CN109374063A - A kind of transformer exception detection method, device and equipment based on cluster management - Google Patents
A kind of transformer exception detection method, device and equipment based on cluster management Download PDFInfo
- Publication number
- CN109374063A CN109374063A CN201811475724.6A CN201811475724A CN109374063A CN 109374063 A CN109374063 A CN 109374063A CN 201811475724 A CN201811475724 A CN 201811475724A CN 109374063 A CN109374063 A CN 109374063A
- Authority
- CN
- China
- Prior art keywords
- cluster
- transformer
- analyzed
- accounting
- initial
- 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
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The transformer exception detection method based on cluster management that this application discloses a kind of, device and equipment, using hierarchical clustering algorithm by the first basic parameter, i.e. account parameter and the similar transformer of floor data are partitioned into same transformer cluster cluster, there is similar state between each cluster transformer on this basis, pass through the second basic parameter of transformer in same cluster cluster again, that is the otherness for mutually knowing quite well equipment in cluster of Condition Monitoring Data, it can judge which platform transformer is in abnormality rapidly, the exception that EARLY RECOGNITION goes out transformer can occur in transformer fault, and abnormality detection accuracy rate is high.
Description
Technical field
This application involves transformer equipment operation troubles diagnostic techniques field more particularly to a kind of transformations based on cluster management
Device method for detecting abnormality, device and equipment.
Background technique
Power transformer is one of most important power transmission and transforming equipment in electric system, and operating status is directly related to entirely
The safety and stablization of electric system, therefore ensure that the reliability service of transformer is most important.
Currently, be that the kernel state amount of transformer is monitored online to the method for the detection of transformer state exception,
Threshold decision is carried out to these kernel state amounts based on national standard simultaneously, so that whether the state for diagnosing transformer is abnormal,
Method line monitoring transformer state amount and carry out threshold value comparison, although can ensure that the safety of transformer is steady to a certain extent
Fixed operation, still, actually when transformer core quantity of state is more than the threshold value of national standard setting, the state of transformer may
It has occurred that variation, affects the safety and stablization of electric system to a certain extent, therefore, how to be sent out in transformer
Before raw failure, the exception of transformer is identified, for ensureing that transformer safety stable operation is significant.
Summary of the invention
The embodiment of the present application provides a kind of transformer exception detection method, device and equipment based on cluster management, energy
Enough exceptions that transformer is identified before transformer breaks down, ensure the safe and stable operation of transformer.
In view of this, the application first aspect provides a kind of transformer exception detection method based on cluster management, packet
It includes:
Using transformer to be analyzed described in each of transformer to be analyzed of the first quantity as an initial transformer
Cluster obtains the first basic parameter of each initial transformer cluster, and first basic parameter includes: transformer production
Producer, military service place mean temperature and humidity, transformer rated power, transformer voltage rating and military service duration；
Based on first basic parameter, according to initial transformer cluster described in Similarity measures formula calculating every two
First similarity carries out hierarchical clustering to the initial transformer cluster, several cluster clusters is obtained, until the cluster cluster
The summation of number and the initial transformer number of clusters not clustered accounts for the accounting of first quantity, accounts for less than or equal to preset
Than when, stop cluster；
The second basic parameter is carried out to the transformer to be analyzed in the cluster cluster mutually to compare, and is judged described to be analyzed
Whether transformer abnormal, second basic parameter include: Gases Dissolved in Transformer Oil H2, CH4, C2H2, C2H4, C2H6,
CO, CO2, oil temperature and shelf depreciation.
Preferably, described to be based on first basic parameter, it is initial according to Similarity measures formula calculating every two
The first similarity of transformer cluster carries out hierarchical clustering to the initial transformer cluster, obtains several cluster clusters, until
The cluster number of clusters accounts for the accounting of first quantity with the summation for the initial transformer number of clusters not clustered, small
When being equal to preset accounting, stops cluster, specifically includes:
Based on first basic parameter, according to initial transformer cluster described in Similarity measures formula calculating every two
First similarity merges into the initial transformer cluster of the first similarity maximum first and the second initial transformer cluster
Cluster cluster；
The number for calculating the cluster cluster accounts for described the with the summation of the initial transformer number of clusters not clustered
First accounting of one quantity stops cluster if first accounting is less than or equal to preset accounting, otherwise, continues to described initial
Transformer cluster is clustered, and several cluster clusters are obtained, judge it is described it is several cluster cluster number with do not cluster it is described
The summation of initial transformer number of clusters accounts for the second accounting of first quantity, if second accounting is less than or equal to preset account for
Than stopping cluster.
Preferably, second basic parameter of the transformer progress to be analyzed in the cluster cluster mutually compares, and sentences
Whether the transformer to be analyzed that breaks abnormal, second basic parameter include: Gases Dissolved in Transformer Oil H2, CH4,
C2H2, C2H4, C2H6, CO, CO2, oil temperature and shelf depreciation, specifically include:
Second basic parameter is normalized；
It is calculated according to preset distance calculation formula between the transformer to be analyzed described two-by-two in the same cluster cluster
Distance；
By any one of transformer i to be analyzed in each cluster cluster, if where the transformer i to be analyzed
The cluster cluster in, remaining described band analysis transformer is with the transformer i to be analyzed at a distance from more than or equal to apart from threshold
The total quantity of value accounts for the accounting threshold value of each cluster cluster or more, then judges that the transformer i to be analyzed is abnormal.
Preferably, the Similarity measures formula are as follows:
Wherein, SijFor the fruit watt coefficient of transformer i and transformer j, SijkIt is transformer i and transformer j in data akOn
Value, akFor the first basic parameter, WijkFor weight variable, if data akFor dichotomic variable, 1-1 matches clock synchronization Sijk=1, other are matched
Clock synchronization Sijk=0；0-0 matches clock synchronization Wijk=0, other match clock synchronization Wijk=1, if data akFor variable of arranging in order, two transformer data phases
1 is taken simultaneously, otherwise takes 0；If data akFor numerical variable, Sijk=1- | aik-ajk|/Rk, wherein aikAnd ajkRespectively transformer i
With transformer j in variable akOn value, RkFor variable akRange.
Preferably, the preset accounting are as follows: 10%.
Preferably, the preset distance calculation formula are as follows:
Wherein, dijFor the Euclidean distance of transformer i and transformer j, b 'ikIt is of i-th transformer after normalized
K class online monitoring data.
Preferably, the accounting threshold value are as follows: 80%.
The application second aspect provides a kind of transformer exception detection device based on cluster management, comprising:
Module is obtained, for using transformer to be analyzed described in each of transformer to be analyzed of the first quantity as one
A initial transformer cluster obtains the first basic parameter of each initial transformer cluster, the first basic parameter packet
It includes: transformer production producer, military service place mean temperature and humidity, transformer rated power, transformer voltage rating and military service
Duration；
Cluster module, for being based on first basic parameter, according to Similarity measures formula calculating every two just
The first similarity of beginning transformer cluster carries out hierarchical clustering to the initial transformer cluster, obtains several cluster clusters, directly
The accounting of first quantity is accounted for the summation for the initial transformer number of clusters not clustered to the cluster number of clusters,
When less than or equal to preset accounting, stop cluster；
Judgment module mutually compares for carrying out the second basic parameter to the transformer to be analyzed in the cluster cluster,
Judge whether the transformer to be analyzed abnormal, second basic parameter include: Gases Dissolved in Transformer Oil H2, CH4,
C2H2, C2H4, C2H6, CO, CO2, oil temperature and shelf depreciation.
Preferably, the cluster module specifically includes: the first cluster module and the second cluster module；
First cluster module calculates every two according to Similarity measures formula for being based on first basic parameter
The first similarity of a initial transformer cluster, by the initial transformer cluster of the first similarity maximum first and
Two initial transformer clusters merge into cluster cluster；
Second cluster module, for calculating the number of the cluster cluster and the initial transformer collection not clustered
The summation of group's number accounts for the first accounting of first quantity, if first accounting is less than or equal to preset accounting, stops cluster,
Otherwise, continue to cluster the initial transformer cluster, obtain several cluster clusters, judge several cluster clusters
The summation of number and the initial transformer number of clusters not clustered accounts for the second accounting of first quantity, if described second
Accounting is less than or equal to preset accounting, stops cluster；
The judgment module specifically includes: normalization module, spacing module and detection module；
The normalization module, for the second basic parameter to be normalized；
The spacing module, it is described two-by-two in the same cluster cluster for being calculated according to preset distance calculation formula
The distance between transformer to be analyzed；
The detection module, for by it is each it is described cluster cluster in any one of transformer i to be analyzed, if described
In the cluster cluster where transformer i to be analyzed, remaining described band analysis transformer and the transformer i to be analyzed away from
From the total quantity for being more than or equal to distance threshold, the accounting threshold value of each cluster cluster or more is accounted for, then judges the change to be analyzed
Depressor i is abnormal.
The application third aspect provides a kind of transformer exception detection device of cluster management, and the equipment includes processing
Device and memory；
Said program code is transferred to the processor for storing program code by the memory；
The processor is used for according to the instruction execution first aspect in said program code based on cluster management
Transformer exception detection method.
As can be seen from the above technical solutions, the embodiment of the present application has the advantage that
In the application, a kind of transformer exception detection method based on cluster management is provided, comprising: by the first quantity
Each of transformer to be analyzed transformer to be analyzed obtains each initial transformer collection as an initial transformer cluster
The first basic parameter of group, the first basic parameter includes: transformer production producer, military service place mean temperature and humidity, transformation
Device rated power, transformer voltage rating and military service duration；Based on the first basic parameter, calculated according to Similarity measures formula every
The first similarity of two initial transformer clusters carries out hierarchical clustering to initial transformer cluster, obtains several cluster clusters,
Until the summation of cluster number of clusters and the initial transformer number of clusters not clustered accounts for the accounting of the first quantity, it is less than or equal to pre-
When setting accounting, stop cluster；The second basic parameter is carried out to the transformer to be analyzed in cluster cluster mutually to compare, and judges change to be analyzed
Whether depressor abnormal, the second basic parameter include: Gases Dissolved in Transformer Oil H2, CH4, C2H2, C2H4, C2H6, CO, CO2,
Oil temperature and shelf depreciation.Method provided by the present application, using hierarchical clustering algorithm by the first basic parameter, i.e. account parameter and work
The similar transformer of condition data is partitioned into same transformer cluster cluster, has between each cluster transformer on this basis
There is similar state, then by the second basic parameter of transformer in same cluster cluster, i.e. the mutual of Condition Monitoring Data compares
The otherness for solving equipment in cluster, can judge rapidly which platform transformer is in abnormality, early stage can occur in transformer fault
Identify the exception of transformer, and abnormality detection accuracy rate is high.
Detailed description of the invention
Fig. 1 is the process signal of transformer exception detection method of one of the embodiment of the present application based on cluster management
Figure；
Fig. 2 is another process signal of transformer exception detection method of one of the embodiment of the present application based on cluster management
Figure；
Fig. 3 is the structural representation of transformer exception detection device of one of the embodiment of the present application based on cluster management
Figure.
Specific embodiment
In order to make those skilled in the art more fully understand application scheme, below in conjunction in the embodiment of the present application
Attached drawing, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described embodiment is only this
Apply for a part of the embodiment, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art exist
Every other embodiment obtained under the premise of creative work is not made, shall fall in the protection scope of this application.
In order to make it easy to understand, referring to Fig. 1, a kind of transformer based on cluster management provided in the embodiment of the present application is different
Normal detection method, comprising:
Step 101 initially becomes using each of the transformer to be analyzed of the first quantity transformer to be analyzed as one
Depressor cluster obtains the first basic parameter of each initial transformer cluster, and the first basic parameter includes: transformer production factory
Family, military service place mean temperature and humidity, transformer rated power, transformer voltage rating and military service duration.
It should be noted that the data of transformer are divided into the first basic parameter a in the embodiment of the present applicationkIt is basic with second
Parameter bkTwo classes, the first basic parameter akIt is averaged for transformer account supplemental characteristic, including transformer production producer, military service place
Temperature and humidity, transformer rated power, transformer voltage rating and six kinds of military service duration, are denoted as a respectively1~a6.Second is basic
Parameter bkFor transformer state detection data, comprising: Gases Dissolved in Transformer Oil H2, CH4, C2H2, C2H4, C2H6, CO,
CO2, oil temperature and nine kinds of shelf depreciation, respectively as b1~b9.It is analysed to the transformer conduct to be analyzed of each of transformer
One initial transformer cluster, i.e. transformer i are represented by xi={ ai1,ai2,...,ai6, wherein xiFor i-th transformer
First basic parameter data, aijFor the jth class basic parameter data of i-th transformer, 1≤j≤6.
Step 102 is based on the first basic parameter, calculates the initial transformer cluster of every two according to Similarity measures formula
First similarity carries out hierarchical clustering to initial transformer cluster, obtains several cluster clusters, up to cluster number of clusters and not
The summation of the initial transformer number of clusters of cluster accounts for the accounting of the first quantity, when being less than or equal to preset accounting, stops cluster.
It should be noted that in the embodiment of the present application, the first basic parameter data based on transformer to be analyzed are treated point
It analyses transformer and carries out clustering, be analysed to transformer and be divided into different transformer clusters, used in the embodiment of the present application
It is hierarchical clustering algorithm.For example, the quantity of transformer to be analyzed is 20, this 20 transformers to be analyzed are carried out similar two-by-two
Property calculate, carry out first time Similarity measures after, by maximum two transformers to be analyzed of similitude, for example, be the 10th and
11st, a cluster cluster is merged into, at this moment, transformer number of clusters becomes 19 by 20, judges whether 19/20 is full
Foot is less than or equal to the condition of preset accounting, if it is satisfied, then stop cluster, if conditions are not met, then continue second of cluster,
The result of second of cluster, it may be possible to which the similitude of the 15th transformer to be analyzed and the 11st transformer to be analyzed is maximum, then
15th transformer to be analyzed is incorporated to the 10th transformer to be analyzed and the 11st transformer institute to be analyzed when clustering for the first time
Cluster cluster in, it is also possible to the similitude of the 15th transformer to be analyzed and the 16th transformer to be analyzed is maximum, then
15th transformer to be analyzed and the 16th transformer to be analyzed are merged into a cluster cluster, at this moment, transformer to be analyzed
Quantity become 18 by 19, judge whether 18/20 meet the condition less than or equal to preset accounting, if it is satisfied, then stopping
It only clusters, if conditions are not met, then continuing to cluster.
Step 103 mutually compares transformer to be analyzed the second basic parameter of progress in cluster cluster, judges transformation to be analyzed
Whether device is abnormal, and the second basic parameter includes: Gases Dissolved in Transformer Oil H2, CH4, C2H2, C2H4, C2H6, CO, CO2, oil
Mild shelf depreciation.
It should be noted that in the embodiment of the present application, based on several cluster clusters obtained in step 102, by right
Transformer to be analyzed in same cluster cluster mutually relatively carry out abnormality detection to transformer to be analyzed.If same poly-
Transformer to be analyzed in type of cluster has occurred more than expected variation, then it is assumed that corresponding transformer to be analyzed has occurred different
Often, alarming processing is carried out, EARLY RECOGNITION can occur in transformer fault and go out transformer exception, ensure the safety and stability of transformer
Operation.
In the application, a kind of transformer exception detection method based on cluster management is provided, comprising: by the first quantity
Each of transformer to be analyzed transformer to be analyzed obtains each initial transformer collection as an initial transformer cluster
The first basic parameter of group, the first basic parameter includes: transformer production producer, military service place mean temperature and humidity, transformation
Device rated power, transformer voltage rating and military service duration；Based on the first basic parameter, calculated according to Similarity measures formula every
The first similarity of two initial transformer clusters carries out hierarchical clustering to initial transformer cluster, obtains several cluster clusters,
Until the summation of cluster number of clusters and the initial transformer number of clusters not clustered accounts for the accounting of the first quantity, it is less than or equal to pre-
When setting accounting, stop cluster；The second basic parameter is carried out to the transformer to be analyzed in cluster cluster mutually to compare, and judges change to be analyzed
Whether depressor abnormal, the second basic parameter include: Gases Dissolved in Transformer Oil H2, CH4, C2H2, C2H4, C2H6, CO, CO2,
Oil temperature and shelf depreciation.Method provided by the present application, using hierarchical clustering algorithm by the first basic parameter, i.e. account parameter and work
The similar transformer of condition data is partitioned into same transformer cluster cluster, has between each cluster transformer on this basis
There is similar state, then by the second basic parameter of transformer in same cluster cluster, i.e. the mutual of Condition Monitoring Data compares
The otherness for solving equipment in cluster, can judge rapidly which platform transformer is in abnormality, early stage can occur in transformer fault
Identify the exception of transformer, and abnormality detection accuracy rate is high.
In order to make it easy to understand, referring to Fig. 2, another kind is examined based on the transformer exception of cluster management in the embodiment of the present application
Survey method, comprising:
Step 201 initially becomes using each of the transformer to be analyzed of the first quantity transformer to be analyzed as one
Depressor cluster obtains the first basic parameter of each initial transformer cluster, and the first basic parameter includes: transformer production factory
Family, military service place mean temperature and humidity, transformer rated power, transformer voltage rating and military service duration.
It should be noted that the step 101 in step 201 and preceding embodiment is consistent, no longer it is described in detail herein.
Step 202 is based on the first basic parameter, calculates the initial transformer cluster of every two according to Similarity measures formula
The initial transformer cluster of first similarity maximum first and the second initial transformer cluster are merged into cluster by first similarity
Cluster.
The summation of step 203, the number for calculating cluster cluster and the initial transformer number of clusters not clustered accounts for the first number
Amount the first accounting, if the first accounting be less than or equal to preset accounting, stop cluster, otherwise, continue to initial transformer cluster into
Row cluster, obtains several cluster clusters, judges the number of several cluster clusters and the initial transformer number of clusters not clustered
Summation accounts for the second accounting of the first quantity, if the second accounting is less than or equal to preset accounting, stops cluster.
Further, Similarity measures formula are as follows:
Wherein, SijFor the fruit watt coefficient of transformer i and transformer j, SijkIt is transformer i and transformer j in data akOn
Value, akFor the first basic parameter, WijkFor weight variable, if data akFor dichotomic variable, 1-1 matches clock synchronization Sijk=1, other are matched
Clock synchronization Sijk=0；0-0 matches clock synchronization Wijk=0, other match clock synchronization Wijk=1, if data akFor variable of arranging in order, two transformer data phases
1 is taken simultaneously, otherwise takes 0；If data akFor numerical variable, Sijk=1- | aik-ajk|/Rk, wherein aikAnd ajkRespectively transformer i
With transformer j in variable akOn value, RkFor variable akRange.
Further, preset accounting are as follows: 10%.
It should be noted that in the embodiment of the present application, the first basic parameter data based on transformer to be analyzed are treated point
It analyses transformer and carries out clustering, be analysed to transformer and be divided into different transformer clusters, used in the embodiment of the present application
It is hierarchical clustering algorithm.For example, the quantity of transformer to be analyzed is 20, this 20 transformers to be analyzed are carried out similar two-by-two
Property calculate, carry out first time Similarity measures after, by maximum two transformers to be analyzed of similitude, for example, be the 10th and
11st, a cluster cluster is merged into, at this moment, transformer number of clusters becomes 19 by 20, judges whether 19/20 is full
Foot is less than or equal to the condition of preset accounting, if it is satisfied, then stop cluster, if conditions are not met, then continue second of cluster,
The result of second of cluster, it may be possible to which the similitude of the 15th transformer to be analyzed and the 11st transformer to be analyzed is maximum, then
15th transformer to be analyzed is incorporated to the 10th transformer to be analyzed and the 11st transformer institute to be analyzed when clustering for the first time
Cluster cluster in, it is also possible to the similitude of the 15th transformer to be analyzed and the 16th transformer to be analyzed is maximum, then
15th transformer to be analyzed and the 16th transformer to be analyzed are merged into a cluster cluster, at this moment, transformer to be analyzed
Quantity become 18 by 19, judge whether 18/20 meet the condition less than or equal to preset accounting, if it is satisfied, then stopping
It only clusters, if conditions are not met, then continuing to cluster.
The second basic parameter is normalized in step 204.
It should be noted that in the embodiment of the present application, the second basic parameter include: Gases Dissolved in Transformer Oil H2,
CH4, C2H2, C2H4, C2H6, CO, CO2, oil temperature and shelf depreciation, are denoted as b respectively1~b9, poor to eliminate different index brings
It is different, to the second substrate parameter in each cluster cluster, i.e., the online monitoring data y of transformer to be analyzedi={ bi1,bi2,...,
bi9Be normalized respectively, normalized is as follows:
Wherein, yiFor the online monitoring data of i-th transformer to be analyzed, bijJth class for i-th transformer is supervised online
Measured data, n are the number of units of transformer in the cluster, b 'ijThe jth class of i-th transformer to be analyzed exists after expression normalized
Line monitoring data.
Step 205 calculates between the same transformer to be analyzed two-by-two clustered in cluster according to preset distance calculation formula
Distance.
Further, preset distance calculation formula are as follows:
Wherein, dijFor the Euclidean distance of transformer i and transformer j, b 'ikIt is of i-th transformer after normalized
K class online monitoring data.
Step 206, by any one transformer i to be analyzed in each cluster cluster, if poly- where transformer i to be analyzed
In type of cluster, remaining is more than or equal to the total quantity of distance threshold with analysis transformer at a distance from transformer i to be analyzed, accounts for each poly-
More than the accounting threshold value of type of cluster, then judge that transformer i to be analyzed is abnormal.
Further, accounting threshold value are as follows: 80%.
It should be noted that being to be set as 80% by accounting threshold value, in Mr. Yu cluster set group in the embodiment of the present application
Transformer i to be analyzed, if being greater than the transformer number to be analyzed of distance threshold in the cluster cluster at a distance from transformer i to be analyzed
Amount is more than the 80% of the number of units n of the cluster cluster transformer, then it is assumed that the transformer i to be analyzed is abnormal, otherwise it is assumed that should be wait divide
It is normal to analyse transformer i.
Transformer exception detection method provided by the embodiments of the present application based on cluster management will using hierarchical clustering algorithm
The similar transformer of account parameter, floor data is partitioned into same transformer cluster, each cluster transformation on this basis
There ought to be similar state between device, then be known quite well in cluster by the mutual of Transformer's Condition Monitoring data in same cluster
The otherness of equipment can judge rapidly which platform transformer is in abnormality, and abnormality detection accuracy rate is high.
Transformer exception detection method provided by the embodiments of the present application based on cluster management is with the transformation in same cluster
Device can occur EARLY RECOGNITION in transformer fault and go out the different of transformer as benchmark by the state of mutual multilevel iudge transformer
Often, for ensureing that transformer safety stable operation is significant.
Transformer exception detection method provided by the embodiments of the present application based on cluster management draws the concept of cluster management
Enter in transformer exception detection, similar transformer is returned and forms each transformer cluster together, is conducive to the length of transformer
Phase supervision.
In order to make it easy to understand, referring to Fig. 3, the embodiment of the present application provides a kind of transformer exception based on cluster management
Detection device, comprising:
Module 301 is obtained, for using each of the transformer to be analyzed of the first quantity transformer to be analyzed as one
A initial transformer cluster obtains the first basic parameter of each initial transformer cluster, and the first basic parameter includes: transformer
Manufacturer, military service place mean temperature and humidity, transformer rated power, transformer voltage rating and military service duration.
Cluster module 302 calculates every two transformer collection according to Similarity measures formula for being based on the first basic parameter
The first similarity of group carries out hierarchical clustering to initial transformer cluster, obtains several cluster clusters, until cluster number of clusters
The accounting of the first quantity is accounted for the summation for the initial transformer number of clusters not clustered, when being less than or equal to preset accounting, is stopped poly-
Class.
Judgment module 303, for cluster cluster in transformer to be analyzed carry out the second basic parameter mutually compare, judge to
Whether abnormal analyze transformer, the second basic parameter include: Gases Dissolved in Transformer Oil H2, CH4, C2H2, C2H4, C2H6,
CO, CO2, oil temperature and shelf depreciation.
Cluster module 302 specifically includes: the first cluster module 3021 and the second cluster module 3022.
First cluster module 3021, for being based on the first basic parameter, at the beginning of calculating every two according to Similarity measures formula
The first similarity of beginning transformer cluster, by the initial transformer cluster of first similarity maximum first and the second initial transformer
Cluster merges into cluster cluster.
Second cluster module 3022, for calculating the number of cluster cluster and the initial transformer number of clusters not clustered
Summation accounts for the first accounting of the first quantity, if the first accounting is less than or equal to preset accounting, stops cluster, otherwise, continues to initial
Transformer cluster is clustered, and several cluster clusters are obtained, and judges the number of several cluster clusters and the initial transformation not clustered
The summation of device number of clusters accounts for the second accounting of the first quantity, if the second accounting is less than or equal to preset accounting, stops cluster.
Judgment module 303 specifically includes: normalization module 3031, spacing module 3032 and detection module 3033；
Module 3031 is normalized, for the second basic parameter to be normalized；
Spacing module 3032, for calculating the change to be analyzed two-by-two in same cluster cluster according to preset distance calculation formula
The distance between depressor；
Detection module 3033, for by it is each cluster cluster in any one transformer i to be analyzed, if transformer to be analyzed
In cluster cluster where i, remaining is more than or equal to the sum of distance threshold with analysis transformer at a distance from transformer i to be analyzed
Amount accounts for the accounting threshold value of each cluster cluster or more, then judges that transformer i to be analyzed is abnormal.
Further, Similarity measures formula are as follows:
Wherein, SijFor the fruit watt coefficient of transformer i and transformer j, SijkIt is transformer i and transformer j in data akOn
Value, akFor the first basic parameter, WijkFor weight variable, if data akFor dichotomic variable, 1-1 matches clock synchronization Sijk=1, other are matched
Clock synchronization Sijk=0；0-0 matches clock synchronization Wijk=0, other match clock synchronization Wijk=1, if data akFor variable of arranging in order, two transformer data phases
1 is taken simultaneously, otherwise takes 0；If data akFor numerical variable, Sijk=1- | aik-ajk|/Rk, wherein aikAnd ajkRespectively transformer i
With transformer j in variable akOn value, RkFor variable akRange.
Further, preset accounting are as follows: 10%.
Further, preset distance calculation formula are as follows:
Wherein, dijFor the Euclidean distance of transformer i and transformer j, b 'ikIt is of i-th transformer after normalized
K class online monitoring data.
Further, accounting threshold value are as follows: 80%.
A kind of transformer exception detection device based on cluster management is provided in the embodiment of the present application, equipment includes processing
Device and memory:
Program code is transferred to processor for storing program code by memory；
Processor is used for according to the transformation based on cluster management in the instruction execution embodiment above-mentioned in program code
Device method for detecting abnormality.
The description of the present application and term " first " in above-mentioned attached drawing, " second ", " third ", " the 4th " etc. are (if deposited
) it is to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should be understood that use in this way
Data are interchangeable under appropriate circumstances, so that embodiments herein described herein for example can be in addition to illustrating herein
Or the sequence other than those of description is implemented.In addition, term " includes " and " having " and their any deformation, it is intended that
Covering non-exclusive includes to be not necessarily limited to clearly for example, containing the process, method of a series of steps or units, product or equipment
Those of list to Chu step or unit, but may include be not clearly listed or for these process, methods, product or
The intrinsic other step or units of equipment.
In several embodiments provided herein, it should be understood that disclosed system, device and method can be with
It realizes by another way.For example, the apparatus embodiments described above are merely exemplary, for example, the unit
It divides, only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components
It can be combined or can be integrated into another system, or some features can be ignored or not executed.Another point, it is shown or
The mutual coupling, direct-coupling or communication connection discussed can be through some interfaces, the indirect coupling of device or unit
It closes or communicates to connect, can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, each functional unit in each embodiment of the application can integrate in one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, the technical solution of the application is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) executes the complete of each embodiment the method for the application
Portion or part steps.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (full name in English: Read-Only
Memory, english abbreviation: ROM), random access memory (full name in English: Random Access Memory, english abbreviation:
RAM), the various media that can store program code such as magnetic or disk.
The above, above embodiments are only to illustrate the technical solution of the application, rather than its limitations；Although referring to before
Embodiment is stated the application is described in detail, those skilled in the art should understand that: it still can be to preceding
Technical solution documented by each embodiment is stated to modify or equivalent replacement of some of the technical features；And these
It modifies or replaces, the spirit and scope of each embodiment technical solution of the application that it does not separate the essence of the corresponding technical solution.
Claims (10)
1. a kind of transformer exception detection method based on cluster management characterized by comprising
Using transformer to be analyzed described in each of transformer to be analyzed of the first quantity as an initial transformer cluster,
Obtain the first basic parameter of each initial transformer cluster, first basic parameter include: transformer production producer,
Military service place mean temperature and humidity, transformer rated power, transformer voltage rating and military service duration；
Based on first basic parameter, according to first of initial transformer cluster described in Similarity measures formula calculating every two
Similitude carries out hierarchical clustering to the initial transformer cluster, several cluster clusters is obtained, until the cluster number of clusters
The accounting of first quantity is accounted for the summation for the initial transformer number of clusters not clustered, is less than or equal to preset accounting
When, stop cluster；
The second basic parameter is carried out to the transformer to be analyzed in the cluster cluster mutually to compare, and judges the transformation to be analyzed
Whether device abnormal, second basic parameter include: Gases Dissolved in Transformer Oil H2, CH4, C2H2, C2H4, C2H6, CO,
CO2, oil temperature and shelf depreciation.
2. the transformer exception detection method according to claim 1 based on cluster management, which is characterized in that described to be based on
First basic parameter, according to Similarity measures formula calculate every two described in initial transformer cluster first similarity,
Hierarchical clustering is carried out to the initial transformer cluster, obtains several cluster clusters, until the cluster number of clusters with do not gather
The summation of the initial transformer number of clusters of class accounts for the accounting of first quantity, when being less than or equal to preset accounting, stops
Cluster, specifically includes:
Based on first basic parameter, according to first of initial transformer cluster described in Similarity measures formula calculating every two
The initial transformer cluster of the first similarity maximum first and the second initial transformer cluster are merged into cluster by similitude
Cluster；
The summation of the number and the initial transformer number of clusters not clustered that calculate the cluster cluster accounts for first number
First accounting of amount stops cluster, otherwise, continues to the initial transformation if first accounting is less than or equal to preset accounting
Device cluster is clustered, and several cluster clusters are obtained, and is judged the number of several cluster clusters and is not clustered described initial
The summation of transformer number of clusters accounts for the second accounting of first quantity, if second accounting is less than or equal to preset accounting,
Stop cluster.
3. the transformer exception detection method according to claim 1 based on cluster management, which is characterized in that described to institute
Second basic parameter of the transformer progress to be analyzed stated in cluster cluster mutually compares, and judges whether the transformer to be analyzed is different
Often, second basic parameter include: Gases Dissolved in Transformer Oil H2, CH4, C2H2, C2H4, C2H6, CO, CO2, oil temperature and
Shelf depreciation specifically includes:
Second basic parameter is normalized；
According to preset distance calculation formula calculate it is same it is described cluster cluster in the transformer to be analyzed described two-by-two between away from
From；
By any one of transformer i to be analyzed in each cluster cluster, if the institute where the transformer i to be analyzed
It states in cluster cluster, remaining described band analysis transformer is at a distance from the transformer i to be analyzed more than or equal to distance threshold
Total quantity accounts for the accounting threshold value of each cluster cluster or more, then judges that the transformer i to be analyzed is abnormal.
4. the transformer exception detection method according to claim 1 based on cluster management, which is characterized in that described similar
Property calculation formula are as follows:
Wherein, SijFor the fruit watt coefficient of transformer i and transformer j, SijkIt is transformer i and transformer j in data akOn take
Value, akFor the first basic parameter, WijkFor weight variable, if data akFor dichotomic variable, 1-1 matches clock synchronization Sijk=1, other pairings
When Sijk=0；0-0 matches clock synchronization Wijk=0, other match clock synchronization Wijk=1, if data akFor variable of arranging in order, two transformer data are identical
When take 1, otherwise take 0；If data akFor numerical variable, Sijk=1- | aik-ajk|/Rk, wherein aikAnd ajkRespectively transformer i and
Transformer j is in variable akOn value, RkFor variable akRange.
5. the transformer exception detection method according to claim 1 based on cluster management, which is characterized in that described preset
Accounting are as follows: 10%.
6. the transformer exception detection method according to claim 3 based on cluster management, which is characterized in that described preset
Distance calculation formula are as follows:
Wherein, dijFor the Euclidean distance of transformer i and transformer j, b 'ikKth class for i-th transformer after normalized exists
Line monitoring data.
7. the transformer exception detection method according to claim 3 based on cluster management, which is characterized in that the accounting
Threshold value are as follows: 80%.
8. a kind of transformer exception detection device based on cluster management characterized by comprising
Module is obtained, for using transformer to be analyzed described in each of transformer to be analyzed of the first quantity as at the beginning of one
Beginning transformer cluster, obtains the first basic parameter of each initial transformer cluster, and first basic parameter includes: to become
Depressor manufacturer, military service place mean temperature and humidity, transformer rated power, transformer voltage rating and military service duration；
Cluster module initially becomes according to Similarity measures formula calculating every two for being based on first basic parameter
The first similarity of depressor cluster carries out hierarchical clustering to the initial transformer cluster, several cluster clusters is obtained, until institute
The accounting that cluster number of clusters accounts for first quantity with the summation of the initial transformer number of clusters not clustered is stated, is less than
When equal to preset accounting, stop cluster；
Judgment module mutually compares for carrying out the second basic parameter to the transformer to be analyzed in the cluster cluster, judges
Whether the transformer to be analyzed abnormal, second basic parameter include: Gases Dissolved in Transformer Oil H2, CH4, C2H2,
C2H4, C2H6, CO, CO2, oil temperature and shelf depreciation.
9. the transformer exception detection device according to claim 8 based on cluster management, which is characterized in that the cluster
Module specifically includes: the first cluster module and the second cluster module；
First cluster module calculates every two institute according to Similarity measures formula for being based on first basic parameter
The first similarity for stating initial transformer cluster, will be at the beginning of the initial transformer cluster of the first similarity maximum first and second
Beginning transformer cluster merges into cluster cluster；
Second cluster module, for calculating the number of the cluster cluster and the initial transformer cluster number not clustered
Purpose summation accounts for the first accounting of first quantity, if first accounting is less than or equal to preset accounting, stops cluster, no
Then, continue to cluster the initial transformer cluster, obtain several cluster clusters, judge the number of several cluster clusters
The summation of mesh and the initial transformer number of clusters not clustered accounts for the second accounting of first quantity, if described second accounts for
Than being less than or equal to preset accounting, stop cluster；
The judgment module specifically includes: normalization module, spacing module and detection module；
The normalization module, for the second basic parameter to be normalized；
The spacing module, it is described wait divide two-by-two in the same cluster cluster for being calculated according to preset distance calculation formula
Analyse the distance between transformer；
The detection module, for by it is each it is described cluster cluster in any one of transformer i to be analyzed, if it is described to point
In the cluster cluster where analysis transformer i, remaining described band analysis transformer is big at a distance from the transformer i to be analyzed
In the total quantity for being equal to distance threshold, the accounting threshold value of each cluster cluster or more is accounted for, then judges the transformer i to be analyzed
It is abnormal.
10. a kind of transformer exception detection device based on cluster management, which is characterized in that the equipment include processor and
Memory:
Said program code is transferred to the processor for storing program code by the memory；
The processor is used to be based on cluster according to the instruction execution claim 1-7 in said program code is described in any item
The transformer exception detection method of management.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811475724.6A CN109374063B (en) | 2018-12-04 | 2018-12-04 | Cluster management-based transformer anomaly detection method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811475724.6A CN109374063B (en) | 2018-12-04 | 2018-12-04 | Cluster management-based transformer anomaly detection method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109374063A true CN109374063A (en) | 2019-02-22 |
| CN109374063B CN109374063B (en) | 2021-04-23 |
Family
ID=65375738
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811475724.6A Active CN109374063B (en) | 2018-12-04 | 2018-12-04 | Cluster management-based transformer anomaly detection method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109374063B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110161382A (en) * | 2019-04-30 | 2019-08-23 | 国网江苏省电力有限公司电力科学研究院 | It is a kind of to judge whether transformer needs to have a power failure the method and apparatus of test |
| CN113077020A (en) * | 2021-06-07 | 2021-07-06 | 广东电网有限责任公司湛江供电局 | Transformer cluster management method and system |
| CN113625058A (en) * | 2021-08-13 | 2021-11-09 | 云南电网有限责任公司电力科学研究院 | Early warning method and system for monitoring transformer bushing based on abnormal point group |
| CN113791300A (en) * | 2021-11-15 | 2021-12-14 | 广东电网有限责任公司东莞供电局 | Charging gun state monitoring method and system for charging station |
| CN113826127A (en) * | 2019-05-23 | 2021-12-21 | 西门子股份公司 | Method and device for evaluating health state of transformer and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5103409A (en) * | 1989-01-09 | 1992-04-07 | Hitachi, Ltd. | Field measuring instrument and its abnormality managing method |
| JPH05336589A (en) * | 1992-05-28 | 1993-12-17 | Hitachi Ltd | Ultrasonic transmission equipment |
| CN105871634A (en) * | 2016-06-01 | 2016-08-17 | 北京蓝海讯通科技股份有限公司 | Method and application for detecting cluster anomalies and cluster managing system |
| CN106338673A (en) * | 2016-09-29 | 2017-01-18 | 国网山东省电力公司电力科学研究院 | Transformer winding fault characteristic parameter extraction method |
| CN106845728A (en) * | 2017-02-14 | 2017-06-13 | 北京邮电大学 | The Forecasting Methodology and device of a kind of power transformer defect |
| CN108629528A (en) * | 2018-06-20 | 2018-10-09 | 国网江苏省电力有限公司电力科学研究院 | Quality of Transformer problem analysis method based on Apriori algorithm |
-
2018
- 2018-12-04 CN CN201811475724.6A patent/CN109374063B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5103409A (en) * | 1989-01-09 | 1992-04-07 | Hitachi, Ltd. | Field measuring instrument and its abnormality managing method |
| JPH05336589A (en) * | 1992-05-28 | 1993-12-17 | Hitachi Ltd | Ultrasonic transmission equipment |
| CN105871634A (en) * | 2016-06-01 | 2016-08-17 | 北京蓝海讯通科技股份有限公司 | Method and application for detecting cluster anomalies and cluster managing system |
| CN106338673A (en) * | 2016-09-29 | 2017-01-18 | 国网山东省电力公司电力科学研究院 | Transformer winding fault characteristic parameter extraction method |
| CN106845728A (en) * | 2017-02-14 | 2017-06-13 | 北京邮电大学 | The Forecasting Methodology and device of a kind of power transformer defect |
| CN108629528A (en) * | 2018-06-20 | 2018-10-09 | 国网江苏省电力有限公司电力科学研究院 | Quality of Transformer problem analysis method based on Apriori algorithm |
Non-Patent Citations (1)
| Title |
|---|
| 王洪授等: "基于聚类主成分分析法的变压器参数关联度评估", 《火力与指挥控制》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110161382A (en) * | 2019-04-30 | 2019-08-23 | 国网江苏省电力有限公司电力科学研究院 | It is a kind of to judge whether transformer needs to have a power failure the method and apparatus of test |
| CN110161382B (en) * | 2019-04-30 | 2022-06-10 | 国网江苏省电力有限公司电力科学研究院 | Method and device for judging whether transformer needs power failure test or not |
| CN113826127A (en) * | 2019-05-23 | 2021-12-21 | 西门子股份公司 | Method and device for evaluating health state of transformer and storage medium |
| CN113077020A (en) * | 2021-06-07 | 2021-07-06 | 广东电网有限责任公司湛江供电局 | Transformer cluster management method and system |
| CN113625058A (en) * | 2021-08-13 | 2021-11-09 | 云南电网有限责任公司电力科学研究院 | Early warning method and system for monitoring transformer bushing based on abnormal point group |
| CN113791300A (en) * | 2021-11-15 | 2021-12-14 | 广东电网有限责任公司东莞供电局 | Charging gun state monitoring method and system for charging station |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109374063B (en) | 2021-04-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109374063A (en) | A kind of transformer exception detection method, device and equipment based on cluster management | |
| CN107301296B (en) | Data-based qualitative analysis method for circuit breaker fault influence factors | |
| CN106502907B (en) | A kind of distributed software abnormality diagnostic method based on execution track tracking | |
| CN112699913A (en) | Transformer area household variable relation abnormity diagnosis method and device | |
| CN106209405B (en) | Method for diagnosing faults and device | |
| CN106156485A (en) | Method for diagnosing fault of power transformer and device | |
| CN110297841B (en) | Transformer fault diagnosis and rapid indexing method and system | |
| CN108182523A (en) | The treating method and apparatus of fault data, computer readable storage medium | |
| CN105868106B (en) | The test method and test system of ultrasonic device reliability | |
| CN106569056A (en) | Power transformer fault diagnosis method and diagnosis device | |
| CN105955214B (en) | Batch process fault detection method based on sample time-series and neighbour's affinity information | |
| CN109409444B (en) | Multivariate power grid fault type discrimination method based on prior probability | |
| CN109389325B (en) | Method for evaluating state of electronic transformer of transformer substation based on wavelet neural network | |
| Koochi et al. | A synchrophasor-based decision tree approach for identification of most coherent generating units | |
| CN110428005A (en) | A kind of safe misclassification constrained procedure of Electrical Power System Dynamic based on umbrella-type algorithm | |
| CN108304567A (en) | High-tension transformer regime mode identifies and data classification method and system | |
| CN109239489A (en) | A kind of health status on-line analysis, device and medium for energy-storage system | |
| CN113125903A (en) | Line loss anomaly detection method, device, equipment and computer-readable storage medium | |
| CN113408548A (en) | Transformer abnormal data detection method and device, computer equipment and storage medium | |
| CN117312611A (en) | Rapid positioning and diagnosing method and related device for power faults | |
| CN111027218B (en) | Complex system risk priority number analysis and calculation method considering weight and correlation | |
| Ren et al. | A clustering-based method for health conditions evaluation of aero-engines | |
| Yao et al. | AdaBoost-CNN: a hybrid method for electricity theft detection | |
| CN113961708B (en) | Power equipment fault tracing method based on multi-level graph convolutional network | |
| Pan et al. | Study on intelligent anti–electricity stealing early-warning technology based on convolutional neural networks |
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 |