CN110334127A - A kind of distribution network line fault law mining method, system and storage medium - Google Patents

A kind of distribution network line fault law mining method, system and storage medium Download PDF

Info

Publication number
CN110334127A
CN110334127A CN201910600307.8A CN201910600307A CN110334127A CN 110334127 A CN110334127 A CN 110334127A CN 201910600307 A CN201910600307 A CN 201910600307A CN 110334127 A CN110334127 A CN 110334127A
Authority
CN
China
Prior art keywords
attribute
fault
distribution network
distribution
frequent
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201910600307.8A
Other languages
Chinese (zh)
Other versions
CN110334127B (en
Inventor
方鑫
史明明
张军
袁宇波
付慧
袁晓冬
费益军
陶加贵
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Corp of China SGCC
State Grid Jiangsu Electric Power Co Ltd
Electric Power Research Institute of State Grid Jiangsu Electric Power Co Ltd
Original Assignee
State Grid Corp of China SGCC
State Grid Jiangsu Electric Power Co Ltd
Electric Power Research Institute of State Grid Jiangsu Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by State Grid Corp of China SGCC, State Grid Jiangsu Electric Power Co Ltd, Electric Power Research Institute of State Grid Jiangsu Electric Power Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN201910600307.8A priority Critical patent/CN110334127B/en
Publication of CN110334127A publication Critical patent/CN110334127A/en
Application granted granted Critical
Publication of CN110334127B publication Critical patent/CN110334127B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2465—Query processing support for facilitating data mining operations in structured databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/28—Databases characterised by their database models, e.g. relational or object models
    • G06F16/284—Relational databases
    • G06F16/285—Clustering or classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Electricity, gas or water supply
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Abstract

The embodiment of the invention discloses a kind of distribution network line fault law mining method, system and storage mediums.Wherein, this method includes acquisition distribution network line fault tripping record, generates the flag event under distribution network line fault and reason scene, and generate distribution line failure information bank；Method of Data Discretization based on k-means clusters each single item attribute information in distribution line failure information bank, obtains attribute preliminary classification collection；Attribute preliminary classification collection is carried out equal frequency to optimize, is based on apriori algorithm, faults frequent item set mining is carried out to one kind fault attribute every in fault attribute category set, constructs frequent Candidate Set, and carry out distribution line fault trip strong correlation Rules Filtering.The technical solution of the embodiment of the present invention realizes the cluster of fault attribute to excavate fault correlation rule, reduces power distribution network and makes an inspection tour O&M range.

Description

A kind of distribution network line fault law mining method, system and storage medium
Technical field
The present embodiments relate to system for distribution network of power running technology field more particularly to a kind of distribution network line faults Law mining method, system and storage medium.
Background technique
Power distribution network is as the power network for being directly facing user, and Distribution Network Equipment scale is big, wide coverage, locating for equipment External environment is complicated, and the failure of current power grid 80% belongs to distribution network failure according to statistics.Cause distribution network failure trip reason phase Closing relation factor includes temperature, line length, the operation time limit, affiliated distribution transforming quantity and dynamic operation data of route etc..For Promotion power supply reliability needs to carry out inspection for route in advance, since power distribution network distribution is wide, the equipment scale scale of construction is big, There is no failure to assist under policy condition, realizes that the tour of whole region route and equipment and maintenance expense are huge, difficult to realize pair The anticipation of the dynamic operation data real time fail of route.
Therefore, the whole region route of existing power distribution network and the tour of equipment and the costly problem of maintenance urgently solve Certainly.
Summary of the invention
The embodiment of the present invention provides a kind of distribution network line fault law mining method, system and storage medium, realization and matches The judgement of electric network fault range reduces power distribution network and patrols dimension range, to solve the whole region route of existing power distribution network and set The costly problem of standby tour and maintenance.
To realize above-mentioned technical problem, the invention adopts the following technical scheme:
In a first aspect, the embodiment of the invention provides a kind of distribution network line fault law mining methods, comprising:
Distribution network line fault tripping record is acquired, the flag event under distribution network line fault and reason scene is generated；
According to the flag event, the distribution line failure information bank for containing more attribute informations is generated；
Method of Data Discretization based on k-means, to each single item attribute information in the distribution line failure information bank It is clustered, obtains attribute preliminary classification collection；
The attribute preliminary classification collection is carried out equal frequency to optimize, obtains fault attribute category set；
Based on apriori algorithm, different faults reason is carried out to one kind fault attribute every in the fault attribute category set The frequent candidate of different frequently attribute numbers is covered in faults frequent item set mining under scene, building；
According to the frequent candidate, it is based on the distribution line failure information bank, carries out distribution line fault trip Strong correlation Rules Filtering.
Further, the acquisition distribution network line fault tripping record, generates distribution network line fault and reason scene Under flag event, comprising:
Acquire the distribution network line fault tripping record；
Based on dispatch automated system or other dynamic monitoring systems, jumped according to the collected distribution network line fault Lock record, generates distribution network line fault trip indication event.
Further, according to the flag event, the distribution line failure information bank for containing more attribute informations, packet are generated It includes:
According to the flag event, the multi-source data based on information system, to distribution network line feeder line grade effective information It is handled；
According to the fault trip record and fault trip time of origin per distribution network line together, the category of fault moment is integrated Property, generate the distribution line failure information bank for containing more attribute informations；
Wherein, the attribute of the fault moment includes temperature, rainfall, wind speed, route overall length, overhead transmission line overall length, distribution Route affiliated area, distribution line have the operation datas such as distribution transforming quantity, line voltage distribution, electric current, active and reactive and route fortune under its command One of row time limit or at least two.
Further, the route overall length is cable run and overhead line conductor section summation；The overhead transmission line overall length For overhead line conductor section summation；
Fault trip record and fault trip time of origin of the basis per distribution network line together, integrate fault moment Attribute, generate contain more attribute informations the distribution line failure information bank, comprising:
The each section lead segment length of distribution line is obtained by topological data, calculates the route overall length and the overhead line Lu overall length.
Further, the overhead transmission line overall length, is calculated by following formula:
Lj=∑ bi
The route overall length, is calculated by following formula:
Lt=∑ ai+∑bi
Wherein, aiAnd biIt is i sections and i sections of aerial condutor section of cable conductor section of length respectively.
Further, based on the Method of Data Discretization of k-means, to each single item in the distribution line failure information bank Attribute information is clustered, and attribute preliminary classification collection is obtained, comprising:
According to the classification quantity of attribute to be sorted, attributive classification initial value is set；
Using k-means algorithm, the number of iterations is set, solves classification point setting value again.
Further, the classification point setting value is calculated by following formula:
Ej=∑ Eij
E=∑ Ej
Wherein, E is the amount of deviation based on all objects under current class, EjIt is amount of deviation in cluster, EijIt is each object To the distance of central point；xijThe i element being classified as in j cluster,It is central point in j cluster, with EjAnd EijMinimum target is constantly adjusted Element x in whole each clusterij。
Further, the attribute preliminary classification collection is carried out equal frequency to optimize, obtains fault attribute category set, comprising:
The sample size concentrated to each attribute preliminary classification is ranked up, and by each attribute preliminary classification sum and is belonged to Property classification quantity preset value compare, when the attribute preliminary classification sum is less than the attributive classification quantity preset value, carry out etc. Frequency optimizes；
The sample size that each attribute preliminary classification is concentrated is tied compared with attributive classification sample preset value according to comparing The sample size that fruit and the attribute preliminary classification are concentrated is ranked up, the sample size concentrated to each attribute preliminary classification Equal frequency is carried out from high in the end to optimize, until attributive classification sum is equal to the attributive classification quantity preset value, and then obtains failure Attributive classification collection.
Further, the equal frequency optimization is calculated by following formula:
X=[x1 x2 ... xN]
X1=[x1 x2 ... xn]
X2=[xn+1 xn+2 ... xn+n]
...
Xk'=[xN-n+1 xN-n+2 ... xN]
Wherein, x is the subclass quantity to be split classified based on a certain classification under k-means algorithm, k ' for x, XiIt is son Class, n are sample sizes under each subclassification.
Further, it is based on apriori algorithm, one kind fault attribute every in the fault attribute category set is carried out different The frequent candidate of different frequently attribute numbers is covered in faults frequent item set mining under failure cause scene, building, comprising:
According to the distribution line failure information bank, it is based on apriori algorithm, to the frequent item set of distribution line failure Determine to support that count threshold is set；
Count threshold is supported according to the judgement of setting, and the faults frequent item that different frequently attribute numbers are covered in building is waited Selected works.
Further, the frequent candidate of different frequently attribute numbers is covered in the building, comprising:
The distribution line failure information bank is scanned, k item collection support is calculated, supports count threshold when number is greater than, meter Enter k frequent candidates；
Matching splicing is carried out according to k frequent candidates, generates k+1 item collection；
Support counting screening by k+1 item collection, generates k+1 frequent candidates；
Cycle calculations, until it can not search out more how often candidate；
Wherein, the k item collection support is the number that k attribute value occurs simultaneously.
Further, according to the frequent candidate, it is based on the distribution line failure information bank, carries out distribution line Fault trip strong correlation Rules Filtering, comprising:
According to the frequent candidate, calculate support in the distribution line failure information bank under sample total, Confidence level and promotion degree；
According to the counted support, the confidence level and the promotion degree is counted, distribution line fault trip is carried out Strong correlation Rules Filtering.
Further, the support is calculated by following formula:
The confidence level is calculated by following formula:
The promotion degree is calculated by following formula:
Wherein,It is the support that A frequent item set occurs under B fault scenes,In B event Hinder the confidence level that A frequent item set occurs under scene,Indicate the simultaneous support meter of A and B Number, count (A) are the support countings that A occurs, and D is fault sample total amount；Lift (A, B) is that A and B are simultaneous The promotion degree, P (A ∪ B) are A and the B simultaneous probability in full dose sample D, and P (B/A) is probability of B under the conditions of A.
Second aspect, the embodiment of the present invention also provide a kind of distribution network line fault law mining system, comprising:
Flag event generation module, for acquire distribution network line fault tripping record, generate distribution network line fault and Flag event under reason scene；
Fault message storehouse generation module, for generating the distribution line for containing more attribute informations according to the flag event Fault message storehouse；
Hierarchical cluster attribute module, for the Method of Data Discretization based on k-means, to the distribution line failure information bank Middle each single item attribute information is clustered, and attribute preliminary classification collection is obtained；
Equal frequencies optimization module obtains fault attribute category set for the attribute preliminary classification collection to be carried out equal optimization frequently；
Faults frequent item set mining module is based on apriori algorithm, to one kind failure every in the fault attribute category set Attribute carries out the faults frequent item set mining under different faults reason scene, and the frequent time of different frequently attribute numbers is covered in building Set of choices；
Strong correlation Rules Filtering module, for being based on the distribution line failure information according to the frequent candidate Library carries out distribution line fault trip strong correlation Rules Filtering.
The third aspect, the embodiment of the present invention also provide a kind of readable storage medium storing program for executing, when the instruction in the storage medium by When the processor of distribution network line fault law mining system executes, distribution network line fault law mining system is held Distribution network line fault law mining method of the row as described in first aspect is any.
A kind of distribution network line fault law mining method provided in an embodiment of the present invention passes through acquisition distribution network line event Barrier tripping record, generates the flag event under distribution network line fault and reason scene；According to the flag event, generation contains The distribution line failure information bank of more attribute informations；Method of Data Discretization based on k-means, to the distribution line failure Each single item attribute information is clustered in information bank, obtains attribute preliminary classification collection；The attribute preliminary classification collection is carried out etc. Frequency optimizes, and obtains fault attribute category set；Based on apriori algorithm, to one kind failure category every in the fault attribute category set Property carry out faults frequent item set mining under different faults reason scene, the frequent candidate of different frequently attribute numbers is covered in building Item collection；According to the frequent candidate, it is based on the distribution line failure information bank, carries out the strong phase of distribution line fault trip Rules Filtering is closed, other multi-source datas merged existing distribution network failure sample data under failure is realized, merges k-means's Method of Data Discretization and frequent item set data digging method excavate fault correlation rule, reduce power distribution network and make an inspection tour O&M range, To improve power distribution network operational reliability, solve the whole region route of existing power distribution network and the tour of equipment and maintenance expense Huge problem.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, institute in being described below to the embodiment of the present invention Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the invention Example, for those of ordinary skill in the art, without creative efforts, can also implement according to the present invention The content of example and these attached drawings obtain other attached drawings.
Fig. 1 is a kind of distribution network line fault law mining method flow diagram provided in an embodiment of the present invention；
Fig. 2 is another distribution network line fault law mining method flow diagram provided in an embodiment of the present invention；
Fig. 3 is another distribution network line fault law mining method flow diagram provided in an embodiment of the present invention；
Fig. 4 is a kind of structural schematic diagram of distribution network line fault law mining system provided in an embodiment of the present invention.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched The specific embodiment stated is used only for explaining the present invention rather than limiting the invention.It also should be noted that in order to just Only the parts related to the present invention are shown in description, attached drawing rather than entire infrastructure.
The embodiment of the present invention provides a kind of distribution network line fault law mining method.Fig. 1 is that the embodiment of the present invention provides A kind of distribution network line fault law mining method flow diagram.Referring to Fig. 1, distribution network line event provided in an embodiment of the present invention Hinder law mining method, comprising:
S101, acquisition distribution network line fault tripping record, generate the label under distribution network line fault and reason scene Event.
Specifically, distribution network line fault tripping record includes distribution network line fault tripping information and distribution network line event Hinder reason, according to the distribution network line fault of acquisition tripping information and distribution network line fault reason, generates distribution network line event Flag event under barrier and reason scene.
S102, according to the flag event, generate the distribution line failure information bank for containing more attribute informations.
Specifically, for per distribution network line fault trip indication event together, it is based on information system multi-source data, it is complete It is handled at distribution line feeder line grade effective information, generates the distribution line failure information bank for containing more attribute informations.
S103, the Method of Data Discretization based on k-means, to each single item attribute in the distribution line failure information bank Information is clustered, and attribute preliminary classification collection is obtained.
Specifically, to each single item attribute information in distribution line failure information bank, the Data Discretization side based on k-means Method is clustered, and attribute preliminary classification collection is obtained, and attribute preliminary classification collection includes the group class that there is like attribute information to constitute Set.
S104, the attribute preliminary classification collection such as is subjected at the frequency optimizes, obtain fault attribute category set.
Specifically, according to inspection O&M actual demand set class condition, the attribute value that attribute preliminary classification is concentrated with Class condition compares, and the attribute value that the attribute preliminary classification for being unsatisfactory for class condition is concentrated is carried out equal frequency optimization, realization attribute Rationalize classification, obtains fault attribute category set.
S105, it is based on apriori algorithm, different faults is carried out to one kind fault attribute every in the fault attribute category set The frequent candidate of different frequently attribute numbers is covered in faults frequent item set mining under reason scene, building.
Specifically, support count threshold needed for setting the judgement of the frequent property set of distribution line failure counts certain The number that specific attribute occurs simultaneously will count the number of obtained certain specific attributes while appearance and support counting threshold Value is compared, cycle calculations, and the frequent candidate of different frequently attribute numbers is covered in building.
S106, distribution line failure is carried out based on the distribution line failure information bank according to the frequent candidate Trip strong correlation Rules Filtering.
Specifically, according to the frequent candidate, it is based on the distribution line failure information bank, calculates distribution line event Hinder support, confidence level and the promotion degree in information bank under sample total, counted support, confidence level and promotion will be counted Degree is compared with corresponding preset threshold respectively, is obtained distribution line fault trip strong correlation rule, is obtained event to excavate Hinder correlation rule, reduces power distribution network and make an inspection tour O&M range, can effectively improve power distribution network operational reliability.
A kind of distribution network line fault law mining method provided in an embodiment of the present invention passes through acquisition distribution network line event Barrier tripping record, generates the flag event under distribution network line fault and reason scene；According to flag event, generates and belong to containing morely The distribution line failure information bank of property information；Method of Data Discretization based on k-means, in distribution line failure information bank Each single item attribute information is clustered, and attribute preliminary classification collection is obtained；The attribute preliminary classification collection is carried out equal frequency to optimize, is obtained To fault attribute category set；Based on apriori algorithm, different faults are carried out to one kind fault attribute every in fault attribute category set The frequent candidate of different frequently attribute numbers is covered in faults frequent item set mining under reason scene, building；According to frequent Candidate is based on distribution line failure information bank, carries out distribution line fault trip strong correlation Rules Filtering, and being able to achieve will be existing There are distribution network failure sample data, including real-time sample data and historical sample data, merge other multi-source datas under failure, Frequent item set data digging method is merged, fault correlation rule is excavated, it is right by the Method of Data Discretization based on k-means Each single item attribute information is clustered in distribution line failure information bank, obtains fault attribute category set, is calculated based on apriori Method carries out the faults frequent item set mining under different faults reason scene to one kind fault attribute every in fault attribute category set, It realizes distribution line fault trip strong correlation Rules Filtering, reduces power distribution network and make an inspection tour O&M range, improve power distribution network operation Reliability solves the problems, such as that the tour of the whole region route of existing power distribution network and equipment and maintenance expense are high.
Optionally, Fig. 2 is another distribution network line fault law mining method flow diagram provided in an embodiment of the present invention. On the basis of the above embodiments, referring to fig. 2, another distribution network line fault law mining side provided in an embodiment of the present invention Method, comprising:
S201, the acquisition distribution network line fault tripping record.
Specifically, distribution network line fault tripping record can be existing distribution network line fault tripping record, that is, match Guangdong power system tripping historical information, specific time, distribution network line fault reason including distribution network line fault tripping And distribution network line fault electric current, information of voltage etc..
S202, dispatch automated system or other dynamic monitoring systems are based on, according to the collected distribution network line Fault trip record, generates distribution network line fault trip indication event.
Specifically, dispatch automated system or other dynamic monitoring systems can provide the remote control of power distribution network, telemetry intelligence (TELINT), In conjunction with collected different distribution network line fault tripping records, different distribution network line fault trip indication events is generated.
S203, according to the flag event, generate the distribution line failure information bank for containing more attribute informations.
S204, the Method of Data Discretization based on k-means, to each single item attribute in the distribution line failure information bank Information is clustered, and attribute preliminary classification collection is obtained.
S205, the attribute preliminary classification collection such as is subjected at the frequency optimizes, obtain fault attribute category set.
S206, it is based on apriori algorithm, different faults is carried out to one kind fault attribute every in the fault attribute category set The frequent candidate of different frequently attribute numbers is covered in faults frequent item set mining under reason scene, building.
S207, distribution line failure is carried out based on the distribution line failure information bank according to the frequent candidate Trip strong correlation Rules Filtering.
Optionally, Fig. 3 is another distribution network line fault law mining method flow diagram provided in an embodiment of the present invention. On the basis of the above embodiments, referring to Fig. 3, another distribution network line fault law mining side provided in an embodiment of the present invention Method, comprising:
S301, acquisition distribution network line fault tripping record, generate the label under distribution network line fault and reason scene Event.
S302, according to the flag event, the multi-source data based on information system is effective to distribution network line feeder line grade Information is handled.
Specifically, information system includes dispatch automated system, fault statistics analysis system, equipment management system, GIS Information system and meteorological system.Wherein, dispatch automated system is recorded for providing the tripping of distribution line outlet breaker, therefore Barrier statistical analysis system is set for providing distribution line failure trip reason information, equipment management system for providing distribution line Standby basis account data, GIS information system is for providing equipment connecting relation, feeder line topological data, and meteorological system is for providing Temperature, rainfall and wind speed including the fault flag moment.
S303, fault trip record and fault trip time of origin according to every distribution network line together, when integrating failure The attribute at quarter generates the distribution line failure information bank for containing more attribute informations；
Wherein, the attribute of the fault moment includes temperature, rainfall, wind speed, route overall length, overhead transmission line overall length, distribution Route affiliated area, distribution line have the operation datas such as distribution transforming quantity, line voltage distribution, electric current, active and reactive and route fortune under its command One of row time limit or at least two.
Specifically, according to the fault trip record and fault trip time of origin per distribution network line together, failure is integrated The temperature at moment, rainfall, wind speed, route overall length, overhead transmission line overall length, distribution line affiliated area, distribution line have distribution transforming under its command Attributes, the generation such as the operation datas such as quantity, line voltage distribution, electric current, active and reactive and the route operation time limit contain various faults The distribution line failure information bank of moment attribute information.
S304, the Method of Data Discretization based on k-means, to each single item attribute in the distribution line failure information bank Information is clustered, and attribute preliminary classification collection is obtained.
S305, the attribute preliminary classification collection such as is subjected at the frequency optimizes, obtain fault attribute category set.
S306, it is based on apriori algorithm, different faults is carried out to one kind fault attribute every in the fault attribute category set The frequent candidate of different frequently attribute numbers is covered in faults frequent item set mining under reason scene, building.
S307, distribution line failure is carried out based on the distribution line failure information bank according to the frequent candidate Trip strong correlation Rules Filtering.
Optionally, route overall length is cable run and overhead line conductor section summation, and overhead transmission line overall length is overhead transmission line Conducting line segment summation.According to the fault trip record and fault trip time of origin per distribution network line together, fault moment is integrated Attribute, generate contain more attribute informations the distribution line failure information bank, comprising:
The each section lead segment length of distribution line is obtained by topological data, calculates the route overall length and the overhead line Lu overall length.
Specifically, the length of each section lead is got by the topological data of power distribution network, overhead transmission line overall length can lead to Following formula are crossed to be calculated:
Lj=∑ bi
Route overall length can be calculated by following formula:
Lt=∑ ai+∑bi
Wherein, aiAnd biIt is i sections and i sections of aerial condutor section of cable conductor section of length respectively.
Optionally, based on the Method of Data Discretization of k-means, to each single item category in the distribution line failure information bank Property information is clustered, and attribute preliminary classification collection is obtained, comprising:
One, the classification quantity according to attribute to be sorted set attributive classification initial value；
Specifically, the classification quantity of attribute to be sorted can be the classification quantity of distribution network line fault to be sorted, if Determine attributive classification initial value, starts to calculate according to the attributive classification initial value of setting.
Secondly, using k-means algorithm, set the number of iterations, solve a classification point setting value again.
Specifically, iteration time is set using k-means algorithm according to the attribute information in distribution line failure information bank Number, dependence classification initial value start to calculate, and solve classification point setting value again, according to classification point setting value to the distribution wire Each single item attribute information is clustered in the fault message storehouse of road, obtains attribute preliminary classification collection.
Optionally, the classification point setting value is calculated by following formula:
Ej=∑ Eij
E=Σ Ej
Wherein, E is the amount of deviation based on all objects under current class, EjIt is amount of deviation in cluster, EijIt is each object To the distance of central point；xijThe i element being classified as in j cluster,It is central point in j cluster, with EjAnd EijMinimum target is constantly adjusted Element x in whole each clusterij。
Optionally, the attribute preliminary classification collection is carried out equal frequency to optimize, obtains fault attribute category set, comprising:
One, the sample size concentrated to each attribute preliminary classification are ranked up, and each attribute preliminary classification is total Number is with attributive classification quantity preset value compared with, when the attribute preliminary classification sum is less than the attributive classification quantity preset value, Equal frequency is carried out to optimize.
Specifically, the sample size concentrated to each attribute preliminary classification is ranked up, and attribute preliminary classification is concentrated Attribute preliminary classification sum is compared with attributive classification quantity preset value, when attribute preliminary classification sum is pre- less than attributive classification quantity If when value, needing to carry out each attribute preliminary classification collection for being unsatisfactory for sample size that frequency is waited to optimize.
Secondly, the sample size of concentrating each attribute preliminary classification compared with attributive classification sample preset value, according to The sample size that comparison result and the attribute preliminary classification are concentrated is ranked up, the sample concentrated to each attribute preliminary classification This quantity carries out that frequency is waited to optimize from high in the end, until attributive classification sum is equal to the attributive classification quantity preset value, and then obtains To fault attribute category set.
Specifically, if the sample size that attribute preliminary classification is concentrated is greater than 1.5 times of attributive classification sample preset values, according to Each attribute preliminary classification concentrates the sequence of sample size from high to low, is greater than 1.5 to the sample size that attribute preliminary classification is concentrated Each attribute preliminary classification collection of times attributive classification sample preset value carries out equal frequency divisions class, until attributive classification sum is equal to attribute point Class quantity preset value and meet batch total requirement, obtain fault attribute category set.
Optionally, the equal frequency optimization is calculated by following formula:
X=[x1 x2 ... xN]
X1=[x1 x2 ... xn]
X2=[xn+1 xn+2 ... xn+n]
...
Xk'=[xN-n+1 xN-n+2 ... xN]
Wherein, x is the subclass quantity to be split classified based on a certain classification under k-means algorithm, k ' for x, XiIt is son Class, n are sample sizes under each subclassification.
Optionally, it is based on apriori algorithm, different events are carried out to one kind fault attribute every in the fault attribute category set Hinder the faults frequent item set mining under reason scene, the frequent candidate of different frequently attribute numbers is covered in building, comprising:
One, according to the distribution line failure information bank, apriori algorithm is based on, to the frequent of distribution line failure The judgement of item collection supports that count threshold is set.
Specifically, according to the flag event in distribution line failure information bank, it is based on apriori algorithm, setting meets need Support count threshold needed for the judgement of the frequent item set for the distribution line failure asked.It should be noted that attribute is item, frequency The set of numerous attribute is frequent item set.
Secondly, count threshold is supported according to the judgement of setting, the failure frequency of different frequently attribute numbers is covered in building Numerous Candidate Set.
Specifically, count threshold is supported according to the judgement of setting, the frequent of different frequently attribute numbers is covered in building Candidate may include: scanning distribution line failure information bank, calculate the k item collection support of flag event, degree of being supported Screening is counted, when number, which is greater than, supports count threshold, is included in k frequent candidates；It is carried out according to k frequent candidates Matching splicing, generates k+1 item collection；Support counting screening by k+1 item collection, generates k+1 frequent candidates；Circulation meter It calculates, until it can not search out more how often candidate.Wherein, the k item collection support is that k attribute value occurs simultaneously Number, n be distribution line failure information bank in flag event fault attribute total quantity, k 1,2,3 ... ..., n.
Optionally, according to the frequent candidate, it is based on the distribution line failure information bank, carries out distribution line event Barrier tripping strong correlation Rules Filtering, comprising:
One, according to the frequent candidate, calculate the branch in the distribution line failure information bank under sample total Degree of holding, confidence level and promotion degree.
Specifically, support can be calculated by following formula:
Confidence level can be calculated by following formula:
Promotion degree can be calculated by following formula:
Wherein,It is the support that A frequent item set occurs under B fault scenes,In B failure field The confidence level that A frequent item set occurs under scape,The expression simultaneous support counting of A and B, count (A), It is the support counting that A occurs, D is fault sample total amount；Lift (A, B) is the simultaneous promotion degree of A and B, and P (A ∪ B) is A and B simultaneous probability in full dose sample D, P (B/A) are probability of B under the conditions of A.
Secondly, according to the counted support, the confidence level and the promotion degree is counted, carry out distribution line failure Trip strong correlation Rules Filtering.
Specifically, the threshold value of support and confidence level is set, the positive threshold and reversed threshold value of promotion degree is set, works as support Degree and confidence level are greater than given threshold, and when promotions degree is greater than positive threshold, then determine that current rule is that positive strong correlation is regular； When support and confidence level are greater than given threshold, and promotion degree is less than reversed threshold value, then it is determined as reversed strong correlation rule；Its Remaining correlation rule screens out.When distribution line fault trip strong correlation Rules Filtering result be strong correlation rule, then illustrate power distribution network Land internal fault is then needed when distribution line fault trip strong correlation Rules Filtering result is positive strong correlation rule Distribution line fault trip in the region is overhauled；When distribution line fault trip strong correlation Rules Filtering result is reversed When strong correlation rule, then foundation is provided for inspection, reduces failure rate by shifting to an earlier date inspection.
Fig. 4 is a kind of structural schematic diagram of distribution network line fault law mining system provided in an embodiment of the present invention.Ginseng See Fig. 4, distribution network line fault law mining system provided in an embodiment of the present invention, comprising:
Flag event generation module 100 generates distribution network line fault for acquiring distribution network line fault tripping record And the flag event under reason scene.
Fault message storehouse generation module 200, for generating the distribution wire for containing more attribute informations according to the flag event Road fault message storehouse.
Hierarchical cluster attribute module 300 believes the distribution line failure for the Method of Data Discretization based on k-means Each single item attribute information is clustered in breath library, obtains attribute preliminary classification collection.
Equal frequencies optimization module 400 obtains fault attribute and classifies for the attribute preliminary classification collection to be carried out equal frequency optimization Collection.
Faults frequent item set mining module 500 is based on apriori algorithm, to every one kind in the fault attribute category set Fault attribute carries out the faults frequent item set mining under different faults reason scene, and the frequency of different frequently attribute numbers is covered in building Numerous candidate.
Strong correlation Rules Filtering module 600, for being believed based on the distribution line failure according to the frequent candidate Library is ceased, distribution line fault trip strong correlation Rules Filtering is carried out.
Distribution network line fault law mining system provided in an embodiment of the present invention, including flag event generation module, event Hinder information bank generation module, hierarchical cluster attribute module, etc. frequencies optimization module, faults frequent item set mining module and strong correlation rule sieve Modeling block acquires distribution network line fault tripping record by flag event generation module, generates distribution network line fault and original Because the flag event under scene is generated by fault message storehouse generation module according to the flag event containing more attribute informations Distribution line failure information bank, Method of Data Discretization of the hierarchical cluster attribute module based on k-means, to the distribution line failure Each single item attribute information is clustered in information bank, obtains attribute preliminary classification collection, by waiting frequencies optimization module by the attribute Preliminary classification collection carries out equal frequency and optimizes, and obtains fault attribute category set, and be based on apriori algorithm, faults frequent item set mining Module carries out the faults frequent item collection under different faults reason scene to one kind fault attribute every in the fault attribute category set It excavates, the frequent candidate that different frequently attribute numbers are covered in building is based on the distribution according to the frequent candidate Line fault information bank, strong correlation Rules Filtering module carry out distribution line fault trip strong correlation Rules Filtering.The present invention is real The distribution network line fault law mining system for applying example offer, which realizes, merges existing distribution network failure sample data under failure Other multi-source datas, merge the Method of Data Discretization and frequent item set data digging method of k-means, excavate fault correlation Rule reduces power distribution network tour O&M range to improve power distribution network operational reliability and solves whole areas of existing power distribution network The tour and the maintenance huge problem of expense of domain route and equipment.
The embodiment of the present invention also provides a kind of readable storage medium storing program for executing, is stored thereon with software program, when the storage medium In instruction by distribution network line fault law mining system processor execute when so that distribution network line fault law mining System is able to carry out the distribution network line fault law mining method as described in above-mentioned any embodiment.This method comprises: acquisition Distribution network line fault tripping record, generates the flag event under distribution network line fault and reason scene；According to the label Event generates the distribution line failure information bank for containing more attribute informations；Method of Data Discretization based on k-means, to institute It states each single item attribute information in distribution line failure information bank to be clustered, obtains attribute preliminary classification collection；By the primary category Property classification carry out equal frequency and optimize, obtain fault attribute category set；Based on apriori algorithm, in the fault attribute category set Every one kind fault attribute carries out the faults frequent item set mining under different faults reason scene, and different frequently attributes are covered in building Several frequent candidates；According to the frequent candidate, it is based on the distribution line failure information bank, carries out distribution line Fault trip strong correlation Rules Filtering.
Certainly, a kind of storage medium comprising computer executable instructions, computer provided by the embodiment of the present invention The distribution network line fault law mining method operation that executable instruction is not limited to the described above can also be performed of the invention any Relevant operation in distribution network line fault law mining method provided by embodiment, and have corresponding function and beneficial effect Fruit.
By the description above with respect to embodiment, it is apparent to those skilled in the art that, the present invention It can be realized by software and required common hardware, naturally it is also possible to which by hardware realization, but in many cases, the former is more Good embodiment.Based on this understanding, technical solution of the present invention substantially in other words contributes to the prior art Part can be embodied in the form of software products, which can store in readable storage medium storing program for executing, such as computer Floppy disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), flash memory (FLASH), hard disk or CD etc., including some instructions are used so that a computer equipment (can be personal meter Calculation machine, server or network equipment etc.) execute distribution network line fault law mining side described in each embodiment of the present invention Method.
Note that the above is only a better embodiment of the present invention and the applied technical principle.It will be appreciated by those skilled in the art that The invention is not limited to the specific embodiments described herein, be able to carry out for a person skilled in the art it is various it is apparent variation, It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out by above embodiments to the present invention It is described in further detail, but the present invention is not limited to the above embodiments only, without departing from the inventive concept, also It may include more other equivalent embodiments, and the scope of the invention is determined by the scope of the appended claims.

Claims (15)

1. a kind of distribution network line fault law mining method, which is characterized in that the described method includes:
Distribution network line fault tripping record is acquired, the flag event under distribution network line fault and reason scene is generated；
According to the flag event, the distribution line failure information bank for containing more attribute informations is generated；
Method of Data Discretization based on k-means carries out each single item attribute information in the distribution line failure information bank Cluster, obtains attribute preliminary classification collection；
The attribute preliminary classification collection is carried out equal frequency to optimize, obtains fault attribute category set；
Based on apriori algorithm, different faults reason scene is carried out to one kind fault attribute every in the fault attribute category set Under faults frequent item set mining, building covers the frequent candidate of different frequently attribute numbers；
According to the frequent candidate, it is based on the distribution line failure information bank, carries out the strong phase of distribution line fault trip Close Rules Filtering.
2. the method according to claim 1, wherein acquisition distribution network line fault tripping record, generates Flag event under distribution network line fault and reason scene, comprising:
Acquire the distribution network line fault tripping record；
Based on dispatch automated system or other dynamic monitoring systems, according to the collected distribution network line fault tripping note Record generates distribution network line fault trip indication event.
3. the method according to claim 1, wherein
According to the flag event, the distribution line failure information bank for containing more attribute informations is generated, comprising:
According to the flag event, the multi-source data based on information system carries out distribution network line feeder line grade effective information Processing；
According to the fault trip record and fault trip time of origin per distribution network line together, the attribute of fault moment is integrated, Generate the distribution line failure information bank for containing more attribute informations；
Wherein, the attribute of the fault moment includes temperature, rainfall, wind speed, route overall length, overhead transmission line overall length, distribution line Affiliated area, distribution line have the operation datas such as distribution transforming quantity, line voltage distribution, electric current, active and reactive and route operation year under its command One of limit or at least two.
4. according to the method described in claim 3, it is characterized in that, the route overall length is cable run and overhead line conductor Section summation；The overhead transmission line overall length is overhead line conductor section summation；
Fault trip record and fault trip time of origin of the basis per distribution network line together, integrate the category of fault moment Property, generate the distribution line failure information bank for containing more attribute informations, comprising:
The each section lead segment length of distribution line is obtained by topological data, the route overall length is calculated and the overhead transmission line is total It is long.
5. according to the method described in claim 3, it is characterized in that, the overhead transmission line overall length, is calculated by following formula It arrives:
Lj=∑ bi
The route overall length, is calculated by following formula:
Lt=∑ ai+∑bi
Wherein, aiAnd biIt is i sections and i sections of aerial condutor section of cable conductor section of length respectively.
6. the method according to claim 1, wherein the Method of Data Discretization based on k-means, matches to described Each single item attribute information is clustered in line fault information bank, obtains attribute preliminary classification collection, comprising:
According to the classification quantity of attribute to be sorted, attributive classification initial value is set；
Using k-means algorithm, the number of iterations is set, solves classification point setting value again.
7. according to the method described in claim 6, it is characterized in that, the classification point setting value is calculated by following formula:
Ej=∑ Eij
E=∑ Ej
Wherein, E is the amount of deviation based on all objects under current class, EjIt is amount of deviation in cluster, EijIt is each object in The distance of heart point；xijThe i element being classified as in j cluster,It is central point in j cluster, with EjAnd EijMinimum target, constantly adjustment are every Element x in a clusterij。
8. optimizing the method according to claim 1, wherein the attribute preliminary classification collection is carried out equal frequency, obtain To fault attribute category set, comprising:
The sample size concentrated to each attribute preliminary classification is ranked up, and each attribute preliminary classification sum and attribute are divided Class quantity preset value compares, and when the attribute preliminary classification sum is less than the attributive classification quantity preset value, it is excellent to carry out equal frequency Change；
The sample size that each attribute preliminary classification is concentrated with attributive classification sample preset value compared with, according to comparison result with The sample size that the attribute preliminary classification is concentrated is ranked up, and the sample size concentrated to each attribute preliminary classification is from height Equal frequency is carried out on earth to optimize, until attributive classification sum is equal to the attributive classification quantity preset value, and then obtains fault attribute Category set.
9. the method according to claim 1, wherein the equal frequency optimization is calculated by following formula:
X=[x1 x2...xN]
X1=[x1 x2...xn]
X2=[xn+1 xn+2...xn+n]
...
Xk'=[xN-n+1 xN-n+2...xN]
Wherein, x is the subclass quantity to be split classified based on a certain classification under k-means algorithm, k ' for x, XiIt is subclass, n is Sample size under each subclassification.
10. classifying the method according to claim 1, wherein being based on apriori algorithm to the fault attribute Every a kind of fault attribute is concentrated to carry out the faults frequent item set mining under different faults reason scene, building is covered difference and frequently belonged to The frequent candidate of property number, comprising:
According to the distribution line failure information bank, it is based on apriori algorithm, the judgement to the frequent item set of distribution line failure Support that count threshold is set；
Count threshold is supported according to the judgement of setting, and the faults frequent item that different frequently attribute numbers are covered in building is candidate Collection.
11. according to the method described in claim 10, it is characterized in that, the frequent of different frequently attribute numbers is covered in the building Candidate, comprising:
The distribution line failure information bank is scanned, k item collection support is calculated, count threshold is supported when number is greater than, is included in k Frequent candidate；
Matching splicing is carried out according to k frequent candidates, generates k+1 item collection；
Support counting screening by k+1 item collection, generates k+1 frequent candidates；
Cycle calculations, until it can not search out more how often candidate；
Wherein, the k item collection support is the number that k attribute value occurs simultaneously.
12. the method according to claim 1, wherein being based on the distribution wire according to the frequent candidate Road fault message storehouse carries out distribution line fault trip strong correlation Rules Filtering, comprising:
According to the frequent candidate, support, the confidence in the distribution line failure information bank under sample total are calculated Degree and promotion degree；
According to the counted support, the confidence level and the promotion degree is counted, the strong phase of distribution line fault trip is carried out Close Rules Filtering.
13. according to the method for claim 12, which is characterized in that the support is calculated by following formula:
The confidence level is calculated by following formula:
The promotion degree is calculated by following formula:
Wherein,It is the support that A frequent item set occurs under B fault scenes,In B failure field The confidence level that A frequent item set occurs under scape,Indicate the simultaneous support counting of A and B, Count (A) is the support counting that A occurs, and D is fault sample total amount；Lift (A, B) is the simultaneous institute of A and B Promotion degree is stated, P (A ∪ B) is A and the B simultaneous probability in full dose sample D, and P (B/A) is probability of B under the conditions of A.
14. a kind of distribution network line fault law mining system characterized by comprising
Flag event generation module generates distribution network line fault and reason for acquiring distribution network line fault tripping record Flag event under scene；
Fault message storehouse generation module, for generating the distribution line failure for containing more attribute informations according to the flag event Information bank；
Hierarchical cluster attribute module, for the Method of Data Discretization based on k-means, to every in the distribution line failure information bank One attribute information is clustered, and attribute preliminary classification collection is obtained；
Equal frequencies optimization module obtains fault attribute category set for the attribute preliminary classification collection to be carried out equal optimization frequently；
Faults frequent item set mining module is based on apriori algorithm, to one kind fault attribute every in the fault attribute category set The faults frequent item set mining under different faults reason scene is carried out, the frequent candidate item of different frequently attribute numbers is covered in building Collection；
Strong correlation Rules Filtering module, for being based on the distribution line failure information bank according to the frequent candidate, into Row distribution line fault trip strong correlation Rules Filtering.
15. a kind of readable storage medium storing program for executing, which is characterized in that when the instruction in the storage medium is by distribution network line fault rule When the processor of digging system executes, so that distribution network line fault law mining system is able to carry out such as claim 1 to 13 Any distribution network line fault law mining method.
CN201910600307.8A 2019-07-04 2019-07-04 Power distribution network line fault rule mining method, system and storage medium Active CN110334127B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910600307.8A CN110334127B (en) 2019-07-04 2019-07-04 Power distribution network line fault rule mining method, system and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910600307.8A CN110334127B (en) 2019-07-04 2019-07-04 Power distribution network line fault rule mining method, system and storage medium

Publications (2)

Publication Number Publication Date
CN110334127A true CN110334127A (en) 2019-10-15
CN110334127B CN110334127B (en) 2022-08-19

Family

ID=68143102

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910600307.8A Active CN110334127B (en) 2019-07-04 2019-07-04 Power distribution network line fault rule mining method, system and storage medium

Country Status (1)

Country Link
CN (1) CN110334127B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112085618A (en) * 2020-06-16 2020-12-15 西安交通大学 Frequent item set-based power distribution equipment fault factor relevance mining method
CN112417763A (en) * 2020-11-25 2021-02-26 杭州凯达电力建设有限公司 Defect diagnosis method, device and equipment for power transmission line and storage medium
CN115278371A (en) * 2021-04-29 2022-11-01 中国移动通信集团河北有限公司 Method, device and equipment for positioning video quality difference content source and storage medium

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103871003A (en) * 2014-03-31 2014-06-18 国家电网公司 Power distribution network fault diagnosis method utilizing historical fault data
CN104820716A (en) * 2015-05-21 2015-08-05 中国人民解放军海军工程大学 Equipment reliability evaluation method based on data mining
CN109358602A (en) * 2018-10-23 2019-02-19 山东中创软件商用中间件股份有限公司 A kind of failure analysis methods, device and relevant device

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103871003A (en) * 2014-03-31 2014-06-18 国家电网公司 Power distribution network fault diagnosis method utilizing historical fault data
CN104820716A (en) * 2015-05-21 2015-08-05 中国人民解放军海军工程大学 Equipment reliability evaluation method based on data mining
CN109358602A (en) * 2018-10-23 2019-02-19 山东中创软件商用中间件股份有限公司 A kind of failure analysis methods, device and relevant device

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112085618A (en) * 2020-06-16 2020-12-15 西安交通大学 Frequent item set-based power distribution equipment fault factor relevance mining method
CN112417763A (en) * 2020-11-25 2021-02-26 杭州凯达电力建设有限公司 Defect diagnosis method, device and equipment for power transmission line and storage medium
CN115278371A (en) * 2021-04-29 2022-11-01 中国移动通信集团河北有限公司 Method, device and equipment for positioning video quality difference content source and storage medium

Also Published As

Publication number Publication date
CN110334127B (en) 2022-08-19

Similar Documents

Publication Publication Date Title
CN110244184A (en) A kind of distribution line fault observer method for digging, system and the medium of frequent item set
CN104899263B (en) A kind of ship track mining analysis and monitoring method based on specific region
CN110334127A (en) A kind of distribution network line fault law mining method, system and storage medium
WO2019191139A1 (en) Processing data for predicting pipe failure
CN110750524A (en) Method and system for determining fault characteristics of active power distribution network
CN104616092B (en) A kind of behavior pattern processing method based on distributed information log analysis
CN104616205A (en) Distributed log analysis based operation state monitoring method of power system
CN105069690A (en) Power grid operating data processing method
CN104239437A (en) Power-network-dispatching-oriented intelligent warning analysis method
CN103902816A (en) Electrification detection data processing method based on data mining technology
CN106373032B (en) The high-incidence region discrimination method of Distribution Network Failure based on big data
CN112131216B (en) Power transmission line self-adaptive database creation method and device based on object model
CN107292478A (en) A kind of disaster influences the acquisition methods of situation on power distribution network
Min et al. A stochastic machine learning based approach for observability enhancement of automated smart grids
CN107784392A (en) A kind of the defects of transmission line of electricity based on machine learning Forecasting Methodology and device
CN110348683A (en) The main genetic analysis method, apparatus equipment of electrical energy power quality disturbance event and storage medium
Eskandarpour et al. Event-driven security-constrained unit commitment with component outage estimation based on machine learning method
CN109902133B (en) Multi-source data error correction processing method and system based on arbitrary partition area of power grid
CN109492055B (en) A method of power grid big data warehouse is established based on CIM model
CN113902583A (en) Distribution network side operation and maintenance method and system using low-voltage network equipment data
EP4186014A1 (en) Spatial power outage estimation for natural hazards leveraging optimal synthetic power networks
JP2003090887A (en) Predication system and prediction method of instantaneous voltage drop by thunderbolt
Munir et al. Extraction of forest power lines from LiDAR point cloud data
CN113011098B (en) Lightning activity rule analysis display model and system for power transmission corridor area
Leite et al. Failure probability metric by machine learning for online risk assessment in distribution 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