CN108226711A - A kind of electric network failure diagnosis method and device based on information fusion - Google Patents
A kind of electric network failure diagnosis method and device based on information fusion Download PDFInfo
- Publication number
- CN108226711A CN108226711A CN201810011115.9A CN201810011115A CN108226711A CN 108226711 A CN108226711 A CN 108226711A CN 201810011115 A CN201810011115 A CN 201810011115A CN 108226711 A CN108226711 A CN 108226711A
- Authority
- CN
- China
- Prior art keywords
- fault
- information
- probability
- suspected
- cluster
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/08—Locating faults in cables, transmission lines, or networks
Abstract
The present invention provides a kind of electric network failure diagnosis method and device based on information fusion, wherein method includes：Fault zone search is carried out to power grid, determines power supply interrupted district；The whole elements obtained in power supply interrupted district form fault element collection, obtain the fault message of whole elements, and fault message includes switching value information and electric quantity information；Information analysis is carried out to fault element collection based on Fuzzy Coloured Petri Nets, obtains the element fault probability of fault element concentration, information analysis is carried out to fault element collection based on HHT technologies are improved, obtains the element fault probability characterization of fault element concentration；Element fault probability is characterized using preset Fusion Model and is merged into row information, obtains fusion results；Utilize preset decision diagnostic model, classified according to fusion results to fault element, it determines the element to break down in power supply interrupted district, solves that there are malfunction, tripping and information uncertain factors, the accuracy of caused diagnostic result is not high the technical issues of such as to lose due to information.
Description
Technical field
The present invention relates to electric network failure diagnosis field more particularly to a kind of electric network failure diagnosis methods based on information fusion
And device.
Background technology
As the premise of social development, effective electric network failure diagnosis system can be fast for the operation that power system security is stablized
Speed is accurately diagnosed to be fault element, significant to the stable operation of power grid.
Major part electric network failure diagnosis method is diagnosed using information such as protection act, breaker hopscotch at present, but
Since information is there are uncertain factors such as malfunction, tripping and information loss, the not high technology of the accuracy of diagnostic result is resulted in
Problem.
Invention content
The present invention provides a kind of electric network failure diagnosis method and devices based on information fusion, solve current major part
Electric network failure diagnosis method is diagnosed using information such as protection act, breaker hopscotch, but due to information there are malfunction, refuse
The technical issues of uncertain factors such as dynamic and information loss, the accuracy of caused diagnostic result is not high.
The present invention provides it is a kind of based on information fusion electric network failure diagnosis method, including：
S1, fault zone search is carried out to power grid, determines power supply interrupted district；
S2, the whole elements obtained in power supply interrupted district form fault element collection, obtain the fault message of whole elements, failure
Information includes switching value information and electric quantity information；
S3, information analysis is carried out to fault element collection based on Fuzzy Coloured Petri Nets, obtains the element of fault element concentration
Probability of malfunction carries out information analysis based on HHT technologies are improved to fault element collection, and the element fault for obtaining fault element concentration is general
Rate characterizes；
S4, element fault probability is characterized using preset Fusion Model and is merged into row information, obtain fusion results；
S5, using preset decision diagnostic model, classified according to fusion results to fault element, determined in power supply interrupted district
The element to break down.
Preferably, carrying out information analysis to fault element collection based on Fuzzy Coloured Petri Nets in step S3, failure is obtained
The element fault probability that element is concentrated specifically includes：
According to network topological information, relaying configuration information and the protection of the power supply interrupted district of fault element concentration and breaker
Action logic, that establishes each initial storehouse institute is compounded with color table；
According to the network topological information of power supply interrupted district and protection and breaker warning message, power supply interrupted district is carried out further
Fault search determines suspected fault element and its corresponding suspected fault element alarm information in power supply interrupted district, and forms first
Suspected fault element collection；
Suspected fault element and its corresponding suspected fault element alarm information, structure are concentrated according to the first possible breakdown element
Inference temporal is built, the initial marking of inference temporal is determined, and inference temporal is solved, obtains being compounded with color table
In element fault feature and its corresponding warning information；
It actual act time and expectation actuation time section in the warning message of element fault feature, calculates respectively
Failure amount beginning confidence level, and using obtained each failure amount beginning confidence level as each in all color tables in Fuzzy Inference Model starting library
The initial confidence level of event determines the initial marking of Fuzzy Inference Model；
Fuzzy Inference Model is solved, determines that the fault element of the first suspected fault element concentration and its element fault are general
Rate.
Preferably, carrying out information analysis to fault element collection based on improvement HHT technologies in step S3, failure member is obtained
The element fault probability characterization that part is concentrated specifically includes：
According to static grid topology data and protection and breaker warning message, further failure is carried out to power supply interrupted district and is searched
Rope determines suspected fault element and its corresponding suspected fault element alarm information in power supply interrupted district, and forms the second suspicious event
Hinder element collection；
It obtains the second suspected fault element in fault recording system and concentrates current signal of the suspected fault element in relation to circuit,
Current signal is current waveform data；
To the empirical mode decomposition that the current signal got is improved, IMF components are obtained；
Hilbert transform is carried out to the IMF components after decomposition, corresponding instantaneous amplitude is calculated, and according to instantaneous width
Value forms marginal spectrum；
The HHT amplitudes degree of distortion of fault current and HHT frequency distortion degree are calculated, the HHT amplitudes degree of distortion and HHT frequencies is abnormal
Variation is characterized as element fault probability, and the failure member for determining that the second suspected fault element is concentrated is characterized according to element fault probability
Part.
Preferably, step S4 is specifically included：
Obtain the fault element and its element fault probability of the first suspected fault element concentration；
Obtain the fault element and its element fault probability of the second suspected fault element concentration；
The fault element and its element for being concentrated the first suspected fault element of acquisition using D-S evidence theory Fusion Model
The fault element and its element fault probability characterization fusion that probability of malfunction and the second suspected fault element are concentrated, obtain fusion knot
Fruit.
Preferably, step S5 is specifically included：
S501, the element fault probability obtained in fusion results characterize, and n element fault probability characterization is used m respectively
(F1), m (F2) ..., m (Fn) represent, and calculate the distance between each element fault probability characterization sample dij=| | m (Fi)-m
(Fj) | |, i ＜ j；
S502, the distance between sample is characterized according to each element fault probability, obtains each element fault probability characterization
The local density of sample, and generate the descending arrangement subscript sequence of local density's sequence；
S503, each element fault probability characterization sample and fault cluster central point and non-event are calculated according to gaussian kernel function
Hinder cluster centre point distance, and by element fault probability characterization sample be divided to distance closer to fault cluster or non-faulting gather
In class, primary fault cluster and initial non-faulting cluster are obtained；
S504, the sample average for calculating primary fault cluster calculate initial non-faulting and gather as primary fault cluster centre
The sample average of class is as initial non-faulting cluster centre；
S505, the error sum of squares that element fault probability characterizes in primary fault cluster and initial non-faulting cluster is calculated,
And classification results when error sum of squares to reach to minimum determine fault element as optimal value；
S506, calculating overall error quadratic sum, simultaneously return to step S503 enters iteration, until overall error quadratic sum is constant.
The present invention provides it is a kind of based on information fusion electric network failure diagnosis device, including：
Power supply interrupted district determination unit for carrying out fault zone search to power grid, determines power supply interrupted district；
Fault element collection construction unit forms fault element collection for obtaining whole elements in power supply interrupted district, obtains complete
The fault message of portion's element, fault message include switching value information and electric quantity information；
Storage unit carries out information analysis to fault element collection for being based on Fuzzy Coloured Petri Nets, obtains failure
The element fault probability that element is concentrated carries out information analysis to fault element collection based on HHT technologies are improved, obtains fault element collection
In element fault probability characterization；
Information fusion unit is merged for being characterized to element fault probability using preset Fusion Model into row information, obtained
Fusion results；
Fault element determination unit for utilizing preset decision diagnostic model, carries out fault element according to fusion results
Classification, determines the element to break down in power supply interrupted district.
Preferably, storage unit specifically includes：
Be compounded with color table structure subelement, for concentrated according to fault element the network topological information of power supply interrupted district, protect
Configuration information and protection and breaker actuation logic are protected, that establishes each initial storehouse institute is compounded with color table；
First suspected fault element collection establishes subelement, for the network topological information according to power supply interrupted district and protection and breaks
Road device warning message carries out further fault search to power supply interrupted district, determines suspected fault element and its correspondence in power supply interrupted district
Suspected fault element alarm information, and form the first suspected fault element collection；
Inference temporal structure solve subelement, for according to the first possible breakdown element concentrate suspected fault element and
Its corresponding suspected fault element alarm information builds inference temporal, determines the initial marking of inference temporal, and right
Inference temporal solves, and obtains being compounded with the element fault feature in color table and its corresponding warning information；
Fuzzy Inference Model builds computation subunit, for the actual act in the warning message according to element fault feature
Time and it is expected actuation time section, respectively calculate failure amount beginning confidence level, and using obtained each failure amount beginning confidence level as
The initial confidence level of each event in all color tables in Fuzzy Inference Model starting library determines the initial marking of Fuzzy Inference Model；
Fisrt fault element determination subelement for being solved to Fuzzy Inference Model, determines the first suspected fault element collection
In fault element and its element fault probability.
Preferably, storage unit specifically further includes：
Second suspected fault element collection establishes subelement, for being warned according to static grid topology data and protection with breaker
Information is accused, further fault search is carried out to power supply interrupted district, determines in power supply interrupted district suspected fault element and its corresponding suspicious
Fault element warning information, and form the second suspected fault element collection；
Current signal obtains subelement, and suspected fault is concentrated for obtaining the second suspected fault element in fault recording system
Current signal of the element in relation to circuit, current signal are current waveform data；
Empirical mode decomposition subelement for the empirical mode decomposition being improved to the current signal got, obtains
IMF components；
Hilbert transform subelement for carrying out Hilbert transform to the IMF components after decomposition, is calculated corresponding
Instantaneous amplitude, and according to instantaneous amplitude formed marginal spectrum；
Second fault element determination subelement calculates the HHT amplitudes degree of distortion of fault current and HHT frequency distortion degree, will
The HHT amplitudes degree of distortion and HHT frequency distortions degree are characterized as element fault probability, and second is determined according to element fault probability characterization
The fault element that suspected fault element is concentrated.
Preferably, information fusion unit specifically includes：
First obtains subelement, general for obtaining the fault element of the first suspected fault element concentration and its element fault
Rate；
Second obtains subelement, general for obtaining the fault element of the second suspected fault element concentration and its element fault
Rate；
Information merges subelement, for utilizing D-S evidence theory Fusion Model by the first suspected fault element collection of acquisition
In the fault element concentrated of fault element and its element fault probability and the second suspected fault element and its element fault probability
Characterization fusion, obtains fusion results.
Preferably, fault element determination unit specifically includes：
Third obtains subelement, for obtaining the characterization of the element fault probability in fusion results, by n element fault probability
Characterization uses m (F respectively1), m (F2) ..., m (Fn) represent, and calculate the distance between each element fault probability characterization sample dij
=| | m (Fi)-m(Fj) | |, i ＜ j；
Local density's sequence generating unit for characterizing the distance between sample according to each element fault probability, obtains
The local density of each element fault probability characterization sample, and generate the descending arrangement subscript sequence of local density's sequence；
Cluster generation subelement gathers for calculating each element fault probability characterization sample according to gaussian kernel function with failure
The distance of class central point and non-faulting cluster centre point, and by element fault probability characterization sample be divided to distance closer to failure
In cluster or non-faulting cluster, primary fault cluster and initial non-faulting cluster are obtained；
Cluster centre computation subunit, during the sample average for calculating primary fault cluster is clustered as primary fault
The heart calculates the sample average of initial non-faulting cluster as initial non-faulting cluster centre；
Fault element determination subelement, for calculating element fault probability in primary fault cluster and initial non-faulting cluster
The error sum of squares of characterization, and classification results when error sum of squares to reach to minimum determine fault element as optimal value；
Iteration subelement enters iteration, until total for calculating overall error quadratic sum and jumping to cluster generation subelement
Error sum of squares is constant.
As can be seen from the above technical solutions, the present invention has the following advantages：
The present invention provides it is a kind of based on information fusion electric network failure diagnosis method, including：S1, failure is carried out to power grid
Range searching determines power supply interrupted district；S2, the whole elements obtained in power supply interrupted district form fault element collection, obtain whole elements
Fault message, fault message include switching value information and electric quantity information；S3, based on Fuzzy Coloured Petri Nets to failure member
Part collection carry out information analysis, obtain fault element concentration element fault probability, based on improve HHT technologies to fault element collection into
Row information is analyzed, and obtains the element fault probability characterization of fault element concentration；It is S4, general to element fault using preset Fusion Model
Rate is characterized to be merged into row information, obtains fusion results；S5, using preset decision diagnostic model, according to fusion results to failure member
Part is classified, and determines the element to break down in power supply interrupted district.
The present invention obtains element fault probability based on Fuzzy Coloured Petri Nets, and element event is obtained based on HHT technologies are improved
Hinder probability characterization, merged two kinds of information by preset Fusion Model, recycle preset diagnosis decision model according to fusion results
Classify to fault element, determine the element to break down in power supply interrupted district, merged multiple data sources information, accuracy,
Higher in terms of real-time and fault-tolerance, it is to utilize protection act, open circuit to solve most of electric network failure diagnosis method at present
The information such as device hopscotch are diagnosed, but there are the uncertain factors such as malfunction, tripping and information loss, caused diagnosis due to information
As a result the technical issues of accuracy is not high.
Description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The embodiment of invention, for those of ordinary skill in the art, without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 is a kind of flow of one embodiment of electric network failure diagnosis method based on information fusion provided by the invention
Schematic diagram；
Fig. 2 is a kind of structure of one embodiment of electric network failure diagnosis device based on information fusion provided by the invention
Schematic diagram.
Specific embodiment
An embodiment of the present invention provides a kind of electric network failure diagnosis method and devices based on information fusion, solve at present
Most of electric network failure diagnosis method is diagnosed using information such as protection act, breaker hopscotch, but since information exists
The technical issues of uncertain factors such as malfunction, tripping and information loss, the accuracy of caused diagnostic result is not high.
Below in conjunction with the attached drawing in the embodiment of the present invention, the technical solution in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other without making creative work
Embodiment shall fall within the protection scope of the present invention.
Referring to Fig. 1, one an embodiment of the present invention provides a kind of electric network failure diagnosis method based on information fusion
Embodiment, including：
101st, fault zone search is carried out to power grid, determines power supply interrupted district；
102nd, the whole elements obtained in power supply interrupted district form fault element collection, obtain the fault message of whole elements, therefore
Hinder information and include switching value information and electric quantity information；
It should be noted that after grid collapses, fault message has a hierarchy, electric current, voltage etc. during electric network fault
Electric quantity information and switching value information can change, and can reflect grid faults characteristics.In general, from SCADA system
Switching value information is obtained, electric quantity information is obtained from fault recording system,
1030th, according to the network topological information of power supply interrupted district, relaying configuration information and the protection that fault element is concentrated with
Breaker actuation logic, that establishes each initial storehouse institute is compounded with color table；
1031st, according to the network topological information of power supply interrupted district and protection and breaker warning message, power supply interrupted district is carried out
Further fault search determines suspected fault element and its corresponding suspected fault element alarm information in power supply interrupted district, and structure
Into the first suspected fault element collection；
1032nd, suspected fault element and its corresponding suspected fault element alarm is concentrated to believe according to the first possible breakdown element
Breath builds inference temporal, determines the initial marking of inference temporal, and inference temporal is solved, obtains compound
There are the element fault feature in color table and its corresponding warning information；
1033rd, the actual act time in the warning message of element fault feature and expectation actuation time section, point
Not Ji Suan failure amount beginning confidence level, and using obtained each failure amount beginning confidence level as Fuzzy Inference Model starting library institute it is coloured
The initial confidence level of each event in table determines the initial marking of Fuzzy Inference Model；
1034th, Fuzzy Inference Model is solved, determines the fault element of the first suspected fault element concentration and its element event
Hinder probability；
1035th, according to static grid topology data and protection and breaker warning message, power supply interrupted district is carried out further
Fault search determines suspected fault element and its corresponding suspected fault element alarm information in power supply interrupted district, and forms second
Suspected fault element collection；
1036th, it obtains the second suspected fault element in fault recording system and concentrates electric current of the suspected fault element in relation to circuit
Signal, current signal are current waveform data；
1037th, the empirical mode decomposition being improved to the current signal got obtains IMF components；
1038th, Hilbert transform is carried out to the IMF components after decomposition, is calculated corresponding instantaneous amplitude, and according to
Instantaneous amplitude forms marginal spectrum；
1039th, the HHT amplitudes degree of distortion of fault current and HHT frequency distortion degree are calculated, by the HHT amplitudes degree of distortion and HHT
Frequency distortion degree is characterized as element fault probability, is characterized according to element fault probability and is determined what the second suspected fault element was concentrated
Fault element；
1041st, fault element and its element fault probability that the first suspected fault element is concentrated are obtained；
1042nd, fault element and its element fault probability that the second suspected fault element is concentrated are obtained；
1043rd, the fault element concentrated the first suspected fault element of acquisition using D-S evidence theory Fusion Model and
The fault element and its element fault probability characterization fusion that its element fault probability and the second suspected fault element are concentrated, are melted
Close result；
1051st, the element fault probability characterization in fusion results is obtained, n element fault probability characterization is used into m respectively
(F1), m (F2) ..., m (Fn) represent, and calculate the distance between each element fault probability characterization sample dij=| | m (Fi)-m
(Fj) | |, i ＜ j；
1052nd, the distance between sample is characterized according to each element fault probability, obtains each element fault probability characterization
The local density of sample, and generate the descending arrangement subscript sequence of local density's sequence；
1053rd, each element fault probability characterization sample and fault cluster central point and non-event are calculated according to gaussian kernel function
Hinder cluster centre point distance, and by element fault probability characterization sample be divided to distance closer to fault cluster or non-faulting gather
In class, primary fault cluster and initial non-faulting cluster are obtained；
1054th, the sample average of primary fault cluster is calculated as primary fault cluster centre, is calculated initial non-faulting and is gathered
The sample average of class is as initial non-faulting cluster centre；
1055th, the error sum of squares that element fault probability characterizes in primary fault cluster and initial non-faulting cluster is calculated,
And classification results when error sum of squares to reach to minimum determine fault element as optimal value；
1056th, it calculates overall error quadratic sum and return to step 1053 enters iteration, until overall error quadratic sum is constant.
It is to a kind of another embodiment of the electric network failure diagnosis method merged based on information provided by the invention above
Illustrate, below by it is provided by the invention it is a kind of based on information fusion electric network failure diagnosis device one embodiment into
Row explanation.
Referring to Fig. 2, the present invention provides a kind of one embodiment of the electric network failure diagnosis device based on information fusion,
Including：
Power supply interrupted district determination unit 201 for carrying out fault zone search to power grid, determines power supply interrupted district；
Fault element collection construction unit 202 forms fault element collection for obtaining whole elements in power supply interrupted district, obtains
The fault message of whole elements, fault message include switching value information and electric quantity information；
Storage unit 203 carries out information analysis to fault element collection for being based on Fuzzy Coloured Petri Nets, obtains
The element fault probability that fault element is concentrated carries out information analysis to fault element collection based on HHT technologies are improved, obtains failure member
The element fault probability characterization that part is concentrated；
Storage unit 203 specifically includes：
Color table structure subelement 2030 is compounded with, for the network topology of the power supply interrupted district letter concentrated according to fault element
Breath, relaying configuration information and protection and breaker actuation logic, that establishes each initial storehouse institute is compounded with color table；
First suspected fault element collection establishes subelement 2031, for the network topological information according to power supply interrupted district and protection
With breaker warning message, further fault search is carried out to power supply interrupted district, determine in power supply interrupted district suspected fault element and its
Corresponding suspected fault element alarm information, and form the first suspected fault element collection；
Inference temporal structure solves subelement 2032, for concentrating suspected fault member according to the first possible breakdown element
Part and its corresponding suspected fault element alarm information build inference temporal, determine the initial marking of inference temporal,
And inference temporal is solved, obtain being compounded with the element fault feature in color table and its corresponding warning information；
Fuzzy Inference Model builds computation subunit 2033, for the reality in the warning message according to element fault feature
Actuation time and expectation actuation time section, calculate failure amount beginning confidence level, and each failure amount beginning confidence level that will be obtained respectively
As the initial confidence level of each event in all color tables in Fuzzy Inference Model starting library, the initial mark of Fuzzy Inference Model is determined
Know；
Fisrt fault element determination subelement 2034 for being solved to Fuzzy Inference Model, determines the first suspected fault member
The fault element and its element fault probability that part is concentrated；
Second suspected fault element collection establishes subelement 2035, for according to static grid topology data and protection and open circuit
Device warning message carries out further fault search to power supply interrupted district, determines in power supply interrupted district suspected fault element and its corresponding
Suspected fault element alarm information, and form the second suspected fault element collection；
Current signal obtains subelement 2036, suspicious for obtaining the second suspected fault element concentration in fault recording system
Current signal of the fault element in relation to circuit, current signal are current waveform data；
Empirical mode decomposition subelement 2037, for the empirical mode decomposition being improved to the current signal got,
Obtain IMF components；
Hilbert transform subelement 2038 for carrying out Hilbert transform to the IMF components after decomposition, is calculated
Corresponding instantaneous amplitude, and marginal spectrum is formed according to instantaneous amplitude；
Second fault element determination subelement 2039 calculates the HHT amplitudes degree of distortion of fault current and HHT frequency distortions
Degree characterizes the HHT amplitudes degree of distortion and HHT frequency distortions degree as element fault probability, is characterized according to element fault probability true
The fault element that fixed second suspected fault element is concentrated；
Information fusion unit 204 is merged for being characterized to element fault probability using preset Fusion Model into row information, obtained
To fusion results；
Information fusion unit 204, specifically includes：
First obtains subelement 2041, for obtaining the fault element and its element fault that the first suspected fault element is concentrated
Probability；
Second obtains subelement 2042, for obtaining the fault element and its element fault that the second suspected fault element is concentrated
Probability；
Information merges subelement 2043, for using D-S evidence theory Fusion Model that the first suspected fault of acquisition is first
The fault element and its element fault that the fault element and its element fault probability and the second suspected fault element that part is concentrated are concentrated
Probability characterization fusion, obtains fusion results.
Fault element determination unit 205, for utilize preset decision diagnostic model, according to fusion results to fault element into
Row classification, determines the element to break down in power supply interrupted district；
Fault element determination unit 205 specifically includes：
Third obtains subelement 2051, for obtaining the characterization of the element fault probability in fusion results, by n element fault
Probability characterization uses m (F respectively1), m (F2) ..., m (Fn) represent, and calculate between each element fault probability characterization sample away from
From dij=| | m (Fi)-m(Fj) | |, i ＜ j；
Local density's sequence generating unit 2052, for characterizing the distance between sample according to each element fault probability,
The local density of each element fault probability characterization sample is obtained, and generates the descending arrangement subscript sequence of local density's sequence；
Generation subelement 2053 is clustered, for calculating each element fault probability characterization sample and event according to gaussian kernel function
Hinder the distance of cluster centre point and non-faulting cluster centre point, and by element fault probability characterization sample be divided to distance closer to
In fault cluster or non-faulting cluster, primary fault cluster and initial non-faulting cluster are obtained；
Cluster centre computation subunit 2054, the sample average for calculating primary fault cluster are clustered as primary fault
Center calculates the sample average of initial non-faulting cluster as initial non-faulting cluster centre；
Fault element determination subelement 2055, for calculating element fault in primary fault cluster and initial non-faulting cluster
The error sum of squares of probability characterization, and classification results when error sum of squares to reach to minimum determine that failure is first as optimal value
Part；
Iteration subelement 2056 changes for calculating overall error quadratic sum and jumping to the cluster generation entrance of subelement 2053
Generation, until overall error quadratic sum is constant.
It is apparent to those skilled in the art that for convenience and simplicity of description, the device of foregoing description
With the specific work process of unit, the corresponding process in preceding method embodiment can be referred to, details are not described herein.
The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although with reference to the foregoing embodiments
The present invention is described in detail, it will be understood by those of ordinary skill in the art that：It still can be to aforementioned each implementation
Technical solution recorded in example modifies or carries out equivalent replacement to which part technical characteristic；And these modification or
It replaces, the spirit and scope for various embodiments of the present invention technical solution that it does not separate the essence of the corresponding technical solution.
Claims (10)
- A kind of 1. electric network failure diagnosis method based on information fusion, which is characterized in that including：S1, fault zone search is carried out to power grid, determines power supply interrupted district；S2, the whole elements obtained in power supply interrupted district form fault element collection, obtain the fault message of whole elements, fault message Including switching value information and electric quantity information；S3, information analysis is carried out to fault element collection based on Fuzzy Coloured Petri Nets, obtains the element fault of fault element concentration Probability carries out information analysis to fault element collection based on HHT technologies are improved, obtains the element fault probability table of fault element concentration Sign；S4, element fault probability is characterized using preset Fusion Model and is merged into row information, obtain fusion results；S5, using preset decision diagnostic model, classified according to fusion results to fault element, determine to occur in power supply interrupted district The element of failure.
- 2. the electric network failure diagnosis method according to claim 1 based on information fusion, which is characterized in that base in step S3 Information analysis is carried out to fault element collection in Fuzzy Coloured Petri Nets, the element fault probability for obtaining fault element concentration is specific Including：According to network topological information, relaying configuration information and the protection of the power supply interrupted district of fault element concentration and breaker actuation Logic, that establishes each initial storehouse institute is compounded with color table；According to the network topological information of power supply interrupted district and protection and breaker warning message, further failure is carried out to power supply interrupted district Search determines suspected fault element and its corresponding suspected fault element alarm information in power supply interrupted district, and it is suspicious to form first Fault element collection；Suspected fault element and its corresponding suspected fault element alarm information are concentrated according to the first possible breakdown element, during structure Sequence inference pattern determines the initial marking of inference temporal, and inference temporal is solved, and obtains being compounded in color table Element fault feature and its corresponding warning information；Actual act time and expectation actuation time section in the warning message of element fault feature, failure is calculated respectively Beginning confidence level is measured, and using obtained each failure amount beginning confidence level as each event in all color tables in Fuzzy Inference Model starting library Initial confidence level, determine the initial marking of Fuzzy Inference Model；Fuzzy Inference Model is solved, determines the fault element and its element fault probability of the first suspected fault element concentration.
- 3. the electric network failure diagnosis method according to claim 2 based on information fusion, which is characterized in that in step S3 Information analysis is carried out to fault element collection based on HHT technologies are improved, obtains the element fault probability characterization tool of fault element concentration Body includes：According to static grid topology data and protection and breaker warning message, further fault search is carried out to power supply interrupted district, It determines suspected fault element and its corresponding suspected fault element alarm information in power supply interrupted district, and forms the second suspected fault member Part collection；It obtains the second suspected fault element in fault recording system and concentrates current signal of the suspected fault element in relation to circuit, electric current Signal is current waveform data；To the empirical mode decomposition that the current signal got is improved, IMF components are obtained；Hilbert transform is carried out to the IMF components after decomposition, corresponding instantaneous amplitude is calculated, and according to instantaneous amplitude shape Into marginal spectrum；The HHT amplitudes degree of distortion of fault current and HHT frequency distortion degree are calculated, by the HHT amplitudes degree of distortion and HHT frequency distortion degree It is characterized as element fault probability, the fault element for determining that the second suspected fault element is concentrated is characterized according to element fault probability.
- 4. the electric network failure diagnosis method according to claim 3 based on information fusion, which is characterized in that step S4 is specific Including：Obtain the fault element and its element fault probability of the first suspected fault element concentration；Obtain the fault element and its element fault probability of the second suspected fault element concentration；The fault element and its element fault for being concentrated the first suspected fault element of acquisition using D-S evidence theory Fusion Model The fault element and its element fault probability characterization fusion that probability and the second suspected fault element are concentrated, obtain fusion results.
- 5. the electric network failure diagnosis method according to claim 4 based on information fusion, which is characterized in that step S5 is specific Including：S501, the element fault probability obtained in fusion results characterize, and n element fault probability characterization is used m (F respectively1), m (F2) ..., m (Fn) represent, and calculate the distance between each element fault probability characterization sample dij=| | m (Fi)-m(Fj) | |, I ＜ j；S502, the distance between sample is characterized according to each element fault probability, obtains each element fault probability characterization sample Local density, and generate local density's sequence descending arrangement subscript sequence；S503, gathered according to each element fault probability characterization sample of gaussian kernel function calculating with fault cluster central point and non-faulting The distance of class central point, and by element fault probability characterization sample be divided to distance closer to fault cluster or non-faulting cluster In, obtain primary fault cluster and initial non-faulting cluster；S504, the sample average for calculating primary fault cluster calculate initial non-faulting cluster as primary fault cluster centre Sample average is as initial non-faulting cluster centre；S505, the error sum of squares that element fault probability characterizes in primary fault cluster and initial non-faulting cluster is calculated, and will Classification results when error sum of squares reaches minimum determine fault element as optimal value；S506, calculating overall error quadratic sum, simultaneously return to step S503 enters iteration, until overall error quadratic sum is constant.
- 6. a kind of electric network failure diagnosis device based on information fusion, which is characterized in that including：Power supply interrupted district determination unit for carrying out fault zone search to power grid, determines power supply interrupted district；Fault element collection construction unit forms fault element collection for obtaining whole elements in power supply interrupted district, obtains all members The fault message of part, fault message include switching value information and electric quantity information；Storage unit carries out information analysis to fault element collection for being based on Fuzzy Coloured Petri Nets, obtains fault element The element fault probability of concentration carries out information analysis to fault element collection based on HHT technologies are improved, obtains fault element concentration Element fault probability characterizes；Information fusion unit is merged for being characterized to element fault probability using preset Fusion Model into row information, is merged As a result；Fault element determination unit for utilizing preset decision diagnostic model, classifies to fault element according to fusion results, Determine the element to break down in power supply interrupted district.
- 7. the electric network failure diagnosis device according to claim 6 based on information fusion, which is characterized in that information analysis list Member specifically includes：Be compounded with color table structure subelement, for concentrated according to fault element the network topological information of power supply interrupted district, protection match Confidence breath and protection and breaker actuation logic, that establishes each initial storehouse institute is compounded with color table；First suspected fault element collection establishes subelement, for the network topological information according to power supply interrupted district and protection and breaker Warning message carries out further fault search to power supply interrupted district, determine in power supply interrupted district suspected fault element and its it is corresponding can Fault element warning information is doubted, and forms the first suspected fault element collection；Inference temporal structure solves subelement, for concentrating suspected fault element and its right according to the first possible breakdown element The suspected fault element alarm information answered builds inference temporal, determines the initial marking of inference temporal, and to sequential Inference pattern solves, and obtains being compounded with the element fault feature in color table and its corresponding warning information；Fuzzy Inference Model builds computation subunit, for the actual act time in the warning message according to element fault feature With expectation actuation time section, failure amount beginning confidence level is calculated respectively, and using obtained each failure amount beginning confidence level as fuzzy The initial confidence level of each event in all color tables in inference pattern starting library determines the initial marking of Fuzzy Inference Model；Fisrt fault element determination subelement for being solved to Fuzzy Inference Model, determines what the first suspected fault element was concentrated Fault element and its element fault probability.
- 8. the electric network failure diagnosis device according to claim 7 based on information fusion, which is characterized in that information analysis list Member specifically further includes：Second suspected fault element collection establishes subelement, for according to static grid topology data and protection and breaker warning letter Breath carries out further fault search to power supply interrupted district, determines suspected fault element and its corresponding suspected fault in power supply interrupted district Element alarm information, and form the second suspected fault element collection；Current signal obtains subelement, and suspected fault element is concentrated for obtaining the second suspected fault element in fault recording system Current signal in relation to circuit, current signal are current waveform data；Empirical mode decomposition subelement for the empirical mode decomposition being improved to the current signal got, obtains IMF points Amount；For carrying out Hilbert transform to the IMF components after decomposition, corresponding wink is calculated in Hilbert transform subelement When amplitude, and according to instantaneous amplitude formed marginal spectrum；Second fault element determination subelement calculates the HHT amplitudes degree of distortion of fault current and HHT frequency distortion degree, by HHT width It is worth the degree of distortion and HHT frequency distortions degree to characterize as element fault probability, determines that second is suspicious according to element fault probability characterization The fault element that fault element is concentrated.
- 9. the electric network failure diagnosis device according to claim 8 based on information fusion, which is characterized in that information fusion is single Member specifically includes：First obtains subelement, for obtaining fault element and its element fault probability that the first suspected fault element is concentrated；Second obtains subelement, for obtaining fault element and its element fault probability that the second suspected fault element is concentrated；Information merges subelement, for concentrated the first suspected fault element of acquisition using D-S evidence theory Fusion Model The fault element and its element fault probability characterization that fault element and its element fault probability and the second suspected fault element are concentrated Fusion, obtains fusion results.
- 10. the electric network failure diagnosis device according to claim 9 based on information fusion, which is characterized in that fault element Determination unit specifically includes：Third obtains subelement, and for obtaining the characterization of the element fault probability in fusion results, n element fault probability is characterized M (F are used respectively1), m (F2) ..., m (Fn) represent, and calculate the distance between each element fault probability characterization sample dij=| | m (Fi)-m(Fj) | |, i ＜ j；Local density's sequence generating unit for characterizing the distance between sample according to each element fault probability, obtains each Element fault probability characterizes the local density of sample, and generates the descending arrangement subscript sequence of local density's sequence；Cluster generation subelement, for being calculated in each element fault probability characterization sample and fault cluster according to gaussian kernel function The distance of heart point and non-faulting cluster centre point, and by element fault probability characterization sample be divided to distance closer to fault cluster Or in non-faulting cluster, primary fault cluster and initial non-faulting cluster are obtained；Cluster centre computation subunit, for calculating the sample average of primary fault cluster as primary fault cluster centre, meter The sample average of initial non-faulting cluster is calculated as initial non-faulting cluster centre；Fault element determination subelement characterizes for calculating element fault probability in primary fault cluster and initial non-faulting cluster Error sum of squares, and classification results when error sum of squares to reach to minimum determine fault element as optimal value；Iteration subelement enters iteration, until overall error for calculating overall error quadratic sum and jumping to cluster generation subelement Quadratic sum is constant.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810011115.9A CN108226711A (en) | 2018-01-05 | 2018-01-05 | A kind of electric network failure diagnosis method and device based on information fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810011115.9A CN108226711A (en) | 2018-01-05 | 2018-01-05 | A kind of electric network failure diagnosis method and device based on information fusion |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108226711A true CN108226711A (en) | 2018-06-29 |
Family
ID=62642935
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810011115.9A Pending CN108226711A (en) | 2018-01-05 | 2018-01-05 | A kind of electric network failure diagnosis method and device based on information fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108226711A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110045207A (en) * | 2019-04-26 | 2019-07-23 | 国网上海市电力公司 | A kind of complex fault diagnostic method based on power grid architecture and multisource data fusion |
| CN110070118A (en) * | 2019-04-10 | 2019-07-30 | 广东电网有限责任公司 | A kind of multi-space data fusion method |
| CN110188882A (en) * | 2018-12-28 | 2019-08-30 | 湖南大学 | A kind of high conflicting evidence fusion method based on fuzzy reasoning |
| CN110632451A (en) * | 2019-08-22 | 2019-12-31 | 国网浙江省电力有限公司衢州供电公司 | Low-voltage active power distribution network fault positioning method |
| CN111338929A (en) * | 2019-12-05 | 2020-06-26 | 国网辽宁省电力有限公司信息通信分公司 | Service application system performance evaluation and analysis technical method |
| CN111489539A (en) * | 2019-01-29 | 2020-08-04 | 珠海格力电器股份有限公司 | Household appliance system fault early warning method, system and device |
| CN111781467A (en) * | 2020-07-13 | 2020-10-16 | 贵州电网有限责任公司 | Power grid fault intelligent discrimination method based on multidimensional panoramic data |
| CN113884805A (en) * | 2021-09-17 | 2022-01-04 | 国网山东省电力公司德州供电公司 | Single-phase earth fault insulation degradation identification method and system based on D-S evidence theory |
| CN115545355A (en) * | 2022-11-30 | 2022-12-30 | 国网天津市电力公司电力科学研究院 | Power grid fault diagnosis method, device and equipment based on multi-class information fusion identification |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101726688A (en) * | 2009-09-17 | 2010-06-09 | 浙江大学 | Method for diagnosing multi-data-source information fusion-based power system fault |
| CN104297637A (en) * | 2014-10-31 | 2015-01-21 | 国家电网公司 | Power system fault diagnosis method comprehensively using electricity amount and timing sequence information |
| CN106226658A (en) * | 2016-09-07 | 2016-12-14 | 上海市政工程设计研究总院（集团）有限公司 | A kind of electric network failure diagnosis method based on multi-data fusion |
-
2018
- 2018-01-05 CN CN201810011115.9A patent/CN108226711A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101726688A (en) * | 2009-09-17 | 2010-06-09 | 浙江大学 | Method for diagnosing multi-data-source information fusion-based power system fault |
| CN104297637A (en) * | 2014-10-31 | 2015-01-21 | 国家电网公司 | Power system fault diagnosis method comprehensively using electricity amount and timing sequence information |
| CN106226658A (en) * | 2016-09-07 | 2016-12-14 | 上海市政工程设计研究总院（集团）有限公司 | A kind of electric network failure diagnosis method based on multi-data fusion |
Non-Patent Citations (1)
| Title |
|---|
| 赵斌炎: "基于多信息融合的电网故障诊断方法研究", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110188882A (en) * | 2018-12-28 | 2019-08-30 | 湖南大学 | A kind of high conflicting evidence fusion method based on fuzzy reasoning |
| CN110188882B (en) * | 2018-12-28 | 2022-05-31 | 湖南大学 | High-conflict evidence fusion method based on fuzzy reasoning |
| CN111489539A (en) * | 2019-01-29 | 2020-08-04 | 珠海格力电器股份有限公司 | Household appliance system fault early warning method, system and device |
| CN110070118A (en) * | 2019-04-10 | 2019-07-30 | 广东电网有限责任公司 | A kind of multi-space data fusion method |
| CN110045207A (en) * | 2019-04-26 | 2019-07-23 | 国网上海市电力公司 | A kind of complex fault diagnostic method based on power grid architecture and multisource data fusion |
| CN110632451A (en) * | 2019-08-22 | 2019-12-31 | 国网浙江省电力有限公司衢州供电公司 | Low-voltage active power distribution network fault positioning method |
| CN111338929A (en) * | 2019-12-05 | 2020-06-26 | 国网辽宁省电力有限公司信息通信分公司 | Service application system performance evaluation and analysis technical method |
| CN111781467A (en) * | 2020-07-13 | 2020-10-16 | 贵州电网有限责任公司 | Power grid fault intelligent discrimination method based on multidimensional panoramic data |
| CN113884805A (en) * | 2021-09-17 | 2022-01-04 | 国网山东省电力公司德州供电公司 | Single-phase earth fault insulation degradation identification method and system based on D-S evidence theory |
| CN113884805B (en) * | 2021-09-17 | 2023-10-20 | 国网山东省电力公司德州供电公司 | Single-phase earth fault insulation degradation identification method and system based on D-S evidence theory |
| CN115545355A (en) * | 2022-11-30 | 2022-12-30 | 国网天津市电力公司电力科学研究院 | Power grid fault diagnosis method, device and equipment based on multi-class information fusion identification |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108226711A (en) | A kind of electric network failure diagnosis method and device based on information fusion | |
| Cui et al. | Enhance high impedance fault detection and location accuracy via $\mu $-PMUs | |
| Pandey et al. | A real time event detection, classification and localization using synchrophasor data | |
| Kamwa et al. | Development of rule-based classifiers for rapid stability assessment of wide-area post-disturbance records | |
| CN106054104B (en) | A kind of intelligent electric meter failure real-time predicting method based on decision tree | |
| Khaledian et al. | Real-time synchrophasor data anomaly detection and classification using isolation forest, kmeans, and loop | |
| CN105470970B (en) | The guard method of subsynchronous and supersynchronous vibration and system in power system | |
| CN103760901A (en) | Rail transit fault identification method based on association rule classifier | |
| US11656263B2 (en) | Effective feature set-based high impedance fault detection | |
| Hossam‐Eldin et al. | Artificial intelligence‐based short‐circuit fault identifier for MT‐HVDC systems | |
| CN107478963A (en) | Single-phase ground fault line selecting method of small-electric current grounding system based on power network big data | |
| CN111398679B (en) | Sub-synchronous oscillation identification and alarm method based on PMU (phasor measurement Unit) | |
| CN109142986A (en) | A kind of Fault Locating Method of AC power distribution line | |
| CN108053128A (en) | A kind of Power Network Transient Stability fast evaluation method based on ELM and TF | |
| CN106411617A (en) | Power communication network fault warning correlation processing method | |
| CN112464995A (en) | Power grid distribution transformer fault diagnosis method and system based on decision tree algorithm | |
| Linda et al. | Improving cyber-security of smart grid systems via anomaly detection and linguistic domain knowledge | |
| De Santis et al. | A learning intelligent system for classification and characterization of localized faults in smart grids | |
| Gholami et al. | Comparative analysis of ml techniques for data-driven anomaly detection, classification and localization in distribution system | |
| CN103311960A (en) | Method for dividing coherent stability regions of strong-connection power grids | |
| Zhang et al. | K-coverage: A monitor node selection algorithm for diffusion source localizations | |
| CN103778479A (en) | Adaptive information fault-tolerant protection method | |
| Wehenkel et al. | Artificial intelligence applied to on-line transient stability assessment of electric power systems | |
| CN105678456A (en) | Method for automatically assessing electric energy metering device operation status and system thereof | |
| Wang | Fault line selection of power distribution system via improved bee colony algorithm based deep neural network |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200115 Address after: 510000 room 501-503, annex building, Huaye building, No.1-3, Chumo new street, Xihua Road, Yuexiu District, Guangzhou City, Guangdong Province Applicant after: Guangdong Electric Power Research Institute of energy technology limited liability company Address before: 510080 Dongfeng East Road, Dongfeng, Guangdong, Guangzhou, Zhejiang Province, No. 8 Applicant before: ELECTRIC POWER RESEARCH INSTITUTE, GUANGDONG POWER GRID CO., LTD. Applicant before: Guangdong Electric Power Research Institute of energy technology limited liability company |
|
| TA01 | Transfer of patent application right | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180629 |
|
| RJ01 | Rejection of invention patent application after publication |