CN112732773A - Uniqueness checking method and system for relay protection defect data - Google Patents
Uniqueness checking method and system for relay protection defect data Download PDFInfo
- Publication number
- CN112732773A CN112732773A CN202011478502.7A CN202011478502A CN112732773A CN 112732773 A CN112732773 A CN 112732773A CN 202011478502 A CN202011478502 A CN 202011478502A CN 112732773 A CN112732773 A CN 112732773A
- Authority
- CN
- China
- Prior art keywords
- defect
- record
- attribute
- difference
- pair
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000007547 defect Effects 0.000 title claims abstract description 858
- 238000000034 method Methods 0.000 title claims abstract description 55
- 238000000926 separation method Methods 0.000 claims abstract description 70
- 238000012706 support-vector machine Methods 0.000 claims abstract description 29
- 230000008030 elimination Effects 0.000 claims description 26
- 238000003379 elimination reaction Methods 0.000 claims description 26
- 230000003252 repetitive effect Effects 0.000 claims description 18
- 230000002950 deficient Effects 0.000 claims description 16
- 238000012549 training Methods 0.000 claims description 13
- 238000007405 data analysis Methods 0.000 abstract description 4
- 238000012216 screening Methods 0.000 abstract description 4
- 238000004140 cleaning Methods 0.000 abstract description 3
- 239000000463 material Substances 0.000 abstract description 3
- 238000012795 verification Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 13
- 238000004590 computer program Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 238000012423 maintenance Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000007726 management method Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 238000012217 deletion Methods 0.000 description 3
- 230000037430 deletion Effects 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000011084 recovery Methods 0.000 description 2
- 102100020786 Adenylosuccinate synthetase isozyme 2 Human genes 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
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
- G06F16/2462—Approximate or statistical queries
-
- 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
- G06F16/242—Query formulation
- G06F16/2428—Query predicate definition using graphical user interfaces, including menus and forms
-
- 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
- G06F16/2477—Temporal data queries
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Fuzzy Systems (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Software Systems (AREA)
- Strategic Management (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Human Computer Interaction (AREA)
- Entrepreneurship & Innovation (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Emergency Protection Circuit Devices (AREA)
Abstract
The invention discloses a method and a system for checking uniqueness of relay protection defect data, which comprises the steps of calculating defect attribute difference, defect time difference and defect time span separation of a defect record pair; determining a defect record uniqueness checking interface by using a support vector machine according to a plurality of sample defect record pairs; determining the uniqueness of the defect record pair to be verified according to the relative position of the data point corresponding to the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of the defect record pair to be verified and the defect record uniqueness verification interface; the method can effectively support relay protection professional managers to analyze the relay protection defects, can accurately judge whether any 2 defect records of the same equipment are repeated or not, can reduce the workload of manual data analysis, reduces the cost of manpower and material resources, and can improve the comprehensive application level of the relay protection defect data by screening and cleaning the repeated defect data.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a uniqueness checking method and a uniqueness checking system for relay protection defect data.
Background
The relay protection data can be divided into types of equipment management, operation maintenance, state evaluation, constant value setting and the like. The equipment management data comprises a protected equipment standing book and the relation between the standing book and the primary equipment and the plant station. The fixed value setting data comprises a protection fixed value model, a fixed value item, a relation between a protection fixed value and a system operation mode and the like. The state evaluation class data model comprises IED equipment information generated based on an intelligent substation configuration file (SCD file), and information models such as state quantity, alarm, action and analog quantity corresponding to each IED equipment. The operation maintenance data comprises routing inspection tour, defects, familial defect data and the like. The relay protection defect data is equipment operation maintenance data for describing the operation abnormity and the fault of the relay protection equipment, and can embody information such as equipment operation reliability, overall quality of different types of equipment and the like.
At present, the management and application of relay protection professional information still stay in the single information acquisition application layer, the problems of accuracy, integrity and standard of data are outstanding, and automatic information acquisition and intelligent input management of the data from a protection device source end to a transformer substation end and a scheduling end are not realized. Therefore, research work of identification technology and data checking technology of the relay protection device needs to be developed, unique identification of data of the whole network protection equipment is achieved, a whole life cycle information model of the protection equipment is comprehensively perfected, a foundation is laid for accurate and complete collection and analysis application of relay protection information, and relevant application of each application system of relay protection specialties based on identification information is achieved.
The uniqueness of relay protection data is an important aspect of data governance. Repeated relay protection defect records often exist in a relay protection professional system, and the application level of defect data is influenced.
Disclosure of Invention
The invention provides a uniqueness checking method and a uniqueness checking system for relay protection defect data, and aims to solve the problem of how to automatically perform uniqueness checking on the relay protection defect data.
In order to solve the above problem, according to an aspect of the present invention, there is provided a uniqueness checking method for relay protection defect data, the method including:
acquiring at least two defect records to be verified corresponding to the same relay protection equipment identifier, and selecting two defect records to be verified in the at least two defect records to be verified as a defect record pair to be verified;
calculating the weight of each defect attribute according to the statistical result of each attribute of the relay protection total defect record, and calculating the defect attribute difference degree of the defect record pair to be verified according to the defect attribute data in the defect record pair to be verified and the weight of each defect attribute
Determining the defect time difference degree and the defect time span separation degree of the defect record pair to be verified according to the defect time in the defect record pair to be verified;
obtaining a plurality of sample defect record pairs, calculating the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of each sample defect record pair, and determining a defect record uniqueness checking interface by using a support vector machine according to data points formed by the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of each sample defect record pair;
determining a uniqueness checking result of the defect record pair to be checked according to the relative position of the data point corresponding to the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of the defect record pair to be checked and the defect record uniqueness checking interface; wherein the uniqueness check result comprises: a repetitive defect recording pair and a non-repetitive defect recording pair.
Preferably, the method calculates the weight of each defect attribute according to the statistical result of each attribute of the relay protection total defect record by using the following method, including:
wherein, ω isiThe weight of the ith defect attribute; pijThe probability of occurrence of the jth value of the ith defect attribute is obtained; k is the number of different values of the ith defect attribute.
Preferably, the method calculates the defect attribute difference of the defect record pair according to the weight of each defect attribute by using the following formula, including:
wherein, V1A defect attribute difference degree for the defect record pair; omegaiThe weight of the ith defect attribute; i isiAnd the data indicating whether the defect record has difference to the i-th defect attribute is 1 if the defect record has difference and 0 if the defect record has no difference.
Preferably, the method for determining the defect time difference degree and the defect time span separation degree of the defect record pair according to the defect time in the defect record pair by using the following modes comprises the following steps:
V2＝(V21+V22)/2，
wherein, V2A defect time difference degree for the defect record pair; v21The degree of difference in defect occurrence time for the defect record pair; v22Eliminating time difference degrees for the defects of the defect record pairs; t isdif1And Tdif2Respectively recording the difference of defect occurrence time and defect elimination time in the defect record pair, and the unit is minute; v3The defect time span separation of pairs is recorded for defects.
Preferably, the determining a defect record uniqueness checking interface by using a support vector machine according to data points consisting of the defect attribute difference, the defect time difference and the defect time span separation of each sample defect record pair includes:
constructing an index space with the defect attribute difference, the defect time difference and the defect time span separation as coordinates;
determining data points consisting of defect attribute difference, defect time difference and defect time span separation of each sample defect record pair in the index space, clustering according to the prior result of whether each sample defect record pair is a repeated defect record, and determining a repeated defect record cluster and a non-repeated defect record cluster;
and training by adopting a support vector machine according to the repeated defective cluster and the non-repeated defective cluster to determine a defect record uniqueness checking interface.
According to another aspect of the present invention, there is provided a system for checking uniqueness of relay protection defect data, the system including:
the device comprises a to-be-verified defect record pair determining unit, a to-be-verified defect record pair determining unit and a to-be-verified defect record pair determining unit, wherein the to-be-verified defect record pair determining unit is used for acquiring at least two to-be-verified defect records corresponding to the same relay protection device identifier, and selecting two to-be-verified defect records in the at least two to-be-verified defect records as to-be-verified defect;
the defect attribute difference degree determining unit is used for calculating the weight of each defect attribute according to the statistical result of each attribute of the whole relay protection defect record, and calculating the defect attribute difference degree of the defect record pair to be verified according to the defect attribute data in the defect record pair to be verified and the weight of each defect attribute;
a defect time difference determining unit, configured to determine a defect time difference and a defect time span separation of the defect record pair to be verified according to a defect time in the defect record pair to be verified;
the defect record uniqueness checking interface determining unit is used for acquiring a plurality of sample defect record pairs, calculating the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of each sample defect record pair, and determining a defect record uniqueness checking interface by using a support vector machine according to data points formed by the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of each sample defect record pair;
the uniqueness checking result determining unit is used for determining the uniqueness checking result of the defect record pair to be checked according to the relative position of the data point corresponding to the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of the defect record pair to be checked and the defect record uniqueness checking interface; wherein the uniqueness check result comprises: a repetitive defect recording pair and a non-repetitive defect recording pair.
Preferably, the system calculates the weight of each defect attribute according to the statistical result of each attribute of the relay protection total defect record by using the following method, including:
wherein, ω isiThe weight of the ith defect attribute; pijThe probability of occurrence of the jth value of the ith defect attribute is obtained; k is the number of different values of the ith defect attribute.
Preferably, the system calculates the defect attribute difference of the defect record pair according to the weight of each defect attribute by using the following formula, including:
wherein, V1A defect attribute difference degree for the defect record pair; omegaiThe weight of the ith defect attribute; i isiAnd the data indicating whether the defect record has difference to the i-th defect attribute is 1 if the defect record has difference and 0 if the defect record has no difference.
Preferably, the system determines the defect time difference degree and the defect time span separation degree of the defect record pair according to the defect time in the defect record pair by using the following modes:
V2＝(V21+V22)/2，
wherein, V2A defect time difference degree for the defect record pair; v21The degree of difference in defect occurrence time for the defect record pair; v22Eliminating time difference degrees for the defects of the defect record pairs; t isdif1And Tdif2Respectively recording the difference of defect occurrence time and defect elimination time in the defect record pair, and the unit is minute; v3The defect time span separation of pairs is recorded for defects.
Preferably, the determining unit of the defect record uniqueness checking interface determines the defect record uniqueness checking interface by using a support vector machine according to data points consisting of the defect attribute difference, the defect time difference and the defect time span separation of each sample defect record pair, including:
constructing an index space with the defect attribute difference, the defect time difference and the defect time span separation as coordinates;
determining data points consisting of defect attribute difference, defect time difference and defect time span separation of each sample defect record pair in the index space, clustering according to the prior result of whether each sample defect record pair is a repeated defect record, and determining a repeated defect record cluster and a non-repeated defect record cluster;
and training by adopting a support vector machine according to the repeated defective cluster and the non-repeated defective cluster to determine a defect record uniqueness checking interface.
The invention provides a method and a system for checking uniqueness of relay protection defect data, which are used for calculating defect attribute difference, defect time difference and defect time span separation of a defect record pair to be checked; determining a defect record uniqueness checking interface by using a support vector machine according to a plurality of sample defect record pairs; determining a uniqueness checking result of the defect record pair to be checked according to the relative position of the data point corresponding to the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of the defect record pair to be checked and the defect record uniqueness checking interface; the method can effectively support relay protection professional managers to analyze the relay protection defects, can accurately judge whether any 2 defect records of the same equipment are repeated or not, can reduce the workload of manual data analysis, reduces the cost of manpower and material resources, and can improve the comprehensive application level of the relay protection defect data by screening and cleaning the repeated defect data.
Drawings
A more complete understanding of exemplary embodiments of the present invention may be had by reference to the following drawings in which:
fig. 1 is a flowchart of a uniqueness checking method 100 for relay protection defect data according to an embodiment of the present invention;
FIG. 2 is a diagram illustrating the determination of the intersection and union of defect time spans for two defect records according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a support vector machine according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating a defect entry uniqueness checking interface according to an embodiment of the invention;
fig. 5 is a schematic structural diagram of a uniqueness checking system 500 for relay protection defect data according to an embodiment of the present invention.
Detailed Description
The exemplary embodiments of the present invention will now be described with reference to the accompanying drawings, however, the present invention may be embodied in many different forms and is not limited to the embodiments described herein, which are provided for complete and complete disclosure of the present invention and to fully convey the scope of the present invention to those skilled in the art. The terminology used in the exemplary embodiments illustrated in the accompanying drawings is not intended to be limiting of the invention. In the drawings, the same units/elements are denoted by the same reference numerals.
Unless otherwise defined, terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Further, it will be understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense.
Fig. 1 is a flowchart of a uniqueness checking method 100 for relay protection defect data according to an embodiment of the present invention. As shown in fig. 1, in the uniqueness checking method for relay protection defect data provided by the embodiment of the present invention, a defect attribute difference, a defect time difference, and a defect time span separation of a defect record pair to be checked are calculated; determining a defect record uniqueness checking interface by using a support vector machine according to a plurality of sample defect record pairs; determining a uniqueness checking result of the defect record pair to be checked according to the relative position of the data point corresponding to the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of the defect record pair to be checked and the defect record uniqueness checking interface; the method can effectively support relay protection professional managers to analyze the relay protection defects, can accurately judge whether any 2 defect records of the same equipment are repeated or not, can reduce the workload of manual data analysis, reduces the cost of manpower and material resources, and can improve the comprehensive application level of the relay protection defect data by screening and cleaning the repeated defect data. In the uniqueness checking method 100 for relay protection defect data provided by the embodiment of the present invention, starting from step 101, at least two defect records to be checked corresponding to the same relay protection device identifier are obtained in step 101, and two defect records to be checked in the at least two defect records to be checked are selected as a defect record pair to be checked.
The physical identification I D of the relay protection device is a unique code of the relay protection device, and is a code specification uniquely representing one protection device in different analysis, application and management occasions. Therefore, the invention establishes the relay protection defect data index and extracts the defect record pair to check the uniqueness of the defect data.
In the embodiment of the invention, the uniqueness checking process of the relay protection defect data comprises the following steps: firstly, with the physical identification ID of the relay protection equipment as a query condition, searching a plurality of defect records corresponding to the same physical identification ID in a relay protection defect data table, and randomly extracting two defect records corresponding to the same physical identification ID according to returned query result information to combine to obtain a defect record pair to be checked. Then, calculating the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of the defect record pair to be detected, judging whether the record is a repeated record according to the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of the defect record pair to be detected, and returning the repeated record to the data source end for processing.
In step 102, calculating the weight of each defect attribute according to the statistical result of each attribute of the relay protection total defect record, and calculating the defect attribute difference degree of the defect record pair to be verified according to the defect attribute data in the defect record pair to be verified and the weight of each defect attribute.
Preferably, the method calculates the weight of each defect attribute according to the statistical result of each attribute of the relay protection total defect record by using the following method, including:
wherein, ω isiThe weight of the ith defect attribute; pijThe probability of occurrence of the jth value of the ith defect attribute is obtained; k is the number of different values of the ith defect attribute.
Preferably, the method calculates the defect attribute difference of the defect record pair according to the weight of each defect attribute by using the following formula, including:
wherein, V1A defect attribute difference degree for the defect record pair; omegaiThe weight of the ith defect attribute; i isiAnd the data indicating whether the defect record has difference to the i-th defect attribute is 1 if the defect record has difference and 0 if the defect record has no difference.
The core of the method is to analyze the typical difference of the repeated record and the non-repeated record on the characteristics, determine the index for judging the repeatability of the defect record pair, determine the classification principle and judge the repeated condition of the defect record pair according to the classification principle.
If two defect records with the same physical ID are different records, the 2 records should have a certain difference in content. For example, for relay protection defect data, different defects should have a certain difference in defect degree, defect equipment classification, defect cause, whether protection exits, defect position, and specific defect cause, and then a difference degree index can be constructed by reflecting whether the to-be-verified defects have a difference in these attributes, and the difference degree index of the defect attributes is set. In addition, the difference in time between the two recordings is a key feature for the uniqueness check for the different recordings. In the relay protection defect data, the attributes related to time include defect occurrence time, defect elimination time, protection exit time, and protection recovery time. Because the defect does not necessarily cause the protection to quit operation, and the protection quit time and the protection recovery time do not necessarily exist, the defect time difference degree is set as a check index of the defect record difference degree based on the difference between the defect occurrence time and the defect elimination time. In addition to the difference between the defect occurrence time and the defect elimination time, the relay protection defect data is operation and maintenance data, the field filling has certain subjectivity, and the defect occurrence time and the defect elimination time filled by operation and maintenance personnel are estimated time, so that the repeatedly filled defect data has certain difference between the defect occurrence time and the defect elimination time, and the judgment method using the absolute value of the single use time as an index judges the repeated data to be non-repeated data by mistake. However, multiple records in the field for the same defect filling are usually intersected in a defect time period, and the uniqueness check of the records by adding indexes according to the intersection degree is strong. Therefore, the defect time span separation degree is used as a check index at the same time.
In the implementation mode of the invention, the weight of each defect attribute is calculated according to the statistical result of each attribute of the whole relay protection defect record, and the defect attribute difference of the defect record pair to be verified is calculated according to the defect attribute data in the defect record pair to be verified and the weight of each defect attribute.
The data of the defect attribute of the relay protection equipment are mainly embodied in the defect degree, the defect equipment classification, the defect reason, whether protection exits, the defect part and the specific defect reason, and the difference index of the defect attribute is defined as the proportion of the number of different attributes between two records to the total number. The defect attribute includes a certain difference in the amount of information, for example: whether the protection exits from only 2 options, and the defect part has 64 options, obviously, the attribute of the defect part can reflect the difference degree of two defects, and the difference degree of the defect information should have higher weight.
Entropy is a most common measure of sample set purity. Assuming that the ratio of the kth type sample in the current sample set D is pk(k ═ 1, 2, …, | y |), then the entropy of information for D is defined as:
the smaller the value of Ent (D), the higher the purity of D.
Therefore, the weights of different defect attributes are set according to the entropy theory. Calculating the weight of each defect attribute according to the statistical result of each attribute of the relay protection overall defect record by using the following method, including:
wherein, ω isiIs as followsThe weight of the i defect attributes; pijThe probability of occurrence of the jth value of the ith defect attribute is obtained; k is the number of different values of the ith defect attribute.
For example, the weights of the different defect attributes calculated are shown in table 1.
TABLE 1 weight setting of different Defect attributes
| Information | Degree of defect | Defect device classification | Cause of defect | Protection of exit | Defective part | Specific defect cause attributes |
| Weight of | 0.078 | 0.197 | 0.138 | 0.049 | 0.295 | 0.242 |
Calculating the defect attribute difference degree of the defect record pair according to the weight of each defect attribute by using the following formula, wherein the defect attribute difference degree comprises the following steps:
wherein, V1A defect attribute difference degree for the defect record pair; omegaiThe weight of the ith defect attribute; i isiAnd the data indicating whether the defect record has difference to the i-th defect attribute is 1 if the defect record has difference and 0 if the defect record has no difference. If the defect attribute is 6 in table 1, then i is 6.
In step 103, determining the defect time difference and the defect time span separation of the defect record pair to be verified according to the defect time in the defect record pair to be verified.
Preferably, the method for determining the defect time difference degree and the defect time span separation degree of the defect record pair according to the defect time in the defect record pair by using the following modes comprises the following steps:
V2＝(V21+V22)/2，
wherein, V2A defect time difference degree for the defect record pair; v21The degree of difference in defect occurrence time for the defect record pair; v22Eliminating time difference degrees for the defects of the defect record pairs; t isdif1And Tdif2Respectively recording the difference of defect occurrence time and defect elimination time in the defect record pair, and the unit is minute; v3The defect time span separation of pairs is recorded for defects.
The time difference of different defect records is distributed in the range of minutes to years, the difference degree index of the defect time performs logarithmic calculation on the time difference, and index values of verification objects with different magnitude time differences are reasonably distributed on a (0,1) interval. For an application scene of uniqueness check of relay protection defect data at a certain degree, the defect recording time difference of the same relay protection object identification ID is not more than 1 year, and the defect time difference degree of a defect record pair is as follows:
V2＝(V21+V22)/2，
wherein, V2A defect time difference degree for the defect record pair; v21The degree of difference in defect occurrence time for the defect record pair; v22Eliminating time difference degrees for the defects of the defect record pairs; t isdif1And Tdif2The difference between the defect occurrence time and the defect removal time in the defect record pair is expressed in minutes, respectively.
The difference index values of different defect difference times calculated according to the above formula are shown in table 2.
TABLE 2 Difference index values at different defect difference times
| Time difference | 1 minute | 2 minutes | 5 minutes | 1 hour | 1 day | 1 month | 1 year |
| V2Value of (A) | 0.0526 | 0.0834 | 0.1360 | 0.3121 | 0.5522 | 0.8103 | 1 |
In the embodiment of the present invention, the defect time span separation degree V of a defect recording pair is calculated by the following formula3The method comprises the following steps:
as shown in fig. 2, the method for calculating the intersection and union of defect time spans is to extract the occurrence and elimination times of 2 defect records in a defect record pair to form a defect duration interval, and the intersection and union of the 2 intervals are the intersection and union of the defect time spans.
In step 104, a plurality of sample defect record pairs are obtained, the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of each sample defect record pair are calculated, and a support vector machine is used for determining the uniqueness checking interface of the defect record according to data points formed by the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of each sample defect record pair.
Preferably, the determining a defect record uniqueness checking interface by using a support vector machine according to data points consisting of the defect attribute difference, the defect time difference and the defect time span separation of each sample defect record pair includes:
constructing an index space with the defect attribute difference, the defect time difference and the defect time span separation as coordinates;
determining data points consisting of defect attribute difference, defect time difference and defect time span separation of each sample defect record pair in the index space, clustering according to the prior result of whether each sample defect record pair is a repeated defect record, and determining a repeated defect record cluster and a non-repeated defect record cluster;
and training by adopting a support vector machine according to the repeated defective cluster and the non-repeated defective cluster to determine a defect record uniqueness checking interface.
The Support Vector Machine (SVM) is a classification tool with good performance, and the basic idea is to construct an optimal hyperplane in a sample space or a feature space, so that the distance between the hyperplane and a sample set of different classes is maximum, and the maximum generalization capability is achieved. The SVM is developed from an optimal classification surface in a linear separable case, and the basic idea can be illustrated by two types of linear separable cases of fig. 3. The solid and open dots represent two types of samples, and the solid lines P0, P1 are classification lines.
Let the linear separable sample set be:
(x1,y1),(x2,y2),…,(xN,yN)；xi∈Rd；yi∈{1,-1} (1)
wherein, x is d-dimensional vector, y is category label, that is, samples are divided into two categories.
The hyperplane formula exists as follows:
f(x)＝(w·x)-b＝0 (2)
all N samples can be separated without error. Wherein w is a d-dimensional weight phasor; b is a constant term.
If a hyperplane exists, the training samples can be separated without errors, and the distance between the sample closest to the hyperplane in each type of training sample and the hyperplane is as large as possible, the hyperplane is called an optimal classification hyperplane. The process of finding the optimal hyperplane is the model training process of the support vector machine.
The algorithm has the advantages that the algorithm is not influenced by noise data too much, overfitting is not easy to occur, and the algorithm is easier to use compared with a neural network algorithm; the disadvantages are that the best model is found by testing different functions and model parameter combinations, the training is slow, and the formed complex black box model is difficult to interpret. The support vector machine algorithm is mainly applied to the fields of pattern recognition, function approximation, probability density estimation and the like.
In the embodiment of the invention, any 2 records are selected from a plurality of records with the same relay protection object identifier I D, a plurality of defect record sample pairs are determined, and the three defect difference indexes are calculated. Because whether the defect record is the repeated record is known, point tracing is carried out on the space, a point cluster with the repeated defect and a point cluster with the non-repeated defect can be obtained, and the defect record uniqueness checking interface can be obtained by utilizing SVM analysis training. And for the defect record pair to be checked, calculating three indexes of the defect record pair, and judging whether the defect record is a repeated defect record or not by utilizing the defect record uniqueness checking interface.
For example, 2019 relay protection data is used as an analysis target. And establishing a relay protection defect index by taking the ID of the physical object of the protection equipment as a key attribute, wherein the relay protection defect index comprises the following steps: {1995305,{0500000000002190,0500000000002083,0500000000002075}}. Wherein 1995305 is the serial number of 3 defect records of the protection device with physical identification ID of protection device, 0500000000002190 and other 3 physical identification IDs of 1995305. And (3) randomly selecting 2 defect records of the protection equipment, and arranging and combining the defect records to obtain 3 defect pairs.
The model for checking the uniqueness of the relay protection defect is a two-class linear separable problem, and the input of the checking of the uniqueness of the relay protection defect is 3 fingersThe standard-defect attribute difference degree, defect time difference degree and defect time span separation degree, and the output is whether the defect pair is repeated, namely a repeated defect record pair and a non-repeated defect record pair. All defect data in 2019 are taken as training samples, 737 defect pairs are collected in total, and the defect pairs come from 624 protection devices. Calculating the difference degree V of defect attributes of 737 defect record pairs1Degree of defect time difference V2Defect time span separation V3Three indexes are provided. In order to obtain the data uniqueness checking rule, it is necessary to manually determine whether the 737 defect pairs are repetitive defects. Then with V1、V2、V3Constructing an index space for coordinates, i.e. by V1、V2、V3And the three indexes are taken as a 3-dimensional rectangular coordinate system of coordinates of a horizontal axis, a vertical axis and a vertical axis, so that any point in the space is a space point taking the three indexes of the defect record pair as coordinates. And respectively describing the repeated defect recording pair and the non-repeated defect recording pair by combining the artificial judgment of whether the repeated recording is actually performed or not to obtain the point cluster. Of the 737 defective recording pairs, 512 recording pairs are non-repeating, the dot clusters of which are shown by the dots in fig. 4, and the remaining 225 recording pairs are repeating, the dot clusters of which are shown by the triangular dots in fig. 4. The optimal classification surface can be obtained by training by adopting a support vector machine, and the equation is as follows:
0.0096V1+0.0504V2+0.982V3-1＝0
this optimal plane is the spatial plane in fig. 4. V1、V2、V3And respectively representing the indexes of the defect attribute difference degree, the defect time difference degree and the defect time span separation degree. In this formula, x ═ V corresponding to formulas (1) and (2)1,V2,V3]，w＝[0.0096,0.0504,0.982],b＝1。
In step 105, determining a uniqueness check result of the defect record pair to be checked according to the relative position of the data point corresponding to the defect attribute difference, the defect time difference and the defect time span separation of the defect record pair to be checked and the defect record uniqueness check interface; wherein the uniqueness check result comprises: a repetitive defect recording pair and a non-repetitive defect recording pair.
Specifically, for a new set of defect record pairs, the defect attribute difference degree, the defect time difference degree and the defect time span separation degree index of the defect record pair are calculated to form [ V1,V2,V3]Substituting into equation f (V)1,V2,V3)＝0.0096V1+0.0504V2+0.982V3-1. If f (V)1,V2,V3) ≧ 0, meaning that the defect pair is non-repetitive, if f (V)1,V2,V3)<0, indicates that the defect pair is repetitive.
In the embodiment of the invention, in the model checking stage, 2020-year-old protection equipment defect data is selected, a defect index is constructed according to the real object identification ID of the relay protection equipment, a defect record pair with the same protection real object identification ID is selected in the index, and three indexes (V) are calculated1,V2,V3) And judging the position of the space point relative to the interface so as to divide the defect record pair into a repeated defect record or a non-repeated defect record area, judging whether the defect record pair is the repeated defect record or not, and realizing the uniqueness check of the relay protection defect data.
Among the 3 indexes of defect uniqueness check, a defect time span separation degree index V3The good classification effect is exerted. V of non-repeating defect3Near 1, repeating defective V3Less than 1. In addition, the difference degree index V of defect time2Also has good classification effect. In general, the degree of defect time difference>0.5 is no repeated defect, and the time difference between the occurrence time and the defect elimination time of 2 defects in the corresponding defect pair is 1 day or more. The classification effect of the difference degree index of the defect attributes is the worst, and the reason is that even if the defects are the same, different operators on site have certain difference in understanding of the defects, so that two defects reflecting the same defects are recorded on the defect parts and the filling of information such as specific defect reasons is inconsistent.
The method can effectively support the screening of the repeated relay protection defect data, and give a conclusion whether any 2 defects of the same device are repeated records, so that the judgment accuracy is high. The suspected repeated defect records are provided for the professionals, the professionals can determine whether the recorded defects are repeated defects only with a small amount of investment, the professionals are prevented from directly carrying out uniqueness judgment and data management on mass relay protection defect data, the workload of manual data analysis is reduced, the repeated defect data are effectively screened and cleaned, and the comprehensive application level of the relay protection defect data is improved.
The application effect of the method in a typical defect uniqueness checking scene is as follows:
scene 1: the method establishes a defect index aiming at the physical identification ID of the equipment. For the defects of different protection devices, the defects are not under the same index, and the method can automatically judge as non-repeated defect data;
scene 2: the 2 defect records of the same protection device have no intersection in the span interval from the defect occurrence time to the defect elimination time, and the defect occurrence time and the defect elimination time of the 2 records have certain difference, and the method can judge that the 2 defect records are non-repeated records, thereby effectively reducing the workload of manual analysis of personnel;
scene 3: the same defect is repeatedly reported, and the difference degree of the reported discrete items (specifically, the defect degree, the defect equipment classification, the defect reason and the like) is large. Because the defect occurrence and elimination time of 2 defect records are close, the method can eliminate the interference of large difference degree of discrete items and accurately judge the defects as repeated defects. For example, the defect of channel abnormality caused by power supply defect occurred in 2020, and the operation and maintenance personnel 1 reports the defect of channel interface equipment, the defect caused by poor device manufacture, the general defect, the responsibility in the manufacturing department, and the defect part in the power supply plug-in; the operation and maintenance personnel 2 report the defect of the communication transmission equipment, the defect caused by other reasons, serious defect, responsibility in other departments and the defect position in an optical cable (OPGW or ADSS). V of the defect pair1,V2,V3]＝[1,0,0]. In fact, the index V1The effect of judging whether the equipment is the repeated defect is limited, for example, the slope of the boundary surface to the index is extremely low in fig. 4, and the index can only play a role in auxiliary judgmentThe function of (1);
scene 4: the same device generates 2 times of defects with the same reason and eliminates the defects by adopting the same method. Because the time span from the occurrence of the 2-time defects to the elimination of the 2-time defects is not overlapped and the occurrence time and the elimination time have certain difference, the method can be used for correctly judging the non-repetitive defects of the 2-time defects. For example, in the record of deletion in 2020, a power supply plug-in unit damaged by a certain device is replaced for 3 days, and then the power supply plug-in unit is damaged and replaced again. V of the defect pair1,V2,V3]＝[0.1275,0.6167,1]；
Scene 5: the defects are temporarily eliminated by adopting different defect eliminating schemes in the same device, but other defects of the device occur in a short period of time, and other defect eliminating means are needed to eliminate the defects. In the same scene 4, the time span from the occurrence of the 2-time defect to the elimination is not overlapped, and the occurrence time and the elimination time have certain difference, so that the invention can also correctly judge the 2-time defect non-repetitive defect. For example, in the defect elimination record of 2020, a device is defective after a short period of operation after the power supply plug-in is replaced, and then the CPU plug-in is replaced to complete defect elimination again. V of the defect pair1,V2,V3]＝[0,0.5756,1.0000]；
Scene 6: the defect occurrence time of two defect records of the same device is similar, and the defect elimination time of the two defect records is similar to the defect occurrence time, and the defect elimination time of one of the two defect records is longer than the defect occurrence time, so that the difference degree V of the defect time is caused2Is relatively large. The 2 defect records are analyzed to have crossed defects in the time span from defect occurrence to elimination, and are most likely to be repeated defects, and the index set by the invention, namely the defect time span separation degree V3The uniqueness judgment of the defect record pair in the scene can be supported. Fig. 4 shows that unless 2 defect entries have no crossover in the time interval from defect occurrence to defect elimination, there is a great probability that the entries are classified as repetitive defect entries. One record shows that the deletion is completed on the same day and the other record shows that the deletion is completed 2 days later in the defect of the damage of the CPU plug-in 2020,indicating that the defect was repopulated repeatedly. V of defect recording pair1,V2,V3]＝[0.0782,0.4952,0.8868]The invention can accurately classify the cases into the repeated defects.
By using the method of the invention, the suspected repeated defect records can be submitted to a professional so as to facilitate the further processing of the data.
Although only the uniqueness check of the relay protection defect data is taken as an analysis object, the repeated characteristics of other types of relay protection data, such as relay protection action, routing inspection, alarm and the like, have certain similarity with the repeated characteristics of the relay protection defect data.
Fig. 5 is a schematic structural diagram of a uniqueness checking system 500 for relay protection defect data according to an embodiment of the present invention. As shown in fig. 5, a system 500 for checking uniqueness of relay protection defect data according to an embodiment of the present invention includes: a defect record pair to be verified determining unit 501, a defect attribute difference determining unit 502, a defect time difference determining unit 503, a defect record uniqueness checking interface determining unit 504 and a uniqueness checking result determining unit 505.
Preferably, the defect record pair determining unit 501 is configured to obtain at least two defect records to be verified corresponding to the same relay protection device identifier, and select two defect records to be verified in the at least two defect records to be verified as a defect record pair to be verified.
Preferably, the defect attribute difference degree determining unit 502 is configured to calculate a weight of each defect attribute according to a statistical result of each attribute of all the defect records of the relay protection, and calculate the defect attribute difference degree of the defect record pair to be verified according to the defect attribute data in the defect record pair to be verified and the weight of each defect attribute.
Preferably, the system calculates the weight of each defect attribute according to the statistical result of each attribute of the relay protection total defect record by using the following method, including:
wherein, ω isiThe weight of the ith defect attribute; pijThe probability of occurrence of the jth value of the ith defect attribute is obtained; k is the number of different values of the ith defect attribute.
Preferably, the system calculates the defect attribute difference of the defect record pair according to the weight of each defect attribute by using the following formula, including:
wherein, V1A defect attribute difference degree for the defect record pair; omegaiThe weight of the ith defect attribute; i isiAnd the data indicating whether the defect record has difference to the i-th defect attribute is 1 if the defect record has difference and 0 if the defect record has no difference.
Preferably, the defect time difference determining unit 503 is configured to determine the defect time difference and the defect time span separation of the defect record pair to be verified according to the defect time in the defect record pair to be verified.
Preferably, the system determines the defect time difference degree and the defect time span separation degree of the defect record pair according to the defect time in the defect record pair by using the following modes:
V2＝(V21+V22)/2，
wherein, V2A defect time difference degree for the defect record pair; v21The degree of difference in defect occurrence time for the defect record pair; v22Eliminating time difference degrees for the defects of the defect record pairs; t isdif1And Tdif2Respectively recording the difference of defect occurrence time and defect elimination time in the defect record pair, and the unit is minute; v3The defect time span separation of pairs is recorded for defects.
Preferably, the defect record uniqueness checking interface determining unit 504 is configured to obtain a plurality of sample defect record pairs, calculate a defect attribute difference, a defect time difference, and a defect time span separation degree of each sample defect record pair, and determine the defect record uniqueness checking interface by using a support vector machine according to a data point composed of the defect attribute difference, the defect time difference, and the defect time span separation degree of each sample defect record pair.
Preferably, the determining unit 504 for determining the uniqueness checking interface of defect record by using a support vector machine according to the data point consisting of the difference of defect attribute, the difference of defect time and the separation of defect time span of each sample defect record pair includes:
constructing an index space with the defect attribute difference, the defect time difference and the defect time span separation as coordinates;
determining data points consisting of defect attribute difference, defect time difference and defect time span separation of each sample defect record pair in the index space, clustering according to the prior result of whether each sample defect record pair is a repeated defect record, and determining a repeated defect record cluster and a non-repeated defect record cluster;
and training by adopting a support vector machine according to the repeated defective cluster and the non-repeated defective cluster to determine a defect record uniqueness checking interface.
Preferably, the uniqueness checking result determining unit 505 is configured to determine the uniqueness checking result of the defect record pair to be checked according to the relative position of the data point corresponding to the defect attribute difference, the defect time difference, and the defect time span separation degree of the defect record pair to be checked and the defect record uniqueness checking interface; wherein the uniqueness check result comprises: a repetitive defect recording pair and a non-repetitive defect recording pair.
The uniqueness checking system 500 for relay protection defect data according to the embodiment of the present invention corresponds to the uniqueness checking method 100 for relay protection defect data according to another embodiment of the present invention, and is not described herein again.
The invention has been described with reference to a few embodiments. However, other embodiments of the invention than the one disclosed above are equally possible within the scope of the invention, as would be apparent to a person skilled in the art from the appended patent claims.
Generally, all terms used in the claims are to be interpreted according to their ordinary meaning in the technical field, unless explicitly defined otherwise herein. All references to "a/an/the [ device, component, etc ]" are to be interpreted openly as referring to at least one instance of said device, component, etc., unless explicitly stated otherwise. The steps of any method disclosed herein do not have to be performed in the exact order disclosed, unless explicitly stated.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims.
Claims (10)
1. A uniqueness checking method for relay protection defect data is characterized by comprising the following steps:
acquiring at least two defect records to be verified corresponding to the same relay protection equipment identifier, and selecting two defect records to be verified in the at least two defect records to be verified as a defect record pair to be verified;
calculating the weight of each defect attribute according to the statistical result of each attribute of the relay protection whole defect record, and calculating the defect attribute difference degree of the defect record pair to be verified according to the defect attribute data in the defect record pair to be verified and the weight of each defect attribute;
determining the defect time difference degree and the defect time span separation degree of the defect record pair to be verified according to the defect time in the defect record pair to be verified;
obtaining a plurality of sample defect record pairs, calculating the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of each sample defect record pair, and determining a defect record uniqueness checking interface by using a support vector machine according to data points formed by the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of each sample defect record pair;
determining a uniqueness checking result of the defect record pair to be checked according to the relative position of the data point corresponding to the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of the defect record pair to be checked and the defect record uniqueness checking interface; wherein the uniqueness check result comprises: a repetitive defect recording pair and a non-repetitive defect recording pair.
2. The method of claim 1, wherein the method calculates the weight of each defect attribute according to the statistical result of each attribute of the relay protection total defect records by using the following method, comprising:
wherein, ω isiThe weight of the ith defect attribute; pijThe probability of occurrence of the jth value of the ith defect attribute is obtained; k is the number of different values of the ith defect attribute.
3. The method of claim 1, wherein calculating the defect attribute variance of the defect entry pair according to the weight of each defect attribute using the following formula comprises:
wherein, V1A defect attribute difference degree for the defect record pair; omegaiThe weight of the ith defect attribute; i isiAnd the data indicating whether the defect record has difference to the i-th defect attribute is 1 if the defect record has difference and 0 if the defect record has no difference.
4. The method of claim 1, wherein determining the defect time disparity and defect time span separation for a defect record pair from defect times in the defect record pair comprises:
V2＝(V21+V22)/2，
wherein, V2A defect time difference degree for the defect record pair; v21The degree of difference in defect occurrence time for the defect record pair; v22Eliminating time difference degrees for the defects of the defect record pairs; t isdif1And Tdif2Respectively recording the difference of defect occurrence time and defect elimination time in the defect record pair, and the unit is minute; v3The defect time span separation of pairs is recorded for defects.
5. The method of claim 1, wherein the determining the defect record uniqueness checking interface using a support vector machine according to the data points consisting of the defect attribute difference, the defect time difference and the defect time span separation of each sample defect record pair comprises:
constructing an index space with the defect attribute difference, the defect time difference and the defect time span separation as coordinates;
determining data points consisting of defect attribute difference, defect time difference and defect time span separation of each sample defect record pair in the index space, clustering according to the prior result of whether each sample defect record pair is a repeated defect record, and determining a repeated defect record cluster and a non-repeated defect record cluster;
and training by adopting a support vector machine according to the repeated defective cluster and the non-repeated defective cluster to determine a defect record uniqueness checking interface.
6. A system for uniqueness checking of relay protection defect data, the system comprising:
the device comprises a to-be-verified defect record pair determining unit, a to-be-verified defect record pair determining unit and a to-be-verified defect record pair determining unit, wherein the to-be-verified defect record pair determining unit is used for acquiring at least two to-be-verified defect records corresponding to the same relay protection device identifier, and selecting two to-be-verified defect records in the at least two to-be-verified defect records as to-be-verified defect;
the defect attribute difference degree determining unit is used for calculating the weight of each defect attribute according to the statistical result of each attribute of the whole relay protection defect record, and calculating the defect attribute difference degree of the defect record pair to be verified according to the defect attribute data in the defect record pair to be verified and the weight of each defect attribute;
a defect time difference determining unit, configured to determine a defect time difference and a defect time span separation of the defect record pair to be verified according to a defect time in the defect record pair to be verified;
the defect record uniqueness checking interface determining unit is used for acquiring a plurality of sample defect record pairs, calculating the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of each sample defect record pair, and determining a defect record uniqueness checking interface by using a support vector machine according to data points formed by the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of each sample defect record pair;
the uniqueness checking result determining unit is used for determining the uniqueness checking result of the defect record pair to be checked according to the relative position of the data point corresponding to the defect attribute difference degree, the defect time difference degree and the defect time span separation degree of the defect record pair to be checked and the defect record uniqueness checking interface; wherein the uniqueness check result comprises: a repetitive defect recording pair and a non-repetitive defect recording pair.
7. The system of claim 6, wherein the system calculates the weight of each defect attribute according to the statistical result of each attribute of the relay protection total defect records by using the following method, comprising:
wherein, ω isiThe weight of the ith defect attribute; pijThe probability of occurrence of the jth value of the ith defect attribute is obtained; k is the number of different values of the ith defect attribute.
8. The system of claim 6, wherein the system calculates the defect attribute variance of the defect record pair based on the weight of each defect attribute using the following formula, comprising:
wherein, V1A defect attribute difference degree for the defect record pair; omegaiThe weight of the ith defect attribute; i isiAnd the data indicating whether the defect record has difference to the i-th defect attribute is 1 if the defect record has difference and 0 if the defect record has no difference.
9. The system of claim 6, wherein the system determines the defect time disparity and defect time span separation for a defect record pair from defect times in the defect record pair by:
V2＝(V21+V22)/2，
wherein, V2A defect time difference degree for the defect record pair; v21The degree of difference in defect occurrence time for the defect record pair; v22Eliminating time difference degrees for the defects of the defect record pairs; t isdif1And Tdif2Recording defects in pairs for defects respectivelyThe difference between the occurrence time of the sink and the defect elimination time in minutes; v3The defect time span separation of pairs is recorded for defects.
10. The system of claim 6, wherein the defect record uniqueness checking interface determining unit determines the defect record uniqueness checking interface by using a support vector machine according to data points consisting of the defect attribute difference, the defect time difference and the defect time span separation of each sample defect record pair, and comprises:
constructing an index space with the defect attribute difference, the defect time difference and the defect time span separation as coordinates;
determining data points consisting of defect attribute difference, defect time difference and defect time span separation of each sample defect record pair in the index space, clustering according to the prior result of whether each sample defect record pair is a repeated defect record, and determining a repeated defect record cluster and a non-repeated defect record cluster;
and training by adopting a support vector machine according to the repeated defective cluster and the non-repeated defective cluster to determine a defect record uniqueness checking interface.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011478502.7A CN112732773B (en) | 2020-12-15 | 2020-12-15 | Method and system for checking uniqueness of relay protection defect data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011478502.7A CN112732773B (en) | 2020-12-15 | 2020-12-15 | Method and system for checking uniqueness of relay protection defect data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112732773A true CN112732773A (en) | 2021-04-30 |
| CN112732773B CN112732773B (en) | 2024-04-16 |
Family
ID=75602327
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011478502.7A Active CN112732773B (en) | 2020-12-15 | 2020-12-15 | Method and system for checking uniqueness of relay protection defect data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112732773B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113377759A (en) * | 2021-07-01 | 2021-09-10 | 贵州电网有限责任公司 | Defect filling data management method based on expert system algorithm |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110286657A1 (en) * | 2010-05-20 | 2011-11-24 | Yasuyuki Yamada | Defect analysis method, defect analysis apparatus, and recording medium |
| CN102930365A (en) * | 2012-09-24 | 2013-02-13 | 中国电力科学研究院 | Method for checking statistical analysis and operation management data of relay protection |
| CN102928720A (en) * | 2012-11-07 | 2013-02-13 | 广东电网公司 | Defect rate detecting method of oil immersed type main transformer |
| CN105427029A (en) * | 2015-11-05 | 2016-03-23 | 广州兴森快捷电路科技有限公司 | Verification system and method for defect detection |
| CN111222582A (en) * | 2020-01-13 | 2020-06-02 | 中国电力科学研究院有限公司 | Method and system for extracting defect mode of relay protection device |
-
2020
- 2020-12-15 CN CN202011478502.7A patent/CN112732773B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110286657A1 (en) * | 2010-05-20 | 2011-11-24 | Yasuyuki Yamada | Defect analysis method, defect analysis apparatus, and recording medium |
| CN102930365A (en) * | 2012-09-24 | 2013-02-13 | 中国电力科学研究院 | Method for checking statistical analysis and operation management data of relay protection |
| CN102928720A (en) * | 2012-11-07 | 2013-02-13 | 广东电网公司 | Defect rate detecting method of oil immersed type main transformer |
| CN105427029A (en) * | 2015-11-05 | 2016-03-23 | 广州兴森快捷电路科技有限公司 | Verification system and method for defect detection |
| CN111222582A (en) * | 2020-01-13 | 2020-06-02 | 中国电力科学研究院有限公司 | Method and system for extracting defect mode of relay protection device |
Non-Patent Citations (2)
| Title |
|---|
| GUO PENG: "Relay protection data integrity check method based on big data association algorithm", 2019 INTERNATIONAL CONFERENCE ON ELECTRONIC ENGINEERING AND INFORMATICS, 31 December 2019 (2019-12-31), pages 506 - 508 * |
| 杨国生 等: "基于模糊综合评估的 继电保护装置运行状态感知方法", 电力信息与通信技术, vol. 18, no. 4, 15 April 2020 (2020-04-15), pages 74 - 83 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113377759A (en) * | 2021-07-01 | 2021-09-10 | 贵州电网有限责任公司 | Defect filling data management method based on expert system algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112732773B (en) | 2024-04-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110634080B (en) | Abnormal electricity utilization detection method, device, equipment and computer readable storage medium | |
| CN104756106B (en) | Data source in characterize data storage system | |
| CN112508105B (en) | Fault detection and retrieval method for oil extraction machine | |
| CN113344134B (en) | Low-voltage distribution monitoring terminal data acquisition abnormality detection method and system | |
| CN110008254B (en) | Transformer equipment standing book checking processing method | |
| CN111340063A (en) | Coal mill data anomaly detection method | |
| CN113189418B (en) | Topological relation identification method based on voltage data | |
| CN102930495B (en) | Steganography evaluation based steganalysis method | |
| CN111177208A (en) | Power consumption abnormity detection method based on big data analysis | |
| CN109063885A (en) | A kind of substation's exception metric data prediction technique | |
| CN111709668A (en) | Power grid equipment parameter risk identification method and device based on data mining technology | |
| CN116986246A (en) | Intelligent inspection system and method for coal conveying belt | |
| CN114676749A (en) | Power distribution network operation data abnormity judgment method based on data mining | |
| CN113657747B (en) | Intelligent assessment system for enterprise safety production standardization level | |
| CN112732773B (en) | Method and system for checking uniqueness of relay protection defect data | |
| CN117057666B (en) | Distribution quality evaluation method and system for digital distribution network | |
| CN117851184A (en) | Alarm grading system and method based on machine learning | |
| CN113393169B (en) | Financial industry transaction system performance index analysis method based on big data technology | |
| CN110188799A (en) | A kind of continuous data multidimensional analysis and intelligent O&M method | |
| CN115358431A (en) | Accurate study and judgment method for power failure range of transformer area based on low-voltage line archive data | |
| CN114897345A (en) | Method and device for automatically generating index scores based on employee data | |
| CN112149969B (en) | Extra-high voltage direct current control protection merging unit operation state evaluation method and system | |
| CN109871998A (en) | A kind of Line Loss of Distribution Network System rate prediction technique and device based on expert sample bank | |
| CN112488572B (en) | Audit object recommendation method, device, equipment and medium | |
| CN108664590A (en) | A kind of matrimony vine data identification method |
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 |