CN108108909A - Data analysing method towards electric power accident, misoperation fault with operating accident against regulations - Google Patents
Data analysing method towards electric power accident, misoperation fault with operating accident against regulations Download PDFInfo
- Publication number
- CN108108909A CN108108909A CN201810011503.7A CN201810011503A CN108108909A CN 108108909 A CN108108909 A CN 108108909A CN 201810011503 A CN201810011503 A CN 201810011503A CN 108108909 A CN108108909 A CN 108108909A
- Authority
- CN
- China
- Prior art keywords
- accident
- data
- electric power
- factor
- operating
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Educational Administration (AREA)
- Tourism & Hospitality (AREA)
- Data Mining & Analysis (AREA)
- General Business, Economics & Management (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Primary Health Care (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Mathematical method of the method provided by the invention based on fuzzy set concept calculates the fuzzy probability that accident occurs, help to analyze a variety of uncertain people because of influence of the factor to the rule and feature of electric power accident, misoperation fault with operating accident against regulations, so as to more fully utilize data.Simultaneously, in view of there are a large amount of linearly inseparable data, therefore using core k means clustering methods in the data of electric power accident, misoperation fault with operating accident against regulations, the sample data in former space is mapped to a higher dimensional space first, that is nuclear space so that sample data becomes linear separability（Or approximately linear can divide）, then classify in nuclear space.The result of classification can help user to determine the excessive risk behavior for causing electric power accident, misoperation fault with operating accident against regulations, contribute to the weak link of deduction accident event, targetedly precautionary measures and pre-control strategy are formulated, so as to reduce or avoid to repeat the probability of happening of electric power accident, misoperation fault with operating accident against regulations.
Description
Technical field
The present invention relates to electric power safety management and control field, more particularly, to one kind towards electric power accident, misoperation fault with
The data analysing method of operation against rules accident.
Background technology
In recent years, as power grid scale is increasingly huge, structure is increasingly sophisticated, and society is to the stability of power supply and quality
It is required that higher and higher, also more and more lower to the tolerance of power-off event, these all propose the safe and stable operation of electric system
Severe challenge.It is continuously improved with the reliability of power equipment, Unsafe behavior or human-equation error have become electric power
The principal risk source of security incident event.How further to strengthen the management of operating personnel's information data and job safety management and control, carry
The security management and control of high electric operating is horizontal, and it is current electricity to reduce the generation of power generation security incident or repeatability as far as possible
The hot issue of power security management and control work.
Currently for electric power accident event it is mostly close with the achievement in research of big data violating the regulations, mostly using qualitative analysis as
Main, operability is not strong enough, not yet sets up the theoretical system of electric power accident event human factors analysis assessment, electric grid operating people because
The quantitative analysis evaluation studies of reliability are also still rare.Such as：Using analytic hierarchy process (AHP), by each target factor according to importance
Size, form relevance presenting levels, establish sequence judgment matrix, and as the foundation of decision-making, there are uniformity and computationally intensive etc.
Problem.Using the analytic hierarchy process (AHP) based on ambiguity function, the above problem can effectively solve the problem that, but be difficult to disclose electric power accident event
Rule and feature.Using the analytic hierarchy process (AHP) based on triangle ambiguity function, improve previous merely using analytic hierarchy process (AHP)
Deficiency, helps somewhat to disclose the genesis mechanism of electric power accident event, but lacks basic data.
Relevant existing patent has：Patent one, electric operating personnel safety quality assessment method and system, by counting electricity
Power operating personnel completes the situation of evaluation and test examination question, analyzes its security capabilities quality.Shortcoming is：One, particular power thing can not be directed to
Story part and act of violating regulations analysis specialized capability, operation experience et al. are because of the influence situation of factor；Two, create effective exam pool not
Disconnected update and maintenance take time and effort, and operation is more difficult.Patent two, a kind of evaluation method for fault probability of power equipment, it is proposed that one
The electrical equipment fault probability calculation formula that kind is multiplied based on factor, realizes quantitative analysis.Shortcoming is：One, only analyze electric power
Probability of equipment failure can not analyze other electric power accident events and act of violating regulations；Two, the species and value range of factor are limited,
And not individually to people because factor is handled.
More than reason is based on, the present invention proposes a kind of method processing human factor data based on FUZZY PROBABILITY ANALYSIS.It should
Method is considered：One, people has weight to electric power accident event because of factor such as personnel specialty quality, operation experience, experience etc. with violating the regulations
It influences；Two, people has very big randomness and uncertainty because of factor.Therefore, using the mathematics side based on fuzzy set concept
Method calculates the fuzzy probability that event occurs, and helps to analyze a variety of uncertain people because factor is to electric power accident event and maloperation
Rule and feature influence, so as to more fully utilize data.
At the same time, electric power accident event with it is violating the regulations have involve a wide range of knowledge, substantial amounts, content is various, data characteristics is multiple
The features such as miscellaneous, using systematic mathematical method, it is necessary to be carried out for Journal of Sex Research；And the characteristics of for big data, traditional, base
It is hard to work in the method that serial process and structural data are handled.It not yet finds to analyze electric power thing using big data analysis method
Story part has with patent violating the regulations, approximate patent：One kind is towards the visual data digging method of electric power big data.The patent
The cluster and classification analysis of power marketing data set are directed to, shortcoming is：One, electric power accident event and number violating the regulations can not be handled
Present in, the data of a large amount of Nonlinear separabilities；Two, not to people because factor is fully handled.
More than reason is based on, on the basis of electric power accident event and data characteristics violating the regulations and stochastic behaviour is had studied,
The present invention proposes a kind of electric power accident event based on core k- mean clusters and big data analysis method violating the regulations.This method is considered
There are a large amount of linearly inseparable data in electric power accident event and violation data, therefore core k- means clustering methods are used, first
The sample data in former space is mapped to a higher dimensional space, i.e. nuclear space so that it is (or approximate that sample data becomes linear separability
Linear separability), then classify in nuclear space.The result of classification can help user determine to cause electric power accident event with
Excessive risk behavior violating the regulations helps to deduce the weak link of accident event, formulates targetedly precautionary measures and pre-control strategy,
So as to reduce or avoid to repeat accident event and probability of happening violating the regulations.
The content of the invention
Human factor it is an object of the invention to solve the prior art is not included to the technological deficiency of crash analysis, is provided
A kind of data analysing method towards electric power accident, misoperation fault with operating accident against regulations.
For realization more than goal of the invention, the technical solution adopted is that：
Data analysing method towards electric power accident, misoperation fault with operating accident against regulations, comprises the following steps：
S1. the historical data of various electric power accidents, misoperation fault with operating accident against regulations is collected, and to the history of collection
Data are pre-processed；
S2. electric power accident, misoperation fault are set or operates accident against regulations as a fuzzy set, corresponding degree of membership is respectively
μ={ μ1,μ2,μ3,…,μn, it is M={ M to cause the set of factors that accident occurs1,M2,M3,…,Mn, then it judges the accident and obscures generally
The step of rate is：
S21. the comparator matrix of each factor is established by comparing two-by-two：D={ dij, wherein i and j represent compared in the of two because
Element when i and j are of equal importance, takes dij=1；When i is more important than j, d is takenij=2；Conversely, take dij=0
S22. the importance ranking index of each factor is calculatedMaximum and minimum value therein are remembered respectively
For γmaxAnd γmin, and establish the judgment matrix B={ b of set of factorsij}：
S23. it will determine that matrix B carries out Regularization after being added by row, obtain weight sets W, the W={ w of each factor1,w2,
w3,…,wk}；
S24. several experts are engaged, using online evaluation, are carried out for the correlation between each subitem of accident and each factor
Scoring, establishes every expert one rating matrix C={ clk, wherein l represents the subitem of accident, and k represents factor, when accident
When subitem l is related to factor k, clk=1；When the subitem l of accident is uncorrelated to factor k, clk=0；By the scoring of each expert
Regularization is carried out by row, you can obtain rating matrix E={ e after additionlk}；
S24. rating matrix E with the weight sets W of factor is multiplied, the comprehensive grading value of accident can be obtained：pl=Ei× w, (i=
1,2 ..., l), then the Comprehensive Evaluation collection of the accident is：P={ p1,p2,…,pl, the fuzzy probability of the accident is as follows：
S3. using the fuzzy probability of the historical data of the accident of step S1 and the accident being calculated as input, core is used
K- means clustering algorithms are clustered；
S4. based on cluster result, analysis various factors is to electric power accident, misoperation fault or the shadow for operating accident against regulations
It rings.
Compared with technology, the beneficial effects of the invention are as follows：
Mathematical method of the method provided by the invention based on fuzzy set concept calculates the fuzzy probability that accident occurs, and contributes to
A variety of uncertain people are analyzed because factor is to electric power accident, misoperation fault and the rule of operation against rules accident and the shadow of feature
It rings, so as to more fully utilize data.Simultaneously, it is contemplated that the data of electric power accident, misoperation fault with operating accident against regulations
In there are a large amount of linearly inseparable data, therefore use core k- means clustering methods, first map the sample data in former space
To a higher dimensional space, i.e. nuclear space so that sample data becomes linear separability (or approximately linear can divide), then in nuclear space
In classify.The result of classification can help user to determine to cause electric power accident, misoperation fault with operating accident against regulations
Excessive risk behavior helps to deduce the weak link of accident event, targetedly precautionary measures and pre-control strategy is formulated, so as to drop
Probability of happening that is low or avoiding repeating electric power accident, misoperation fault and operation against rules accident.
Description of the drawings
Fig. 1 is the flow chart of data analysing method.
Fig. 2 is the flow chart of data prediction.
Fig. 3 is the flow chart of FUZZY PROBABILITY ANALYSIS.
Fig. 4 is the flow chart that core k- means clustering algorithms are clustered.
Specific embodiment
Attached drawing is only for illustration, it is impossible to be interpreted as the limitation to this patent；
Below in conjunction with drawings and examples, the present invention is further elaborated.
Embodiment 1
As shown in Figure 1, method provided by the invention includes following steps：
1) data collection
Collect various electric power accident events and the historical data of operation against rules in nearly 3 years or 5 years nearly, large enterprise can be with
Acquisition time is appropriately extended, with abundant data.The object of collection includes but not limited to：Once cause more than 30 people dead, 10 people
More than below 30 people more than dead, 3 people below 10 people below dead, 3 people more than dead and 100 people more than severely injured, 50 people 100
More than severely injured, 10 people especially great, great, larger, general electric power person of severe injury etc. below severely injured, 10 people below 50 people below people
Accident；Once cause 100,000,000 yuan or more, less than 50,000,000 yuan or more 100,000,000 yuan, less than 10,000,000 yuan or more 50,000,000 yuan, 1,000,000 yuan
More than 10,000,000 yuan of especially great, great, larger, general misoperation faults of economic loss etc. directly below.The content of collection
Including the data of personal injury event, maloperation and violation event in itself.Such as number of injured people, death toll, economic loss volume
The place or work that degree, accident voltage class, the power generation working link at place or power construction working link, accident occur
Region, maloperation type, incident classification etc..Power generation working link includes：Transmission of electricity, power transformation, distribution, power supply, electricity consumption, tune
Degree, operation, tour, maintenance, maintenance, experiment, meter reading examination, power utility check, repairing, technological transformation, industry expand；Power construction working link
Including：The construction of power transmission and transformation distribution engineering, installation, debugging and management；Accident occurs place and includes but not limited to：Repair and spare parts workshop, change
The maloperations such as power station, power plant, power distribution room, control room, construction site, production warehouse, garage, circuit and power communication corridor
Type includes but not limited to：On-load draws (conjunction) isolation switch (rules and regulations such as electric power dispatch management regulation, electric operation directive/guide
Except the situation of permission), band cable address (conjunctions) ground wire (grounding switch), line with ground (grounding switch) combination switch (disconnecting link) etc..
Meanwhile to accident event and break rules and regulations to classify according to relevant criterions such as GB6441, determine grade.
The content of collection further includes correlated characteristic element, such as on personnel (such as gender, age, educational background, recruitment form, work
Age, work post, this work post length of service, educational training, safety examination, credentials, occupation taboo etc.), machine (such as species, use
Time, rated voltage, high-tension current etc.), the data of environment (such as temperature, humidity, wind speed) etc..The passerby that applies of the present invention also may be used
With the type of data collected by being determined according to actual demand.
2) data prediction
As shown in Fig. 2, data prediction includes data scrubbing, data integration, data conversion and four steps of hough transformation.
(2.1) data scrubbing
Data scrubbing refers to data imperfect, inconsistent in processing database and noise data, will be complete, correct, consistent
Data message deposit data warehouse in.Since electric power accident event and violation data source are complicated, data volume is huge, characteristic
Amount is more, and specific difference is very big, so when handling missing values, missing values is filled in manually, workload is too big；Using global normal
Amount fills in missing values, unreliable.Therefore, following methods should be used：One, for missing data, if other information can be used
It at (such as can be according to ID card No. completion gender, native place, date of birth and the age) of completion, should be mended using other information
Entirely；Can not completion, consider using the attribute average completion missing values or use most possible full missing values of value complement, such as
The instrument or Decision Tree Inductive based on reasoning formalized using logistic regression, Bayes determines missing values, carries out completion.Two,
For noise data, data smoothing technology should be used, to whole sample data using multicomponent linear regressioning technology, data are fitted
To a hypersurface, so as to detect and remove outlier；Local the progress of branch mailbox technology should be used local smooth, it is workable
Technology has：It is smooth, smooth and smooth with case median with case border with case average.The passerby that applies of the present invention can be as the case may be
Selection technique.
(2.2) data integration
Data integration, which refers to, logically or physically organically concentrates the data of separate sources, form, feature property, from
And provide comprehensive data sharing to the user.Due to electric power accident event with violation data in the presence of a large amount of semi-structured and non-structural
Change data, and data volume is huge, feature quantity is more, is stored and is managed so being difficult with traditional relevant database.
Therefore, the mixing of MPP (Massive Parallel Processing) parallel database clusters and Hadoop clusters should be used,
To realize storage and management to hundred PB magnitudes, EB magnitude data.During specific implementation：One, manage calculating high quality using MPP
Structural data, powerful SQL and OLTP type services are provided；Two, it is realized using Hadoop to semi-structured and unstructured
The processing of data, to support the new applications such as content retrieval, depth excavation and comprehensive analysis.
(2.3) data convert
Data conversion refers to the shape converted the data by modes such as data normalization, Data generalizations suitable for data mining
Therefore formula, should use following methods：One, data normalization refers to data bi-directional scaling, is allowed to fall into a smaller spy
Determine section, particular technique has：Min-max standardization, z-score standardization and decimal calibration standardization；Two, Data generalization,
Refer to using Concept Hierarchies, bottom data or initial data are replaced using high-level concept, it can be according to histogram analysis etc.
Mathematical method generate numerical value Concept Hierarchies, can also by user expert pattern grade, explicitly declared attribute level；Three, belong to
Property construction, refer to the needs according to data mining, construct new attribute and be added in property set.The passerby that applies of the present invention can basis
Concrete condition selection technique.
(2.4) hough transformation
Hough transformation refers to the stipulations expression for obtaining data set.Since the electric power accident event to magnanimity and violation data carry out
Data analysis and data mining should use hough transformation technology, it is necessary to long time, ensure former data integrity
In the case of, efficiently, data analysis and data mining task are completed in high quality.Therefore, following methods should be used：One, data are stood
Cube is assembled：Using data cube storage multidimensional aggregation information, to the data being stored in data cube structure, using poly-
Collection operation；Two, dimension stipulations, using data encoding or conversion, the stipulations or compression expression of the former data of acquisition.The skill that may be employed
Art has：Wavelet transformation and principal component analysis.Three, numerical value stipulations, by selecting replacement, the smaller data expression shape of memory space
Formula reduces data volume.The numerical value reduction techniques for having ginseng can be selected：Multiple regression and multidimensional probability distribution；Numerical value reduction without ginseng
Technology can be selected：Histogram and cluster.The passerby that applies of the present invention can selection technique as the case may be.
3) human factor data FUZZY PROBABILITY ANALYSIS
For each accident, to be directed to personnel specialty quality, operation experience and experience etc. with very big randomness and
Probabilistic factor is, it is necessary to do FUZZY PROBABILITY ANALYSIS, as shown in figure 3, being as follows：
Assuming that certain accident is a fuzzy set, corresponding degree of membership is respectively μ={ μ1,μ2,μ3,…,μn, cause the thing
Therefore the set of factors occurred is M={ M1,M2,M3,…,Mn, then the step of judging the fuzzy probability is：
(3.1) weight of each factor is determined
The comparator matrix of each factor is established by comparing two-by-two first：D={ dij, wherein i and j represent compared in the of two because
Element when i and j are of equal importance, takes dij=1；When i is more important than j, d is takenij=2；Conversely, take dij=0.
Then the importance ranking index of each factor is calculatedRemember respectively it is therein most
Big value and minimum value are γmaxAnd γmin, establish the judgment matrix B={ b of set of factorsij, shown in computational methods such as formula (1).With
Afterwards, after this judgment matrix B is added by row, Regularization is carried out, you can obtain weight sets W, the W={ w of each factor1,w2,
w3,…,wk}。
(3.2) rating matrix is obtained
Several experts are engaged, using online evaluation, are carried out for the correlation between each subitem of fuzzy event and each factor
Scoring, establishes every expert one rating matrix C={ clk, wherein l represents each subitem, and k represents factor, when each subitem l with
During factor k correlations, clk=1；When each subitem l is uncorrelated to factor k, clk=0.By row after the scoring of each expert is added
Carry out Regularization, you can obtain rating matrix E={ elk}。
(3.3) fuzzy probability is obtained
Rating matrix with the weight sets of factor is multiplied, the comprehensive grading value of fuzzy event can be obtained：pl=Ei× w, (i=1,
2 ..., l), then the Comprehensive Evaluation collection of the event is：P={ p1,p2,…,p1, shown in the fuzzy probability such as formula (2) of the event.
4) core k- mean clusters
Using the data Jing Guo second step data prediction and the accident fuzzy probability of the 3rd step as input data, core is used
K- means clustering algorithms are clustered.As shown in figure 4, it is as follows：
(4.1) in order to protrude the feature difference between different classes of sample so that sample becomes linear separability or approximation
Linear separability, first with a Nonlinear MappingBy former space RnIn sample x be mapped to one
In higher-dimension nuclear space F, sample becomes
(4.2) K sample is arbitrarily selected in nuclear space as initial cluster center, i.e.,
(4.3) in nuclear space, by each sampleIt is assigned to according to nearest neighbouring rule in each classification, i.e., using public affairs
Formula (3) calculates the distance of sample and each cluster centre, chooses the class of distance value minimum.Wherein K is kernel function, usually using full
The kernel function of sufficient Mercer conditions, such as formula (4) and formula (5), can also use condition positive definite kernel function (i.e. CPD cores), such as
Formula (6).In order to save computing event, it may be considered that using CPD cores, and make q=1.
Polynomial:K(xi,xj)=(1+xi·xj)p,p∈N (4)
CPD:K(xi,xj)=- | | xi-xj||q+1,0<q≤2 (6)
(4.4) cluster centre is recalculatedWithValue.Since in nuclear space, cluster centre cannot be bright
It really calculates, a sample generation can only be selected to gather for center in every class, specific practice is：In the classification obtained in step (4.3),
Respectively using each sample as cluster centre, calculate class in other each sample points arrive cluster centre distance, calculate apart from the sum of,
Then the minimum cluster centre of the sum of distance is such cluster centre, and the sum of minimum distance is such error sum of squaresIt will be all kinds ofIt is added and obtainsValue.Distance calculation formula is described consistent with (4.3).
(4.5) step (4.3) and step (4.4) are repeated, until continuous n times iterationUntil being worth constant (or varying less).
According to cluster result, analysis various factors is to the rule towards electric power accident, misoperation fault with operating accident against regulations
The influence of rule and feature determines to cause the excessive risk behavior towards electric power accident, misoperation fault with operating accident against regulations, analysis
Pests occurrence rule towards electric power accident, misoperation fault with operating accident against regulations deduces the weak link of accident event, so as to make
Fixed targetedly precautionary measures and pre-control strategy.
Obviously, the above embodiment of the present invention is only intended to clearly illustrate example of the present invention, and is not pair
The restriction of embodiments of the present invention.For those of ordinary skill in the art, may be used also on the basis of the above description
To make other variations or changes in different ways.There is no necessity and possibility to exhaust all the enbodiments.It is all this
All any modification, equivalent and improvement made within the spirit and principle of invention etc., should be included in the claims in the present invention
Protection domain within.
Claims (2)
1. the data analysing method towards electric power accident, misoperation fault with operating accident against regulations, it is characterised in that：Including following
Step：
S1. the historical data of various electric power accidents, misoperation fault with operating accident against regulations is collected, and to the historical data of collection
It is pre-processed；
S2. set electric power accident, misoperation fault or operate against regulations accident as a fuzzy set, corresponding degree of membership be respectively μ=
{μ1, μ2, μ3..., μn, it is M={ M to cause the set of factors that accident occurs1, M2, M3..., Mn, then it judges the accident and obscures generally
The step of rate is：
S21. the comparator matrix of each factor is established by comparing two-by-two：D={ dij, wherein i represents the factor compared in the of two with j,
When i and j are of equal importance, d is takenij=1；When i is more important than j, d is takenij=2；Conversely, take dij=0
S22. the importance ranking index of each factor is calculatedRemember that maximum and minimum value therein are respectively
γmaxAnd γmin, and establish the judgment matrix B={ b of set of factorsij}：
S23. it will determine that matrix B carries out Regularization after being added by row, obtain weight sets W, the W={ w of each factor1, w2,
w3..., wk}；
S24. several experts are engaged, using online evaluation, are scored for each subitem of accident with the correlation between each factor,
A rating matrix C={ c is established to every expertlk, wherein l represents the subitem of accident, and k represents factor, as the subitem l of accident
When related to factor k, clk=1；When the subitem l of accident is uncorrelated to factor k, clk=0；The scoring of each expert is added
Regularization is carried out by row, you can obtain rating matrix E={ e afterwardslk}；
S24. rating matrix E with the weight sets W of factor is multiplied, the comprehensive grading value of accident can be obtained：pl=Ei× w, (i=1,
2 ..., l), then the Comprehensive Evaluation collection of the accident is：P={ p1, p2..., pl, the fuzzy probability of the accident is as follows：
S3. it is equal using core k- using the fuzzy probability of the historical data of the accident of step S1 and the accident being calculated as input
Value clustering algorithm is clustered；
S4. based on cluster result, analysis various factors is to electric power accident, misoperation fault or the influence for operating accident against regulations.
2. the data analysing method according to claim 1 towards electric power accident, misoperation fault with operating accident against regulations,
It is characterized in that：The pretreatment of the step S1 includes four data scrubbing, data integration, data conversion and hough transformation steps
Suddenly：
Data scrubbing refers to data imperfect, inconsistent in processing database and noise data, by complete, correct, consistent number
It is believed that in breath deposit data warehouse；
Data integration, which refers to, logically or physically organically concentrates the data of separate sources, form, feature property, so as to for
User provides comprehensive data sharing；
Data conversion refers to by smoothly assembling, Data generalization, standardization processing convert the data into shape suitable for data mining
Formula；
Hough transformation refers to the stipulations expression for obtaining data set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810011503.7A CN108108909A (en) | 2018-01-05 | 2018-01-05 | Data analysing method towards electric power accident, misoperation fault with operating accident against regulations |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810011503.7A CN108108909A (en) | 2018-01-05 | 2018-01-05 | Data analysing method towards electric power accident, misoperation fault with operating accident against regulations |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108108909A true CN108108909A (en) | 2018-06-01 |
Family
ID=62219766
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810011503.7A Pending CN108108909A (en) | 2018-01-05 | 2018-01-05 | Data analysing method towards electric power accident, misoperation fault with operating accident against regulations |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108108909A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111144751A (en) * | 2019-12-26 | 2020-05-12 | 国家电网有限公司 | Method and system for determining power accident occurrence risk based on behavior analysis |
| CN115936753A (en) * | 2022-12-14 | 2023-04-07 | 浙江中控技术股份有限公司 | Industrial chain analysis method and system |
| CN117686757A (en) * | 2023-12-07 | 2024-03-12 | 浙江万昌电力设备有限公司 | Intelligent early warning method and system for outdoor power metering box |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020069091A1 (en) * | 2000-10-02 | 2002-06-06 | Steven Wahlbin | Computerized method and system of liability assessment for an accident |
| CN102426670A (en) * | 2011-11-17 | 2012-04-25 | 重庆市科学技术研究院 | Technical value evaluation method and evaluation model of technical economic life |
| CN104392393A (en) * | 2014-11-20 | 2015-03-04 | 三峡大学 | DEMATEL-ANP-VIKOR mixed selection method of power system security risk reduction schemes |
| CN105894189A (en) * | 2016-03-31 | 2016-08-24 | 西安石油大学 | Complex system risk inducement analysis method using FTA and GRA-coupled FAHP |
-
2018
- 2018-01-05 CN CN201810011503.7A patent/CN108108909A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020069091A1 (en) * | 2000-10-02 | 2002-06-06 | Steven Wahlbin | Computerized method and system of liability assessment for an accident |
| CN102426670A (en) * | 2011-11-17 | 2012-04-25 | 重庆市科学技术研究院 | Technical value evaluation method and evaluation model of technical economic life |
| CN104392393A (en) * | 2014-11-20 | 2015-03-04 | 三峡大学 | DEMATEL-ANP-VIKOR mixed selection method of power system security risk reduction schemes |
| CN105894189A (en) * | 2016-03-31 | 2016-08-24 | 西安石油大学 | Complex system risk inducement analysis method using FTA and GRA-coupled FAHP |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111144751A (en) * | 2019-12-26 | 2020-05-12 | 国家电网有限公司 | Method and system for determining power accident occurrence risk based on behavior analysis |
| CN115936753A (en) * | 2022-12-14 | 2023-04-07 | 浙江中控技术股份有限公司 | Industrial chain analysis method and system |
| CN117686757A (en) * | 2023-12-07 | 2024-03-12 | 浙江万昌电力设备有限公司 | Intelligent early warning method and system for outdoor power metering box |
| CN117686757B (en) * | 2023-12-07 | 2024-06-11 | 浙江万昌电力设备有限公司 | Intelligent early warning method and system for outdoor power metering box |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113435759B (en) | Primary equipment risk intelligent assessment method based on deep learning | |
| CN108491969A (en) | Spatial Load Forecasting model building method based on big data | |
| CN106980905A (en) | Power distribution network power supply reliability prediction method and system | |
| CN110807550A (en) | Distribution transformer overload identification early warning method based on neural network and terminal equipment | |
| Zhang et al. | Critical node identification in active distribution network using resilience and risk theory | |
| CN108964023A (en) | A kind of busbar voltage situation short term prediction method and system for power grid | |
| CN117726161A (en) | Risk assessment method, system, terminal and medium for power grid power transmission equipment | |
| Hou et al. | Spatial distribution assessment of power outage under typhoon disasters | |
| CN112200458A (en) | Power distribution network planning data application method and system | |
| CN110348683A (en) | The main genetic analysis method, apparatus equipment of electrical energy power quality disturbance event and storage medium | |
| Ying et al. | Research on state evaluation and risk assessment for relay protection system based on machine learning algorithm | |
| CN112308425A (en) | Method for constructing distribution transformer health evaluation index system | |
| CN108108909A (en) | Data analysing method towards electric power accident, misoperation fault with operating accident against regulations | |
| CN103914482B (en) | Centralized Monitoring event influence property based on CMDB determines method | |
| CN104037760B (en) | Anticipated accident selection method for electric power system with uncertain injection power | |
| CN114662589B (en) | A method, device, equipment and readable storage medium for analyzing and judging electric meter faults | |
| CN113052250A (en) | Decision support method, system, device and medium based on meteorological disaster | |
| Li et al. | A probabilistic data-driven method for response-based load shedding against fault-induced delayed voltage recovery in power systems | |
| Ma et al. | A probabilistic cascading failure model for dynamic operating conditions | |
| Wang et al. | Deep learning-based short-term load forecasting for transformers in distribution grid | |
| Hou et al. | Research on real power grid security and stability based on big data analysis method and online data of power grid | |
| CN116318894A (en) | Power node network attack threat degree calculation method and terminal equipment | |
| CN115146715A (en) | Method, device, equipment and storage medium for diagnosing potential safety hazards in electricity use | |
| Lu et al. | Anomaly recognition method for massive data of power internet of things based on bayesian belief network | |
| Dongli et al. | Research on fault risk ranking and screening of distribution network based on uncertainty theory |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180601 |