CN106483947A - Distribution Running State assessment based on big data and method for early warning - Google Patents
Distribution Running State assessment based on big data and method for early warning Download PDFInfo
- Publication number
- CN106483947A CN106483947A CN201610837927.XA CN201610837927A CN106483947A CN 106483947 A CN106483947 A CN 106483947A CN 201610837927 A CN201610837927 A CN 201610837927A CN 106483947 A CN106483947 A CN 106483947A
- Authority
- CN
- China
- Prior art keywords
- data
- running state
- power distribution
- index
- module
- 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
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0224—Process history based detection method, e.g. whereby history implies the availability of large amounts of data
- G05B23/0227—Qualitative history assessment, whereby the type of data acted upon, e.g. waveforms, images or patterns, is not relevant, e.g. rule based assessment; if-then decisions
- G05B23/0229—Qualitative history assessment, whereby the type of data acted upon, e.g. waveforms, images or patterns, is not relevant, e.g. rule based assessment; if-then decisions knowledge based, e.g. expert systems; genetic algorithms
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0243—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model
- G05B23/0245—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model based on a qualitative model, e.g. rule based; if-then decisions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of distribution Running State assessment based on big data and method for early warning are mainly monitored to current power distribution Running State, ensure assessment and the early warning mechanism of power distribution network safe operation, it is mainly made up of the part such as data processing module, running status evaluation module, Risk-warning module and operational control module.Mainly first pass through the comprehensively given index parameter of PCA, then the functional relation between comprehensive score and each achievement data is excavated using the algorithm idea of gene expression programming, finally using current power distribution network running state data as this functional relation input, current power distribution Running State risk class is evaluated according to its result of calculation, and give feasible referential suggestion, user and decision-maker is helped preferably to plan and manage whole power distribution network so that power distribution network can more efficient more safely run.
Description
Technical field
The present invention relates to a kind of distribution Running State assessment and method for early warning, it is mainly used in solving distribution Running State
Assessment and the problem of early warning, belong to security information for power system field.
Background technology
Power distribution network is the important component part of power system, the carrying of the development with social economy and people's living standard
Rise, people put forward higher requirement to the construction of power distribution network and management.But in traditional distribution Running State is assessed,
It is limited to narrower data acquisition channel or relatively low data integration and disposal ability so that research worker is difficult to from wherein excavating
Go out more valuable information.Now due to the popularization and application of the application systems such as power distribution automation, power information collection, for having thousand
For the large-scale distribution network of bar feeder line, the magnanimity isomery of exponential growth, polymorphic data can be produced in power distribution network, therefore such as
What to analyze and process these data with the big data technology that reaches its maturity, and effectively excavate relation between horizontal data and
Prediction data is moved towards, and the safe operation for power distribution network and good service provide support to become our problem demanding prompt solutions.For
For Utilities Electric Co., how using the data basis of power distribution network accumulation, the running status of power distribution network is estimated and early warning more
It is an important job.
Distribution Running State is estimated and early warning mainly uses data analysiss data digging technology, in conjunction with going through
History and as-is data, by calculation risk index, judge institute's risk type, in prediction following a period of time from now on
The risk situation that power distribution network is faced, then according to risk classifications identification result, generates corresponding prevention and control scheme.At present,
In terms of power distribution network assessment, the integrated evaluating method of application has a lot, such as:Analytic hierarchy process (AHP), Field Using Fuzzy Comprehensive Assessment, Information Entropy
And PCA etc..These methods are all the applications of classical theory, have higher practicality.But, its anthropic factor
The keynote message that projects of the rather serious and different evaluation theory of impact different, these all may make evaluation conclusion drift
Move, make evaluation result that error to occur.Therefore study effective distribution Running State assessment and method for early warning can preferably carry
High power distribution network operation and management level, has great importance.
The assessment of distribution Running State and early warning mainly need to consider the problem of two aspects:(1) how to set assessment distribution
The index parameter of Running State;(2) a kind of method how is selected effectively to assess current power distribution Running State and to be given
Early warning.Set suitable evaluation index parameter for according to actual demand, have factors to need to consider, such as:Set comments
Estimate the running status whether index can comprehensively reflect whole power distribution network, if subjective and objective factor can be taken into account and respectively have
Emphasis, and whether can reduce index parameter number etc. on the premise of satisfaction refers to requirement as far as possible.And how according to setting
Index parameter select a kind of effective method assessing the key technology that the running status of power distribution network is also whole evaluation system.
Content of the invention
It is an object of the invention to provide a kind of distribution Running State assessment based on big data and method for early warning,
The risk being likely to occur under distribution Running State is estimated and early warning.Present mechanism is a kind of tactic method, by making
Effectively distribution Running State can be estimated and early warning with this method, for ensureing that power distribution network being capable of safe operation offer
A kind of technological means.
The technical solution of the present invention is:
A kind of distribution Running State assessment based on big data and method for early warning, is characterized in that:By data processing module
Carry out data processing, running status assessment is carried out to the data processing through data processing module by running status evaluation module, fortune
The comprehensive score that row state estimation obtains submits Risk-warning resume module to, and Risk-warning module evaluates current power distribution network operation shape
The risk class of state, provides feasible behavior command to operational control module simultaneously;
Comprise the following steps:
Step one:The actual demand of problem analysis, builds rational distribution Running State evaluation index system;
Step 2:Under collection evaluation system, the relevant historical data corresponding to two-level index, is processed to it:Umber of defectives
According to detection, the identification of bad data and repairing and missing values filling, form the training sample data that need to process further
Collection X;
Step 3:This training sample data collection X is standardized process, forms training sample data collection to be excavated,
Assume still to be represented with X;
Step 4:With PCA, synthesis is carried out to all kinds of index parameter values of collection, index number is subtracted from p
As little as m, obtains New Set zi(i=1,2 ..., calculation expression m), thus generate new training sample data collection Z;
Step 5:Gene expression programming is used to initialize population according to the sample data set Z obtaining;
Step 6:Build suitable fitness function, set and stop producing newly for the calculating reaching needed for stylish generation
Accuracy rating;
Step 7:Carry out selection operation, mutation operation, map function and reorganization operation for the individuality producing, produce
New individual;
Step 8:The new individual producing is sorted according to adaptive value size, retains excellent individual;
Step 9:Select multiformity operator pb、pr, p in certain proportionbRetain the poor individuality of fitness, with certain
Probability prRandomly generate a part of new individual, then produce new generation;
Step 10:Judge the new computational accuracy whether reaching setting for adaptive value producing, if reaching, exiting, going to step
Rapid 11, otherwise go to step 7 and repeat;
Step 11:Return the functional relation between comprehensive score and each achievement data, i.e. F=f (z1,z2,…,zm);
Step 12:The related index parameter sample number of obtained power distribution network current operating conditions after gathering and process
According to the independent variable as this functional relation, calculate comprehensive score F of current power distribution Running State;
Step 13:Inquiry risk class knowledge base, the comprehensive risk class waiting corresponding to point F of evaluation, and be given
Accordingly feasible behavior command;
Step 14:If power distribution network is in automatic mode, it is immediately performed command adapted thereto;Otherwise, user is transferred to voluntarily to process;
Step 15:The assessment of distribution Running State and prealarming process terminate.
Architecture
Fig. 1 gives a kind of assessment of the distribution Running State based on big data and early warning structure chart, and it mainly includes four
Individual module:Data processing module, running status evaluation module, Risk-warning module and operational control module.The operation of in figure
Control module includes in the case of giving respective risk advanced warning grade to distribution Running State, to distribution Running State
All concrete operations of corresponding decision are implemented in scheduling;Also include the artificial active operation of possible some simultaneously.The present invention increases
Added other parts have no effect on distribution Running State is estimated and early warning action, simply ensure power distribution network can
It is in safer environment to run.
Concrete introduction is given below:
Data processing module:When with power distribution network resource big data a large amount of for object acquisition, it is no lack of in the data being gathered
The presence of some bad data, directly bring carry out calculating will certainly to distribution Running State assessment result produce unfavorable
Impact.Accordingly, it would be desirable to process to bad data, corresponding completion is made to the data value of some disappearances, thus improving simultaneously
In the quality of state estimation when institute input data, improve the accuracy of the last each index comprehensive score affecting running status.
In this patent any restriction is not done to implementing of data processing module.
Running status evaluation module:For improving the safety of distribution network operation, build the power distribution network fortune of a reasonable
Row state estimation model is extremely important.In power distribution network running, need to select or one or more assessment algorithm of design is made
For technological means, to build assessment models in conjunction with the historical data after being gathered and processing;Then power distribution network is currently monitored
Data as the input of assessment models, obtain corresponding assessment result, be scheduling and the management of later stage distribution Running State
Provide theoretical foundation.In patent of the present invention mainly in conjunction with PCA and based on gene expression programming come structure
Build distribution Running State assessment models.
Risk-warning module:In addition it is also necessary to according to assessment result after the state that power distribution network is currently run is estimated
Set corresponding advanced warning grade and give to feed back;Meanwhile, in order to be preferably power specialty personnel's implementation decision and planning construction
Basis for estimation is provided, Risk-warning module can according to the demand of user selective by some Monitoring Data or evaluation status with
Different visualization techniques is shown.In addition, when assessment result display distribution Running State abnormal, Risk-warning
Module also needs to trial and provides the user solution, and has the authority automated toing respond to command adapted thereto under high advanced warning grade.
Operational control module:Operational control module mainly runs operational order, is the scheduling of distribution Running State and pipe
The operational control center of reason.It includes both of which:Automatic mode and artificial mode.If this operational control module is in automatic mold
Under formula, it needs according to configured good instruction or program management and controls whole power distribution network, such as occurs in Risk-warning module
During the high advanced warning grade setting, operational control module needs to be immediately performed given instruction.If at this operational control module
In a manual mode, it mainly executes instruction or the program of user input according to the demand of user.Generally, operation control
Molding block is in automatic mode, but the authority of artificial mode is higher than automatic mode.
Method flow
1st, data processing module
The historical data being gathered is carried out processing and is a need for, it improves structure distribution Running State assessment mould
The accuracy of data set during type.In the method, data is carried out with process and be broadly divided into 3 steps:1) detection of bad data.
Using the bad data detection of comparative maturity, and with reference to the correlation theory of power system, the historical data being gathered is carried out
Detection, judges to whether there is bad data in this measurement sampling.As it is contemplated that using 6 in standardized residual and power system
Rule and 4 rules are detected and are judged.2) identification of bad data and repairing.When the gathered historical data of discovery exists
After bad data, need to further determine that metric data is bad data to which (or which), and to the greatest extent may be used according to suitable method
Bad data can be adjusted or assignment again.Bad data is recognized using decision-making tree theory as being contemplated that, and base
In the power distribution network rail voltage general fluctuation range knowledge such as less, bad data is repaired.3) filling of missing values.Institute
In the historical data of collection, due to misoperation and omission, or it is that information the reason such as cannot obtain and can lead to shortage of data,
And the presence of null value can impact to whole data mining process, so we need by special method to disappearance
Data carries out deriving, fills, to reduce the gap between data mining algorithm and practical application.Can for missing data not
Same type or attribute use different processing empty value methods, such as it is contemplated that all kinds of methods such as particular value filling and homing method are comprehensive
Mode realizing.
Further, since the data of the different electric power index of collection can directly affect distribution Running State assessment models and set up
Correctness, should be as far as possible it is therefore desirable to set up an appropriate electrical network assessment indicator system of comparison according to final demand
Comprehensively reflect electrical network practical situation, consider the practical situation such as data acquisition difficulty, amount of calculation again, accomplish as far as possible neither
Repeat also not omit.In patent of the present invention, because safe and reliable and economy are the basic demands that electrical network is powered, meanwhile, right
Also need when distribution Running State is estimated to consider the power supply capacity of electrical network, quantitative analysiss carried out to possible risk indicator,
Therefore, to sum up consider, and combine electric power expert many-side suggestion, patent of the present invention determines distribution network operation shape as shown in Figure 2
State evaluation index system structure.
This evaluation index system includes safety evaluatio, power supply capacity evaluation, reliability and power supply quality evaluation, economy
Evaluate and fault identification and 5 first class index of risk indicator.Each index comprises multinomial subordinate's index, i.e. two-level index, with from
Different angles is quantified.This 5 first class index constitute an entirety, can the current running status of effective evaluation power distribution network.
It is specifically intended that patent of the present invention is two-level index under this evaluation system in the historical data that data processing module is gathered
Quantized data collection.
Running status evaluation module
Power distribution network operationally, need cycle or persistence to whole operation of power networks state implementing monitoring.By selecting
Suitable data analysiss and mining algorithm, and combine the historical data of power distribution network, to build fairly perfect assessment models;Then
The current service data of collection power distribution network, calculates its risk indicator using this assessment models and gives advanced warning grade, judge distribution
The risk classifications that net may face, the whole service situation of power distribution network in prediction following a period of time from now on.Due to multiple
The leading indicator of data meeting dispersive effects assessment result, reduces the correctness of assessment result, and increases difficulty in computation;In addition, it is existing
Some having primarily focus on expertise with regard to power distribution network synthesis state evaluating method, lack objectivity, are difficult to entirely joining
Operation of power networks state is estimated.
Therefore, in the method, the structure of distribution Running State assessment models is broadly divided into two aspects:First with
PCA carries out dimensionality reduction to multiple indexes parameter, comprehensive effectively index parameter, is then based on gene expression programming
Algorithm realizes assessment and the early warning of distribution Running State.
Using PCA, the historical data obtained by data processing module is carried out dimensionality reduction, comment in comprehensive each index
Simplify the input of index amount on the basis of valency.Assume that the sample observation data matrix through data processing module output is:
Wherein xijWith regard to the numerical value corresponding to index j, the index number selected is p to i-th gathered sample of expression, altogether
Acquire n sample.
Main working process is as follows:
The power distribution network historical data sample collecting is standardized process, forms training sample data to be excavated
Collection.Matrix samples X of (1) formula are standardized processing, relevant treatment formula is as follows:
Wherein
Calculate the correlation matrix R of this training sample:
For convenience it is assumed that the sample after initial data standardization is still represented with X, then the phase of the data after normalized process
Closing coefficient is:
Seek the eigenvalue (λ of correlation matrix R with Jacobian technique1,λ2…λp), that is, solve characteristic equation | λ I-R |=0,
Order arranges by size to make the eigenvalue tried to achieve:
λ1≥λ2≥…≥λp≥0
It is also desirable to obtain respectively corresponding to eigenvalue λiCharacteristic vector ei(i=1,2 ..., p) it is desirable to | | ei||
=1, that is,eijRepresent vectorial eiJ-th component.
Calculate principal component contributor rate and contribution rate of accumulative total.Here contribution rate just refers to that the variance of certain main constituent accounts for all
The proportion of variance, actual namely certain eigenvalue accounts for the proportion that All Eigenvalues add up to, that is,:
Contribution rate is bigger, illustrates that the information of the original variable that this main constituent is comprised is stronger.
Select important main constituent, write out main constituent expression formula.Here the contribution rate of accumulative total mainly according to main constituent Lai
Choose main constituent number, typically take contribution rate of accumulative total to reach 85%~95% eigenvalue λ1,λ2,…,λmCorresponding 1st,
2nd ..., the individual main constituent of m (m≤p).Therefore main constituent load lijComputing formula is as follows:
Assume the historical data that we are gathered index variable be xj, now main constituent ziFor:
Through above-mentioned process and analysis, we instead of p original index variable with m main constituent, p that will be original
The index of individual impact distribution Running State reduces to m, and now our training sample set X is changed into Z through process, that is,:
By PCA, the multiple indexes affecting distribution Running State are carried out after screening and synthesis, in we
In method, we also need to using based on gene expression programming come from the training sample data collection Z obtaining after treatment
In excavate F and zj(j=1,2 ..., m) between functional relationship F=f (z1,z2,…,zm), zjRepresent through PCA
Calculated comprehensive index value.By gathering the related data of power distribution network current operating conditions, at PCA
After reason, using calculate gained numerical value according to corresponding relation as this functional relation input, calculate F value, then calculating
Result as Risk-warning module input, thus being estimated to current distribution Running State and giving to feed back.
Main working process is as follows:
According to training sample set Z (see (2) formula) the initialization population obtaining through PCA.
Build suitable fitness function, set and stop producing newly for the computational accuracy model reaching needed for stylish generation
Enclose.
Carry out selection operation, mutation operation, map function and reorganization operation for the individuality producing, produce new individual.
The new individual producing is sorted according to adaptive value size, retains excellent individual.
Select multiformity operator pb、pr, p in certain proportionbRetain the poor individuality of fitness, with certain Probability pr
Randomly generate a part of new individual, then produce new generation.
Judge the new computational accuracy whether reaching setting for adaptive value producing, if reaching, exiting, going to (7), otherwise
Go to (3) to repeat.
Return tried to achieve functional relation, power distribution network current operating conditions phase obtained after then gathering and process
The index parameter sample data closed, as the independent variable of this functional relation, is calculated the synthesis of current power distribution Running State
F is finally committed to Risk-warning module by score F, is easy to the evaluation of follow-up risk class.
Risk-warning module
The comprehensive grading that Risk-warning module mainly can be submitted to by running status evaluation module evaluates corresponding wind
Dangerous grade, and provide the referential suggestion of some reasonables, it is scheduled for decision-maker's reference.In the method, mainly logical
Cross and set up a risk class knowledge base to realize the evaluation function needed for this module.In order to preferably describe this module work(
The design of energy, it is assumed that built risk class knowledge base as shown in table 1.
Table 1 risk class knowledge base
Scoring is interval | [a1,a2) | [a2,a3) | [a3,a4) | [a4,a5) | [a5,a6) |
Rating | A | B | C | D | E |
Behavior command | Action1 | Action2 | Action3 | Action4 | Action5 |
As shown in Table 1, have 5 grades in the risk class knowledge base supposing, and assume risk between each grade
Order of severity relation is:A > B > C > D > E.When running status evaluation module calculates obatained score in the interval [a of scoring1,a2)
When, it is A by inquire about risk class knowledge base being apparent from affiliated grade, feed back to user also has its behavior command simultaneously
Action1.If grade A represents distribution, Running State is in the edge of collapsing, and that Action1 may represent to need to back up immediately and join
Power distribution network operational mode is switched to safe mode by operation of power networks data simultaneously, waits user to be recovered and reconstruct.Work as operation
When control module is in automatic mode, behavior instruction is executed by power distribution network immediately;If operational control module is in artificial mode,
Behavior instruction transfers to user voluntarily to process.
When being designed to risk class knowledge base, need the actual demand according to problem, in conjunction with expert opinion
More rational risk class is set with theoretical knowledge, and the behavior command that corresponding fraction is interval Yu feasible.Separately
Outward, if there being demand, to what user proposed, this Risk-warning module can also check that more detailed assessment result responds, by its front mould
Block collection and the data submitted to carry out Treatment Analysis, are more intuitively shown its result to user with visual technology.Simultaneously
User and decision-maker can also be helped preferably to assess the current operating conditions of this power distribution network, preferably plan and management is entirely joined
The operation of electrical network.
The inventive method proposes a kind of assessment of the distribution Running State based on big data and method for early warning, is mainly used in
Solve the running status of assessment current power distribution network, the method by using proposing in the present invention can be according to current power distribution network operation
Status data effectively assesses this distribution Running State, and feeds back to the accordingly feasible referential suggestion of user;Simultaneously can also be by
According to the demand of user, selectively assessment result visualization technique is more intuitively shown, help user and decision-maker more
Plan well and manage the operation of whole power distribution network.
Running status evaluation module, by using PCA, can screen and overall power distribution Running State effectively
Two-level index parameter under evaluation index system.For the rational distribution Running State assessment models to one comparison of structure,
The multiple influence index factor can disperse the shared weight in assessment models of each index, simultaneously may also can be each other between all kinds of indexs
Impact, thus reduce the accuracy of whole assessment models result of calculation;Or artificially as desired all kinds of indexs are carried out
Screening, can lead to select all kinds of influence index factors to consider entirely, even being to completely eliminate certain class influence index factor commenting
Estimate the effect played in model, and the above-mentioned problem being likely to occur is not for we gladly sees.Therefore, in order to preferably
Dimensionality reduction is carried out to given multiple indexes, the historical data after gathering and processing is analyzed with PCA,
Main constituent between all kinds of indexs can be efficiently differentiated by the interval of contribution rate of accumulative total in given calculating process so that institute
The aggregative indicator extracting both had considered the weight ratio of each Index Influence assessment models, and the index number simultaneously being reduced also drops
The low subsequently intractability to training sample data collection.
In addition, also needing to calculate the overall score of each aggregative indicator on the basis of obtaining aggregative indicator, for subsequently commenting
Fixed corresponding risk class provides basis for estimation.In existing power distribution network evaluation system, it is using analytic hierarchy process (AHP) mostly, obscures
The method based on expertise and expert opinion such as assessment method calculates the comprehensive score of all kinds of indexs, leads to institute's result of calculation more
Mostly it is the result that subjective desire is assessed, whole distribution Running State is lacked with the assurance of more comprehensively more integration.Therefore, at this
Introduce in invention and be based on gene expression programming, the shadow to whole assessment result for the subjective desire can be reduced to a certain extent
Ring.Initialize population by using the training sample data collection that PCA obtains, and build suitable fitness evaluation
Function, the precision that setup algorithm stops, selecting multiformity genetic factor, finally can construct a functional relationship meeting demand
Formula.By this functional relation, user can be using current power distribution network running state data as its independent variable, such that it is able to calculate
Go out the comprehensive score of this power distribution network current operating conditions, for follow-up, corresponding risk etc. is evaluated to this power distribution network current operating conditions
Level provides strong basis for estimation.
Brief description
The invention will be further described with reference to the accompanying drawings and examples.
Fig. 1 is the composition structure chart of the assessment of distribution Running State and early warning.Main inclusion:Data processing module, operation
State estimation module, Risk-warning module and operational control module.
Fig. 2 is distribution Running State evaluation index system structural representation.
Fig. 3 is reference architecture schematic diagram.Represent the assembly that the inventive method includes.
Fig. 4 is the schematic flow sheet of the inventive method.
Specific embodiment
Description is it will be assumed that there is following application example for convenience:
On the basis of having distribution Running State historical data, need current power distribution Running State is estimated
With early warning it is desirable to a series of input as running status evaluation module after process using the historical data collecting, first
With PCA combined influence index parameter, obtain new training sample data collection, then use gene expression to compile
Journey algorithm excavates the functional relation between comprehensive score and each achievement data, finally by current power distribution network running state data
As the input of its functional relation, calculate comprehensive score, result of calculation is committed to Risk-warning module, inquire about risk class
Knowledge base evaluates risk class for current power distribution Running State, and provides feasible referential suggestion.
Its specific embodiment is:
A kind of distribution Running State assessment based on big data and method for early warning, carry out data by data processing module
Reason, carries out running status assessment by running status evaluation module to the data processing through data processing module, and running status is assessed
The comprehensive score obtaining submits Risk-warning resume module to, and Risk-warning module evaluates risk of current power distribution Running State etc.
Level, provides feasible behavior command to operational control module simultaneously;
Comprise the following steps:
Step one:The actual demand of problem analysis, builds rational distribution Running State evaluation index system;
Step 2:Under collection evaluation system, the relevant historical data corresponding to two-level index, is processed to it:Umber of defectives
According to detection, the identification of bad data and repairing and missing values filling, form the training sample data that need to process further
Collection X;
Step 3:This training sample data collection X is standardized process, forms training sample data collection to be excavated,
Assume still to be represented with X;
Step 4:With PCA, synthesis is carried out to all kinds of index parameter values of collection, index number is subtracted from p
As little as m, obtains New Set zi(i=1,2 ..., calculation expression m), thus generate new training sample data collection Z;
Step 5:Gene expression programming is used to initialize population according to the sample data set Z obtaining;
Step 6:Build suitable fitness function, set and stop producing newly for the calculating reaching needed for stylish generation
Accuracy rating;
Step 7:Carry out selection operation, mutation operation, map function and reorganization operation for the individuality producing, produce
New individual;
Step 8:The new individual producing is sorted according to adaptive value size, retains excellent individual;
Step 9:Select multiformity operator pb、pr, p in certain proportionbRetain the poor individuality of fitness, with certain
Probability prRandomly generate a part of new individual, then produce new generation;
Step 10:Judge the new computational accuracy whether reaching setting for adaptive value producing, if reaching, exiting, going to step
Rapid 11, otherwise go to step 7 and repeat;
Step 11:Return the functional relation between comprehensive score and each achievement data, i.e. F=f (z1,z2,…,zm);
Step 12:The related index parameter sample number of obtained power distribution network current operating conditions after gathering and process
According to the independent variable as this functional relation, calculate comprehensive score F of current power distribution Running State;
Step 13:Inquiry risk class knowledge base, the comprehensive risk class waiting corresponding to point F of evaluation, and be given
Accordingly feasible behavior command;
Step 14:If power distribution network is in automatic mode, it is immediately performed command adapted thereto;Otherwise, user is transferred to voluntarily to process;
Step 15:The assessment of distribution Running State and prealarming process terminate.
Described data processing module is broadly divided into three steps to data processing:1) detection of bad data:Using comparing
Ripe bad data detection, and with reference to the correlation theory of power system, the historical data being gathered is detected, sentence
Break during this measurement is sampled and whether there is bad data;2) identification of bad data and repairing:When the gathered historical data of discovery
After there is bad data, need to further determine which or which metric data is bad data, and according to suitable method to the greatest extent
Bad data may be adjusted or assignment again;3) filling of missing values:In the historical data being gathered, due to operation
Improper and omit, or be that information the reason such as cannot obtain and can lead to shortage of data, and the presence of null value can be to entirely counting
Impact according to mining process, so we need by special method, the data of disappearance to be carried out deriving, fills, to reduce
Gap between data mining algorithm and practical application;The dissimilar or attribute that missing data can be directed to uses different null values
Processing method.
Running status evaluation module carries out running status assessment to the data processing through data processing module, first with master
Componential analysis carry out dimensionality reduction to multiple indexes parameter, comprehensive effectively index parameter, are then based on gene expression programming and calculate
Method realizes assessment and the early warning of distribution Running State.
(A) described using PCA, multiple indexes parameter is carried out with dimensionality reduction, comprehensive effectively index parameter, bag
Include:Using PCA, the historical data obtained by data processing module is carried out dimensionality reduction, in comprehensive each metrics evaluation
On the basis of simplify index amount input;Assume that the sample observation data matrix through data processing module output is:
Wherein xijWith regard to the numerical value corresponding to index j, the index number selected is p to i-th gathered sample of expression, altogether
Acquire n sample;
(B) described assessment and the early warning realizing distribution Running State based on gene expression programming, including:Pass through
PCA by affect distribution Running State multiple indexes carry out screening and synthesis after, in the method in addition it is also necessary to
Excavate F and z using based on gene expression programming from the training sample data collection Z obtaining after treatmentj(j
=1,2 ..., m) between functional relationship F=f (z1,z2,…,zm), zjRepresent through the calculated synthesis of PCA
Desired value;By gathering the related data of power distribution network current operating conditions, after PCA is processed, gained will be calculated
Numerical value according to corresponding relation as this functional relation input, calculate F value, then using calculate result pre- as risk
The input of alert module, thus being estimated to current distribution Running State and giving to feed back.
The main working process of step (B) is as follows:
(1) according to the training sample set Z initialization population obtaining through PCA;
(2) build suitable fitness function, set and stop producing newly for the computational accuracy reaching needed for stylish generation
Scope;
(3) it is directed to the individuality producing and carries out selection operation, mutation operation, map function and reorganization operation, produce newly individual
Body;
(4) new individual producing is sorted according to adaptive value size, retain excellent individual;
(5) select multiformity operator pb、pr, p in certain proportionbRetain the poor individuality of fitness, with certain probability
prRandomly generate a part of new individual, then produce new generation;
(6) judge the new computational accuracy whether reaching setting for adaptive value producing, if reaching, exiting, going to step
(7), otherwise go to step (3) to repeat;
(7) return tried to achieve functional relation, after then gathering and process, obtained power distribution network currently runs shape
The related index parameter sample data of state, as the independent variable of this functional relation, is calculated current power distribution Running State
F is finally committed to Risk-warning module by comprehensive score F, is easy to the evaluation of follow-up risk class.
The main working process of step (A) is as follows:
(1) the power distribution network historical data sample collecting is standardized processing, forms training sample to be excavated
Data set;Matrix samples X of (1) formula are standardized processing, relevant treatment formula is as follows:
Wherein
(2) calculate the correlation matrix R of this training sample:
For convenience it is assumed that the sample after initial data standardization is still represented with X, then the phase of the data after normalized process
Closing coefficient is:
(3) eigenvalue (λ of correlation matrix R is sought with Jacobian technique1,λ2…λp), that is, solution characteristic equation | λ I-R |=
0, order arranges by size to make the eigenvalue tried to achieve:
λ1≥λ2≥…≥λp≥0
It is also desirable to obtain respectively corresponding to eigenvalue λiCharacteristic vector ei(i=1,2 ..., p) it is desirable to | | ei||
=1, that is,eijRepresent vectorial eiJ-th component.
(4) principal component contributor rate and contribution rate of accumulative total are calculated;Here contribution rate just refers to that the variance of certain main constituent accounts for
All proportion of variance, actual namely certain eigenvalue accounts for the proportion that All Eigenvalues add up to, that is,:
Contribution rate is bigger, illustrates that the information of the original variable that this main constituent is comprised is stronger;
(5) select important main constituent, write out main constituent expression formula;Here the contribution rate of accumulative total mainly according to main constituent
To choose main constituent number, typically to take contribution rate of accumulative total to reach 85%~95% eigenvalue λ1,λ2,…,λmCorresponding
1st, the 2nd ..., the individual main constituent of m (m≤p);Therefore main constituent load lijComputing formula is as follows:
Assume the historical data that we are gathered index variable be xj, now main constituent ziFor:
Through above-mentioned process and analysis, instead of p original index variable with m main constituent, p shadow that will be original
The index ringing distribution Running State reduces to m, and now training sample set X is changed into Z through process, that is,:
Risk-warning module is passed through to inquire about risk of risk class knowledge base evaluation current power distribution Running State etc.
Level, provides feasible behavior command simultaneously, if operational control module is in automatic mode, this distribution network operation command adapted thereto;No
Then, user is transferred to voluntarily to process;If in addition, there being demand, this Risk-warning module can also respond checking of user's proposition and specifically comment
Estimate the request of result, assessment result is shown with visualization technique, help user and decision-maker preferably to plan and manage whole
The operation of individual power distribution network.
Claims (7)
1. a kind of distribution Running State assessment based on big data and method for early warning, is characterized in that:Entered by data processing module
Row data processing, carries out running status assessment by running status evaluation module to the data processing through data processing module, runs
The comprehensive score that state estimation obtains submits Risk-warning resume module to, and Risk-warning module evaluates current power distribution Running State
Risk class, provide feasible behavior command to operational control module simultaneously;
Comprise the following steps:
Step one:The actual demand of problem analysis, builds rational distribution Running State evaluation index system;
Step 2:Under collection evaluation system, the relevant historical data corresponding to two-level index, is processed to it:Bad data
The filling of detection, the identification of bad data and repairing and missing values, forms the training sample data collection X that need to process further;
Step 3:This training sample data collection X is standardized process, form training sample data collection to be excavated it is assumed that
Still represented with X;
Step 4:With PCA to collection all kinds of index parameter values carry out synthesis, by index number from p reduce to
M, obtains New Set zi(i=1,2 ..., calculation expression m), thus generate new training sample data collection Z;
Step 5:Gene expression programming is used to initialize population according to the sample data set Z obtaining;
Step 6:Build suitable fitness function, set and stop producing newly for the computational accuracy reaching needed for stylish generation
Scope;
Step 7:Carry out selection operation, mutation operation, map function and reorganization operation for the individuality producing, produce newly individual
Body;
Step 8:The new individual producing is sorted according to adaptive value size, retains excellent individual;
Step 9:Select multiformity operator pb、pr, p in certain proportionbRetain the poor individuality of fitness, with certain probability
prRandomly generate a part of new individual, then produce new generation;
Step 10:Judge the new computational accuracy whether reaching setting for adaptive value producing, if reaching, exiting, going to step 10
One, otherwise go to step 7 and repeat;
Step 11:Return the functional relation between comprehensive score and each achievement data, i.e. F=f (z1,z2,…,zm);
Step 12:The related index parameter sample data of obtained power distribution network current operating conditions after gathering and process is made
For the independent variable of this functional relation, calculate comprehensive score F of current power distribution Running State;
Step 13:Inquiry risk class knowledge base, the comprehensive risk class waiting corresponding to point F of evaluation, and be given corresponding
Feasible behavior command;
Step 14:If power distribution network is in automatic mode, it is immediately performed command adapted thereto;Otherwise, user is transferred to voluntarily to process;
Step 15:The assessment of distribution Running State and prealarming process terminate.
2. the distribution Running State assessment based on big data according to claim 1 and method for early warning, is characterized in that:Institute
State data processing module and data processing is broadly divided into three steps:1) detection of bad data:Bad using comparative maturity
Data detection method, and with reference to the correlation theory of power system, the historical data being gathered is detected, judge that this measurement is adopted
Whether there is bad data in sample;2) identification of bad data and repairing:When the historical data finding gathered has umber of defectives
According to rear, need to further determine which or which metric data is bad data, and according to suitable method as far as possible to bad
Data is adjusted or assignment again;3) filling of missing values:In the historical data being gathered, due to misoperation and something lost
Leakage, or be that information the reason such as cannot obtain and can lead to shortage of data, and the presence of null value can be to whole data mining
Journey impacts, so we need by special method, the data of disappearance to be carried out deriving, fills, to reduce data mining
Gap between algorithm and practical application;The dissimilar or attribute that missing data can be directed to uses different processing empty value sides
Method.
3. the distribution Running State assessment based on big data according to claim 1 and 2 and method for early warning, its feature
It is:Running status evaluation module carries out running status assessment to the data processing through data processing module, first with main constituent
Analytic process carries out dimensionality reduction to multiple indexes parameter, comprehensive effectively index parameter, is then based on gene expression programming real
The assessment of existing distribution Running State and early warning.
4. the distribution Running State assessment based on big data according to claim 3 and method for early warning, is characterized in that:
(A) described using PCA, multiple indexes parameter is carried out with dimensionality reduction, comprehensive effectively index parameter, including:Using master
Historical data obtained by data processing module is carried out dimensionality reduction by componential analysis, simplifies on the basis of comprehensive each metrics evaluation
The input of index amount;Assume that the sample observation data matrix through data processing module output is:
Wherein xijWith regard to the numerical value corresponding to index j, the index number selected is p to i-th gathered sample of expression, gathers altogether
N sample;
(B) described assessment and the early warning realizing distribution Running State based on gene expression programming, including:By main one-tenth
The multiple indexes affecting distribution Running State are carried out after screening and synthesis by point analytic process, in the method in addition it is also necessary to utilize
F and z is excavated from the training sample data collection Z obtaining after treatment based on gene expression programmingj(j=1,
2 ..., m) between functional relationship F=f (z1,z2,…,zm), zjRepresent through the calculated aggregative indicator of PCA
Value;By gathering the related data of power distribution network current operating conditions, after PCA is processed, the number of gained will be calculated
Value calculates the value of F according to corresponding relation as the input of this functional relation, then using the result calculating as Risk-warning mould
The input of block, thus being estimated to current distribution Running State and giving to feed back.
5. the distribution Running State assessment based on big data according to claim 4 and method for early warning, is characterized in that:Step
Suddenly the main working process of (B) is as follows:
(1) according to the training sample set Z initialization population obtaining through PCA;
(2) build suitable fitness function, set and stop producing newly for the computational accuracy scope reaching needed for stylish generation;
(3) it is directed to the individuality producing and carries out selection operation, mutation operation, map function and reorganization operation, produce new individual;
(4) new individual producing is sorted according to adaptive value size, retain excellent individual;
(5) select multiformity operator pb、pr, p in certain proportionbRetain the poor individuality of fitness, with certain Probability prWith
Machine produces a part of new individual, then produces new generation;
(6) judge the new computational accuracy whether reaching setting for adaptive value producing, if reaching, exiting, going to step (7), no
Then go to step (3) to repeat;
(7) return tried to achieve functional relation, power distribution network current operating conditions phase obtained after then gathering and process
The index parameter sample data closed, as the independent variable of this functional relation, is calculated the synthesis of current power distribution Running State
F is finally committed to Risk-warning module by score F, is easy to the evaluation of follow-up risk class.
6. the distribution Running State assessment based on big data according to claim 4 and method for early warning, is characterized in that:Step
Suddenly the main working process of (A) is as follows:
(1) the power distribution network historical data sample collecting is standardized processing, forms training sample data to be excavated
Collection;Matrix samples X of (1) formula are standardized processing, relevant treatment formula is as follows:
Wherein
(j=1,2 ..., p)
(2) calculate the correlation matrix R of this training sample:
For convenience it is assumed that the sample after initial data standardization is still represented with X, then the phase relation of the data after normalized process
Number is:
(i, j=1,2 ..., p)
(3) eigenvalue (λ of correlation matrix R is sought with Jacobian technique1,λ2…λp), that is, solve characteristic equation | λ I-R |=0, make
Order arranges the eigenvalue tried to achieve by size:
λ1≥λ2≥…≥λp≥0
It is also desirable to obtain respectively corresponding to eigenvalue λiCharacteristic vector ei(i=1,2 ..., p) it is desirable to | | ei| |=1,
I.e.eijRepresent vectorial eiJ-th component.
(4) principal component contributor rate and contribution rate of accumulative total are calculated;Here contribution rate just refers to that the variance of certain main constituent accounts for all
The proportion of variance, actual namely certain eigenvalue accounts for the proportion that All Eigenvalues add up to, that is,:
Contribution rate is bigger, illustrates that the information of the original variable that this main constituent is comprised is stronger;
(5) select important main constituent, write out main constituent expression formula;Here mainly selected according to the contribution rate of accumulative total of main constituent
Take main constituent number, typically take contribution rate of accumulative total to reach 85%~95% eigenvalue λ1,λ2,…,λmCorresponding the 1st,
2nd ..., the individual main constituent of m (m≤p);Therefore main constituent load lijComputing formula is as follows:
Assume the historical data that we are gathered index variable be xj, now main constituent ziFor:
Through above-mentioned process and analysis, instead of p original index variable with m main constituent, p impact that will be original is joined
The index of operation of power networks state reduces to m, and now training sample set X is changed into Z through process, that is,:
7. the distribution Running State assessment based on big data according to claim 1 and 2 and method for early warning, its feature
It is:Risk-warning module is passed through to inquire about the risk class that risk class knowledge base evaluates current power distribution Running State, with
When provide feasible behavior command, if operational control module is in automatic mode, this distribution network operation command adapted thereto;Otherwise, hand over
Voluntarily processed by user;If in addition, there being demand, this Risk-warning module can also respond user proposition check concrete assessment result
Request, assessment result is shown with visualization technique, helps user and decision-maker preferably to plan and manage whole distribution
The operation of net.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610837927.XA CN106483947A (en) | 2016-09-21 | 2016-09-21 | Distribution Running State assessment based on big data and method for early warning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610837927.XA CN106483947A (en) | 2016-09-21 | 2016-09-21 | Distribution Running State assessment based on big data and method for early warning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106483947A true CN106483947A (en) | 2017-03-08 |
Family
ID=58268707
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610837927.XA Pending CN106483947A (en) | 2016-09-21 | 2016-09-21 | Distribution Running State assessment based on big data and method for early warning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106483947A (en) |
Cited By (29)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107274067A (en) * | 2017-05-22 | 2017-10-20 | 国网江西省电力公司电力科学研究院 | A kind of distribution transformer overloads methods of risk assessment |
CN107527134A (en) * | 2017-07-11 | 2017-12-29 | 中国农业大学 | A kind of distribution transformer state evaluating method and device based on big data |
CN107609769A (en) * | 2017-09-06 | 2018-01-19 | 重庆邮电大学 | A kind of intelligent distribution network fault early warning method based on failure gene table |
CN107992959A (en) * | 2017-04-26 | 2018-05-04 | 国网浙江省电力公司 | A kind of power failure Forecasting Methodology based on electric power big data visualization Neural Network Data digging technology |
CN108090703A (en) * | 2018-01-17 | 2018-05-29 | 许军营 | A kind of analysis of distribution operation risk and assessment system based on big data |
CN108269007A (en) * | 2017-12-29 | 2018-07-10 | 广州供电局有限公司 | Method, apparatus, computer equipment and the storage medium of transformer data analysis |
CN108459582A (en) * | 2018-03-01 | 2018-08-28 | 中国航空无线电电子研究所 | Comprehensive health assessment method towards IMA systems |
CN108510147A (en) * | 2018-01-23 | 2018-09-07 | 浙江工业大学 | A kind of Comprehensive assessment of power quality method based on residual error volatility model |
CN108805427A (en) * | 2018-05-29 | 2018-11-13 | 深圳众厉电力科技有限公司 | A kind of distribution Running State Warning System based on big data |
CN109377016A (en) * | 2018-09-26 | 2019-02-22 | 广东电网有限责任公司 | O&M risk supervises method, apparatus and computer readable storage medium |
CN109540212A (en) * | 2018-11-30 | 2019-03-29 | 湖南世优电力科技股份有限公司 | A kind of cable shaft status assessment early warning system and method based on big data |
CN109670676A (en) * | 2018-11-26 | 2019-04-23 | 安徽继远软件有限公司 | Distributing net platform region method for prewarning risk and system based on Support Vector data description |
CN110110977A (en) * | 2019-04-19 | 2019-08-09 | 广西电网有限责任公司钦州供电局 | It is a kind of based on improve principal component-step analysis power distribution network appraisal procedure |
CN110889613A (en) * | 2019-11-21 | 2020-03-17 | 中国南方电网有限责任公司 | Power grid system splitting risk analysis method based on SCADA big data |
CN110990656A (en) * | 2019-10-14 | 2020-04-10 | 广西电网有限责任公司贺州供电局 | System for visual analysis of big data of smart power grid |
CN111126810A (en) * | 2019-12-16 | 2020-05-08 | 国网河北省电力有限公司电力科学研究院 | Method for evaluating operation safety of source side generator set |
CN111391694A (en) * | 2020-02-29 | 2020-07-10 | 国网江苏省电力有限公司苏州供电分公司 | Multi-level rapid data monitoring method and system for operation and maintenance of charging station |
CN111428780A (en) * | 2020-03-20 | 2020-07-17 | 上海理工大学 | Power grid abnormal operation state identification method based on data driving |
CN111614491A (en) * | 2020-05-06 | 2020-09-01 | 国网电力科学研究院有限公司 | Power monitoring system oriented safety situation assessment index selection method and system |
CN112053086A (en) * | 2020-09-18 | 2020-12-08 | 国网河南省电力公司濮阳供电公司 | Working method of operation diagnosis system based on big data power grid |
CN112668940A (en) * | 2021-01-18 | 2021-04-16 | 南京地铁建设有限责任公司 | Sharing-based subway power supply system operation evaluation method |
CN112734204A (en) * | 2020-12-01 | 2021-04-30 | 国家电网有限公司 | Power distribution network fault line risk assessment method and system |
CN112907096A (en) * | 2021-03-08 | 2021-06-04 | 江苏维诗环境科技有限公司 | Heavy metal pollution risk assessment system |
CN113036917A (en) * | 2021-03-08 | 2021-06-25 | 国网浙江省电力有限公司金华供电公司 | Power distribution network monitoring information monitoring system and method based on machine learning |
CN113379248A (en) * | 2021-06-11 | 2021-09-10 | 国网上海市电力公司 | Power grid risk assessment and early warning method based on complex network theory |
CN113807690A (en) * | 2021-09-09 | 2021-12-17 | 国网江苏省电力有限公司苏州供电分公司 | Online evaluation and early warning method and system for operation state of regional power grid regulation and control system |
CN116187745A (en) * | 2022-12-12 | 2023-05-30 | 安徽继远软件有限公司 | Situation awareness-based power grid risk assessment method |
CN116611712A (en) * | 2023-07-17 | 2023-08-18 | 国网信息通信产业集团有限公司 | Semantic inference-based power grid work ticket evaluation system |
CN116780782A (en) * | 2023-08-24 | 2023-09-19 | 成都易电云商工程服务有限公司 | Comprehensive maintenance system for electrical equipment of power distribution network |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101232182A (en) * | 2008-01-18 | 2008-07-30 | 清华大学 | Three-dimensional coordinated electric network energy managing system and method for controlling and evaluating electric network |
CN104217261A (en) * | 2014-09-22 | 2014-12-17 | 国家电网公司 | Operating state risk early-warning method for main transformer system |
CN104820953A (en) * | 2015-05-19 | 2015-08-05 | 重庆大学 | Power distribution network operation safety assessment and warning system |
CN105303468A (en) * | 2015-11-20 | 2016-02-03 | 国网天津市电力公司 | Comprehensive evaluation method of smart power grid construction based on principal component cluster analysis |
CN105654229A (en) * | 2015-11-26 | 2016-06-08 | 国家电网公司 | Power grid automation system and equipment running state risk assessment algorithm |
-
2016
- 2016-09-21 CN CN201610837927.XA patent/CN106483947A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101232182A (en) * | 2008-01-18 | 2008-07-30 | 清华大学 | Three-dimensional coordinated electric network energy managing system and method for controlling and evaluating electric network |
CN104217261A (en) * | 2014-09-22 | 2014-12-17 | 国家电网公司 | Operating state risk early-warning method for main transformer system |
CN104820953A (en) * | 2015-05-19 | 2015-08-05 | 重庆大学 | Power distribution network operation safety assessment and warning system |
CN105303468A (en) * | 2015-11-20 | 2016-02-03 | 国网天津市电力公司 | Comprehensive evaluation method of smart power grid construction based on principal component cluster analysis |
CN105654229A (en) * | 2015-11-26 | 2016-06-08 | 国家电网公司 | Power grid automation system and equipment running state risk assessment algorithm |
Non-Patent Citations (1)
Title |
---|
潘金月: "基于基因表达式编程的电力系统负荷建模", 《中国优秀硕士学位论文全文数据库(电子期刊)工程科技II辑》 * |
Cited By (38)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107992959A (en) * | 2017-04-26 | 2018-05-04 | 国网浙江省电力公司 | A kind of power failure Forecasting Methodology based on electric power big data visualization Neural Network Data digging technology |
CN107274067A (en) * | 2017-05-22 | 2017-10-20 | 国网江西省电力公司电力科学研究院 | A kind of distribution transformer overloads methods of risk assessment |
CN107527134A (en) * | 2017-07-11 | 2017-12-29 | 中国农业大学 | A kind of distribution transformer state evaluating method and device based on big data |
CN107609769B (en) * | 2017-09-06 | 2020-07-31 | 重庆邮电大学 | Intelligent power distribution network fault early warning method based on fault gene table |
CN107609769A (en) * | 2017-09-06 | 2018-01-19 | 重庆邮电大学 | A kind of intelligent distribution network fault early warning method based on failure gene table |
CN108269007A (en) * | 2017-12-29 | 2018-07-10 | 广州供电局有限公司 | Method, apparatus, computer equipment and the storage medium of transformer data analysis |
CN108090703A (en) * | 2018-01-17 | 2018-05-29 | 许军营 | A kind of analysis of distribution operation risk and assessment system based on big data |
CN108510147B (en) * | 2018-01-23 | 2022-01-11 | 浙江工业大学 | Electric energy quality comprehensive evaluation method based on residual error fluctuation model |
CN108510147A (en) * | 2018-01-23 | 2018-09-07 | 浙江工业大学 | A kind of Comprehensive assessment of power quality method based on residual error volatility model |
CN108459582A (en) * | 2018-03-01 | 2018-08-28 | 中国航空无线电电子研究所 | Comprehensive health assessment method towards IMA systems |
CN108805427A (en) * | 2018-05-29 | 2018-11-13 | 深圳众厉电力科技有限公司 | A kind of distribution Running State Warning System based on big data |
CN109377016A (en) * | 2018-09-26 | 2019-02-22 | 广东电网有限责任公司 | O&M risk supervises method, apparatus and computer readable storage medium |
CN109670676A (en) * | 2018-11-26 | 2019-04-23 | 安徽继远软件有限公司 | Distributing net platform region method for prewarning risk and system based on Support Vector data description |
CN109540212A (en) * | 2018-11-30 | 2019-03-29 | 湖南世优电力科技股份有限公司 | A kind of cable shaft status assessment early warning system and method based on big data |
CN110110977A (en) * | 2019-04-19 | 2019-08-09 | 广西电网有限责任公司钦州供电局 | It is a kind of based on improve principal component-step analysis power distribution network appraisal procedure |
CN110990656A (en) * | 2019-10-14 | 2020-04-10 | 广西电网有限责任公司贺州供电局 | System for visual analysis of big data of smart power grid |
CN110889613A (en) * | 2019-11-21 | 2020-03-17 | 中国南方电网有限责任公司 | Power grid system splitting risk analysis method based on SCADA big data |
CN111126810B (en) * | 2019-12-16 | 2023-08-22 | 国网河北省电力有限公司电力科学研究院 | Method for evaluating operation safety of source side generator set |
CN111126810A (en) * | 2019-12-16 | 2020-05-08 | 国网河北省电力有限公司电力科学研究院 | Method for evaluating operation safety of source side generator set |
CN111391694A (en) * | 2020-02-29 | 2020-07-10 | 国网江苏省电力有限公司苏州供电分公司 | Multi-level rapid data monitoring method and system for operation and maintenance of charging station |
CN111391694B (en) * | 2020-02-29 | 2023-08-11 | 国网江苏省电力有限公司苏州供电分公司 | Multi-level rapid data monitoring method and system for operation and maintenance of charging station |
CN111428780B (en) * | 2020-03-20 | 2023-04-07 | 上海理工大学 | Power grid abnormal operation state identification method based on data driving |
CN111428780A (en) * | 2020-03-20 | 2020-07-17 | 上海理工大学 | Power grid abnormal operation state identification method based on data driving |
CN111614491A (en) * | 2020-05-06 | 2020-09-01 | 国网电力科学研究院有限公司 | Power monitoring system oriented safety situation assessment index selection method and system |
CN112053086A (en) * | 2020-09-18 | 2020-12-08 | 国网河南省电力公司濮阳供电公司 | Working method of operation diagnosis system based on big data power grid |
CN112734204A (en) * | 2020-12-01 | 2021-04-30 | 国家电网有限公司 | Power distribution network fault line risk assessment method and system |
CN112668940A (en) * | 2021-01-18 | 2021-04-16 | 南京地铁建设有限责任公司 | Sharing-based subway power supply system operation evaluation method |
CN112668940B (en) * | 2021-01-18 | 2022-04-19 | 南京地铁建设有限责任公司 | Sharing-based subway power supply system operation evaluation method |
CN112907096A (en) * | 2021-03-08 | 2021-06-04 | 江苏维诗环境科技有限公司 | Heavy metal pollution risk assessment system |
CN113036917A (en) * | 2021-03-08 | 2021-06-25 | 国网浙江省电力有限公司金华供电公司 | Power distribution network monitoring information monitoring system and method based on machine learning |
CN113379248A (en) * | 2021-06-11 | 2021-09-10 | 国网上海市电力公司 | Power grid risk assessment and early warning method based on complex network theory |
CN113807690A (en) * | 2021-09-09 | 2021-12-17 | 国网江苏省电力有限公司苏州供电分公司 | Online evaluation and early warning method and system for operation state of regional power grid regulation and control system |
CN116187745A (en) * | 2022-12-12 | 2023-05-30 | 安徽继远软件有限公司 | Situation awareness-based power grid risk assessment method |
CN116187745B (en) * | 2022-12-12 | 2023-11-14 | 安徽继远软件有限公司 | Situation awareness-based power grid risk assessment method |
CN116611712A (en) * | 2023-07-17 | 2023-08-18 | 国网信息通信产业集团有限公司 | Semantic inference-based power grid work ticket evaluation system |
CN116611712B (en) * | 2023-07-17 | 2023-12-08 | 国网信息通信产业集团有限公司 | Semantic inference-based power grid work ticket evaluation system |
CN116780782A (en) * | 2023-08-24 | 2023-09-19 | 成都易电云商工程服务有限公司 | Comprehensive maintenance system for electrical equipment of power distribution network |
CN116780782B (en) * | 2023-08-24 | 2023-10-31 | 成都易电云商工程服务有限公司 | Comprehensive maintenance system for electrical equipment of power distribution network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106483947A (en) | Distribution Running State assessment based on big data and method for early warning | |
Wan et al. | A group decision making method with interval valued fuzzy preference relations based on the geometric consistency | |
CN103870659B (en) | A kind of fault of numerical control machine tool analyzes method | |
CN103793854B (en) | The overhead transmission line operation risk informatization evaluation method that Multiple Combination is optimized | |
CN107301296A (en) | Circuit breaker failure influence factor method for qualitative analysis based on data | |
CN108664010A (en) | Generating set fault data prediction technique, device and computer equipment | |
CN111598352A (en) | Concrete beam type bridge comprehensive evaluation method based on Bayesian network | |
CN105825342A (en) | Pipeline failure possibility evaluation method and system | |
CN107292502B (en) | Power distribution network reliability assessment method | |
CN109583520B (en) | State evaluation method of cloud model and genetic algorithm optimization support vector machine | |
CN103488869A (en) | Wind power generation short-term load forecast method of least squares support vector machine | |
CN105678481A (en) | Pipeline health state assessment method based on random forest model | |
CN110826237B (en) | Wind power equipment reliability analysis method and device based on Bayesian belief network | |
CN109522962B (en) | Chemical plant safety quantitative evaluation method | |
CN101320449A (en) | Power distribution network estimation method based on combination appraisement method | |
Siew-Hong et al. | Selection of optimal maintenance policy by using fuzzy multi criteria decision making method | |
Zhu et al. | A novel decision-making model for selecting a construction project delivery system | |
CN105975797A (en) | Product early-fault root cause recognition method based on fuzzy data processing | |
CN116522088B (en) | Nuclear power plant operation data analysis method and system based on machine learning | |
CN105574592B (en) | Aeroengine maintenance engineering management tool | |
CN116467674A (en) | Intelligent fault processing fusion updating system and method for power distribution network | |
CN112836921A (en) | Data monitoring analysis and optimization method based on power grid micro-data fusion | |
CN112508232A (en) | Short-circuit current limitation measure evaluation method based on multi-level fuzzy comprehensive evaluation model | |
CN103902845A (en) | Method for evaluating risk of simulation system based on fuzzy FMEA | |
CN110781206A (en) | Method for predicting whether electric energy meter in operation fails or not by learning meter-dismantling and returning failure characteristic rule |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | 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 |
Application publication date: 20170308 |
|
RJ01 | Rejection of invention patent application after publication |