《变压器》 *
邹林等: "基于相对比较法的容性设备绝缘监测方法", 《电力与电工》 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115267641A (en) * 2022-09-22 2022-11-01 武汉格蓝若智能技术有限公司 Method and system for identifying error abnormity of current transformer in same-tower double-circuit power transmission line
CN115267641B (en) * 2022-09-22 2023-01-17 武汉格蓝若智能技术有限公司 Method and system for identifying error abnormity of current transformer in same-tower double-circuit power transmission line
CN117169669A (en) * 2023-11-02 2023-12-05 国网江西省电力有限公司供电服务管理中心 Breakdown capacitance monitoring method and device, electronic equipment and readable storage medium
CN117169669B (en) * 2023-11-02 2024-03-12 国网江西省电力有限公司供电服务管理中心 Breakdown capacitance monitoring method and device, electronic equipment and readable storage medium

Also Published As

Publication number Publication date
CN114814501B (en) 2022-09-16

Similar Documents

Publication Publication Date Title
CN114814501B (en) On-line diagnosis method for capacitor breakdown fault of capacitor voltage transformer
CN109164343B (en) Transformer fault diagnosis method based on characteristic information quantization and weighted KNN
Zhang et al. Feature selection for high-dimensional machinery fault diagnosis data using multiple models and radial basis function networks
CN107505133A (en) The probability intelligent diagnosing method of rolling bearing fault based on adaptive M RVM
CN111722145A (en) Method for diagnosing slight fault of turn-to-turn short circuit of excitation winding of synchronous motor
CN109670242B (en) Transformer winding deformation unsupervised online monitoring method based on elliptical envelope curve
Li et al. Condition assessment of power transformers using a synthetic analysis method based on association rule and variable weight coefficients
CN109298225B (en) Automatic identification model system and method for abnormal state of voltage measurement data
CN109375010A (en) Winding in Power Transformer method for diagnosing faults based on vibration distribution characteristic model
CN109657720B (en) On-line diagnosis method for turn-to-turn short circuit fault of power transformer
CN110197222A (en) A method of based on multi-category support vector machines transformer fault diagnosis
CN109376626A (en) A kind of GIS switch defect diagnostic method based on radiated electric field characteristic parameter support vector machines
CN114966295A (en) Super capacitor operation power frequency modulation detection system based on big data
CN113987294A (en) CVT (continuously variable transmission) online fault diagnosis method based on genetic optimization GRU (generalized regression Unit) neural network
Pandarakone et al. Frequency spectrum investigation and analytical diagnosis method for turn-to-turn short-circuit insulation failure in stator winding of low voltage induction motor
Mao et al. Winding type recognition through supervised machine learning using frequency response analysis (FRA) data
CN115759761A (en) Intelligent operation data management system for electric energy metering device
CN115511387A (en) Oil immersed transformer state evaluation method and system based on multi-source information fusion
CN117849558B (en) Method and device for on-line monitoring breakdown fault of main capacitor of capacitive voltage transformer
CN113866552B (en) Medium voltage distribution network user electricity consumption abnormality diagnosis method based on machine learning
CN117131786B (en) Voltage transformer insulation fault online identification method
CN112801135B (en) Fault line selection method and device for power plant service power system based on characteristic quantity correlation
CN116520068B (en) Diagnostic method, device, equipment and storage medium for electric power data
CN109784777B (en) Power grid equipment state evaluation method based on time sequence information fragment cloud similarity measurement
CN116975389A (en) Transformer substation fault analysis method and device

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