CN107918639A - Based on electric power big data main transformer peak load forecasting method and data warehouse - Google Patents
Based on electric power big data main transformer peak load forecasting method and data warehouse Download PDFInfo
- Publication number
- CN107918639A CN107918639A CN201710979712.6A CN201710979712A CN107918639A CN 107918639 A CN107918639 A CN 107918639A CN 201710979712 A CN201710979712 A CN 201710979712A CN 107918639 A CN107918639 A CN 107918639A
- Authority
- CN
- China
- Prior art keywords
- data
- mrow
- load
- main transformer
- electric power
- 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
- 238000013277 forecasting method Methods 0.000 title claims description 7
- 238000000034 method Methods 0.000 claims abstract description 33
- 238000012706 support-vector machine Methods 0.000 claims abstract description 22
- 238000004458 analytical method Methods 0.000 claims abstract description 21
- 238000007418 data mining Methods 0.000 claims abstract description 15
- 238000000605 extraction Methods 0.000 claims abstract description 11
- 238000000926 separation method Methods 0.000 claims abstract description 6
- 238000007621 cluster analysis Methods 0.000 claims abstract description 4
- 230000006870 function Effects 0.000 claims description 40
- 238000012545 processing Methods 0.000 claims description 22
- 238000012549 training Methods 0.000 claims description 13
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000007405 data analysis Methods 0.000 claims description 8
- 240000002853 Nelumbo nucifera Species 0.000 claims description 7
- 235000006508 Nelumbo nucifera Nutrition 0.000 claims description 7
- 235000006510 Nelumbo pentapetala Nutrition 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 7
- 238000002360 preparation method Methods 0.000 claims description 6
- 230000001373 regressive effect Effects 0.000 claims description 6
- 238000004140 cleaning Methods 0.000 claims description 5
- 230000010354 integration Effects 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 239000012141 concentrate Substances 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 238000007619 statistical method Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000003672 processing method Methods 0.000 abstract description 5
- 238000007726 management method Methods 0.000 description 5
- 241001269238 Data Species 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013481 data capture Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000013439 planning Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000032258 transport Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/283—Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Water Supply & Treatment (AREA)
- Quality & Reliability (AREA)
- Primary Health Care (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to technical field of power systems, is related to a kind of method and data warehouse based on electric power big data main transformer peak load forecasting.Comprise the following steps：S1. statistics and analysis is carried out to network system historical data, establishes data warehouse；S2. series pattern analysis and cluster analysis data digging method are used, the knowledge of extraction is analyzed according to the decision-making purpose of end user, the data separation of most worthy is come out, submits to user；S3. by the historical data of arrangement, main transformer peak load data are pre-processed, establish the load forecasting model based on support vector machines, predict following main transformer load condition.A kind of method based on electric power big data main transformer peak load forecasting provided by the invention, by establishing data warehouse, establish data mining processing method, main transformer load variations situation is predicted using support vector machine method, realize accurate, reliable main transformer peak load forecasting, rational management system makes the operation of its safety economy.
Description
Technical field
The present invention relates to technical field of power systems, is born more particularly, to one kind based on electric power big data main transformer peak
The method of lotus prediction.
Background technology
With the commercialization and the marketization of electric system, the accuracy of load forecast transports power system security economy
Row and the national economic development are of great significance.The horizontal of load forecast work be as the management of an electric power enterprise
One of no distinctive marks for moving towards modernization.Load Prediction In Power Systems value is electric power as the important evidence in power trade
Company's customization electric energy quotation, operating scheme and Electric Power Network Planning provide necessary guiding, its precision of prediction will influence closely
Market, electric load are moved towards in the economic benefit of electric power enterprise, today that especially electric utility develops on an unprecedented scale in China, power consumption management
Oneself warp of forecasting problem becomes the important and difficult task that electric system faces.
The result of load forecast plays the quality of whole planning the influence of substantivity, numerous power workers
The target for making great efforts to seek is to accomplish accurately to predict.But reaching this point has very big difficulty, because electric quantity change
Influence factor is extremely complex, such load forecast, the simple data and information for relying on electric system itself
It can not be completed, in addition the relation between environmental factor and electricity is cannot to be easily described with function, is rephrased the statement
They are exactly a kind of fuzzy relation, and with the advance and development of power industry, electric system scale is increasing, scale
Become larger so that the factor such as numerous politics, economic, society or even meteorology is also and then added to the overall background of prediction together so that
The difficulty of prediction environment greatly increases.
The content of the invention
It is an object of the invention to overcome the deficiencies of the prior art and provide one kind to be based on electric power big data main transformer peak load
The method of prediction, by establishing data warehouse, establishes data mining processing method, is predicted and led using support vector machine method
Varying duty situation of change, realizes accurate, reliable main transformer peak load forecasting, and rational management system makes the fortune of its safety economy
OK.
To solve the above problems, technical solution provided by the invention is：One kind is based on electric power big data main transformer peak load
The method of prediction, wherein, comprise the following steps：
S1. statistics and analysis is carried out to network system historical data, establishes data warehouse；
S2. series pattern analysis and cluster analysis data digging method are used, according to the decision-making purpose of end user to carrying
The knowledge taken is analyzed, and the data separation of most worthy is come out, submits to user；
S3. by the historical data of arrangement, main transformer peak load data are pre-processed, foundation is based on support vector machines
Load forecasting model, predict following main transformer load condition.
Further, the S1 steps further include data processing, wherein, data processing includes the extraction of data, data
Reprint loading, the foundation of source data, data warehouse modeling.
Further, the data mining process in the S2 steps includes data preparation, data mining and explains assessment,
Wherein, the data preparation comprises the following steps：
Data in multifile or multiple database running environment are merged processing, solve the meaning of one's words by S201. data integration
Ambiguity, handles the omission in data and cleaning hash；
S202. data select, and are required, data are handled using some database manipulations, from data according to user
Extract the data acquisition system for needing to excavate；
S203. data prediction, reprocesses the data in S202 steps, check the integralities of data with it is consistent
Property, noise data therein is handled, the data of loss are filled up using statistical method, is done for further analysis
Prepare, and determine the type of dredge operation that will be carried out；
S204. data conversion, according to the needs of data mining, carries out mutual between discrete Value Data and continuous Value Data
The operation of calculating combination between conversion, the packet classification of data value, data item.
In the present invention, data mining uses series pattern analysis and clustering method, is needed from extracting data user
The knowledge wanted；Wherein, explain that assessment analyzes the knowledge of extraction according to the decision-making purpose of end user, most worthy
Data separation comes out, and submits to user.In this process, not only knowledge is expressed in a manner of it can be more readily understood,
Also efficiency evaluation is carried out to it, if cannot meet that user requires, above-mentioned data mining process should be repeated.
Further, load prediction includes in the S3 steps：
S301. the selection of sample and its input and output amount, input vector x in the load regressive prediction model of support vector machines
∈RmFor historical load, meteorologic factor and day type loading effects factor, output y is load to be predicted；Built according to own primary data
Vertical training sample set and forecast sample collection, establish support vector machines regressive object function, solve optimal solution and bring back to recurrence and determine
Plan functional equation, obtains returning decision function, finally calculates prediction result；
S302. the selection of kernel function, chooses gaussian radial basis function as the kernel function in regression model, Gauss is radially
Base kernel function formula is as follows：
S303. historical data is pre-processed and normalized；Smooth and normalized is carried out to historical data,
I.e. by initial data by linear transformation into [- 1,1] section, composing training sample set；Normalizing formula is：
Wherein, x 'iFor normalized data value, xiFor measured value, ximin=min (xi),ximax=max (xi), n is input
Vector dimension, that is, influence the number of load factor；
S304. SVM electric load models are established, SVM is established according to obtained training sample set and the kernel function of selection and is returned
Return object function；
S305. the average load of following some day is carried out using forecast sample and decision-making regression equation obtained in the previous step
Prediction；
S306. after the completion of predicting, secondary month/daily load truthful data is considered as own primary data, training sample is added and concentrates, according to
This analogizes the load prediction for completing whole month/day.
It is used for the data warehouse based on electric power big data main transformer peak load forecasting method the present invention also provides a kind of,
Wherein, the data warehouse includes data source modules, data processing module, data warehouse module, data analysis and digging
Dig module, application and display module, the data source modules, data processing module, data warehouse module, data analysis and digging
Dig module, be linked in sequence using with display module.Wherein, data warehouse module storage data are for analysis use, according to difference
Analysis requirement, data store by different degree of integration, it is similar to a central database, but is different from traditional number
According to storehouse.Since data are huge, using C/S structures；Establish information of the data warehouse not only for amount of storage, it is often more important that
These substantial amounts of data are analyzed and processed, so as to provide the effective service life for decision-making.Inquired about using the front end of data warehouse
Instrument, there is provided the function of random challenge, can carry out any condition, the inquiry of arbitrary patterns combination, any without writing
Program；The final purpose of data warehouse is built, allows staff can be easily using this integrated decision-making branch of data warehouse
Held in ring border is to obtain valuable information, so as to make fast accurately judgement and system to continually changing network operation situation
Fixed corresponding countermeasure.Using B/S modes, client service is integrated into general frame, is each work station by WEB server
The services such as inquiry, analysis are provided.
Further, the data source includes：Main transformer load data, real-time data of power grid, power equipment account over the years
And geography information.
In the present invention, any load prediction is all based on initial data, the collection of historical summary and arrangement it is bad,
The quality of load prediction can be directly affected.In historical load sequence, on the one hand, due to the influence of enchancement factor, load can be
The a certain moment produces the load point of the method for operation different from the past, in the case where there is improper load point, load sequence
Regularity is destroyed, and the similitude of load curve reduces, then the predictability of load is destroyed, and influences precision of prediction；It is another
Aspect, initial data capture system disturb if there is failure or outer signals, error of transmission just occur, causes data to be not allowed
True or shortage of data.These are all bad data Producing reasons.The present invention is picked when historical summary carries out data analysis pretreatment
Except these bad datas, ensure the perfect of data, numeral is accurate, and reflection is all level under normal condition, i.e., flat
Steadyization exceptional value and addendum missing data.The processing method that the present invention uses：With the data of the same type day of adjacent a cycle
A complete daily load sequential value is averagely obtained, then each daily load and this average load are carried out to the differentiation to ratio error
Analysis, error are modified more than 10%, can be replaced with average load value.
Load forecasting model is the summary of statistics track, prediction model be it is diversified, solve finite sample,
Support vector machines, which protrudes, in the identification of non-linear and higher-dimension embodies its advantage, and influences power system load factor with non-linear
Feature, using the superior Nonlinear Learning of support vector machines and estimated performance, the present invention proposes negative based on support vector machines
Lotus Forecasting Methodology model.The theoretical frame of machine learning and logical under the finite sample for the complete set established by support vector machines
With method, the practical problem such as small sample, non-linear, high dimension drawn game portion minimal point can be preferably solved, is born for main transformer peak
Lotus is predicted, can obtain higher accuracy.
Further, the data processing module includes：Data extracting unit, data reprint load units, source data
Establish unit, data warehouse modeling unit.Extraction, conversion, the cleaning of data are carried out to data source by special data-interface,
Into in data warehouse.The function to be realized of data processing module includes the extraction and conversion loading, the foundation of former data of data
With the modeling of data warehouse.Data processing is needed for different data sources by regular hour rule to the number in data warehouse
According to carry out refresh with integrating again.
Compared with prior art, beneficial effect is：One kind provided by the invention is based on electric power big data main transformer peak load
The method of prediction, is analyzed by a large amount of historical datas produced to Operation of Electric Systems and management, to these substantial amounts of numbers
According to effectively statistics, prediction and assessment is carried out, the letter for electric power enterprise science decision is therefrom rapidly and accurately extracted
Breath, provides foundation for quick, Accurate Prediction future main transformer peak load change, by establishing data warehouse, establishes data
Processing method is excavated, main transformer load variations situation is predicted using support vector machine method, realizes that accurate, reliable main transformer peak is born
Lotus predicts that rational management system makes the operation of its safety economy.
Brief description of the drawings
Fig. 1 is overall flow figure of the present invention.
Fig. 2 is load prediction flow chart of the present invention.
Fig. 3 is data warehouse schematic diagram of the present invention.
Embodiment
Attached drawing is only for illustration, it is impossible to is interpreted as the limitation to this patent；It is attached in order to more preferably illustrate the present embodiment
Scheme some components to have omission, zoom in or out, do not represent the size of actual product；To those skilled in the art,
Some known features and its explanation may be omitted and will be understood by attached drawing.Being given for example only property of position relationship described in attached drawing
Explanation, it is impossible to be interpreted as the limitation to this patent.
As shown in Figure 1, a kind of method based on electric power big data main transformer peak load forecasting, wherein, comprise the following steps：
S1. statistics and analysis is carried out to network system historical data, establishes data warehouse；
S2. series pattern analysis and cluster analysis data digging method are used, according to the decision-making purpose of end user to carrying
The knowledge taken is analyzed, and the data separation of most worthy is come out, submits to user；
S3. by the historical data of arrangement, main transformer peak load data are pre-processed, foundation is based on support vector machines
Load forecasting model, predict following main transformer load condition.
In certain embodiments, S1 steps further include data processing, wherein, data processing includes the extraction of data, data
Reprint loading, the foundation of source data, data warehouse modeling.
In certain embodiments, the data mining process in S2 steps includes data preparation, data mining and explains assessment,
Wherein, the data preparation comprises the following steps：
Data in multifile or multiple database running environment are merged processing, solve the meaning of one's words by S201. data integration
Ambiguity, handles the omission in data and cleaning hash；
S202. data select, and are required, data are handled using some database manipulations, from data according to user
Extract the data acquisition system for needing to excavate；
S203. data prediction, reprocesses the data in S202 steps, check the integralities of data with it is consistent
Property, noise data therein is handled, the data of loss are filled up using statistical method, is done for further analysis
Prepare, and determine the type of dredge operation that will be carried out；
S204. data conversion, according to the needs of data mining, carries out mutual between discrete Value Data and continuous Value Data
The operation of calculating combination between conversion, the packet classification of data value, data item.
In the present invention, data mining uses series pattern analysis and clustering method, is needed from extracting data user
The knowledge wanted；Wherein, explain that assessment analyzes the knowledge of extraction according to the decision-making purpose of end user, most worthy
Data separation comes out, and submits to user.In this process, not only knowledge is expressed in a manner of it can be more readily understood,
Also efficiency evaluation is carried out to it, if cannot meet that user requires, above-mentioned data mining process should be repeated.
In certain embodiments, load prediction includes in S3 steps：
S301. the selection of sample and its input and output amount, input vector x in the load regressive prediction model of support vector machines
∈RmFor historical load, meteorologic factor and day type loading effects factor, output y is load to be predicted；Built according to own primary data
Vertical training sample set and forecast sample collection, establish support vector machines regressive object function, solve optimal solution and bring back to recurrence and determine
Plan functional equation, obtains returning decision function, finally calculates prediction result；Sample input quantity characteristic item is chosen to be followed successively by：
1-24：24 months/day are averaged load data A={ a1, a2 ..., a24 } a few days ago for prediction；
25-48：Predict 24 month/day peak load data Max={ m1, m2 ..., m24 } a few days ago；
49-72：Predict 24 month/day minimum load data Min={ n1, n2 ..., n24 } a few days ago；
73-96：Predict the standard deviation STD={ S1, S2 ..., S24 } of 24 month/day average loads a few days ago；
97-120：24 months/day are averaged weather temperature T={ t1, t2 ..., t24 } a few days ago for prediction；
This 120 characteristic items are to realize the feature vector of load forecast to participate in the training of support vector machines.
S302. the selection of kernel function, SVM are stated by training sample set and kernel function, choose various forms of kernel functions
Different SVM regression models can be generated.Kernel function mainly has：Linear kernel function, Polynomial kernel function and gaussian radial basis function
The performance of kernel function SVM and the type of relationship of selected kernel function are little, nuclear parameter (the parameter σ in kernel function) and balance system
Number C is only the principal element for influencing SVM performances.But choose suitable kernel function and be conducive to reduce calculation amount, the present invention chooses
Gaussian radial basis function is as follows as the kernel function in regression model, gaussian radial basis function formula：
S303. historical data is pre-processed and normalized；Smooth and normalized is carried out to historical data,
I.e. by initial data by linear transformation into [- 1,1] section, composing training sample set；Normalizing formula is：
Wherein, x 'iFor normalized data value, xiFor measured value, ximin=min (xi),ximax=max (xi), n is input
Vector dimension, that is, influence the number of load factor；
S304. SVM electric load models are established, SVM is established according to obtained training sample set and the kernel function of selection and is returned
Return object function；
S305. the average load of following some day is carried out using forecast sample and decision-making regression equation obtained in the previous step
Prediction；
S306. after the completion of predicting, secondary month/daily load truthful data is considered as own primary data, training sample is added and concentrates, according to
This analogizes the load prediction for completing whole month/day.
It is used for the data warehouse based on electric power big data main transformer peak load forecasting method the present invention also provides a kind of,
Wherein, the data warehouse includes data source modules, data processing module, data warehouse module, data analysis and digging
Dig module, application and display module, the data source modules, data processing module, data warehouse module, data analysis and digging
Dig module, be linked in sequence using with display module.Wherein, data warehouse module storage data are for analysis use, according to difference
Analysis requirement, data store by different degree of integration, it is similar to a central database, but is different from traditional number
According to storehouse.Since data are huge, using C/S structures；Establish information of the data warehouse not only for amount of storage, it is often more important that
These substantial amounts of data are analyzed and processed, so as to provide the effective service life for decision-making.Inquired about using the front end of data warehouse
Instrument, there is provided the function of random challenge, can carry out any condition, the inquiry of arbitrary patterns combination, any without writing
Program；The final purpose of data warehouse is built, allows staff can be easily using this integrated decision-making branch of data warehouse
Held in ring border is to obtain valuable information, so as to make fast accurately judgement and system to continually changing network operation situation
Fixed corresponding countermeasure.Using B/S modes, client service is integrated into general frame, is each work station by WEB server
The services such as inquiry, analysis are provided.
In certain embodiments, data source includes：Main transformer load data, real-time data of power grid, power equipment account over the years
And geography information.
In the present invention, any load prediction is all based on initial data, the collection of historical summary and arrangement it is bad,
The quality of load prediction can be directly affected.In historical load sequence, on the one hand, due to the influence of enchancement factor, load can be
The a certain moment produces the load point of the method for operation different from the past, in the case where there is improper load point, load sequence
Regularity is destroyed, and the similitude of load curve reduces, then the predictability of load is destroyed, and influences precision of prediction；It is another
Aspect, initial data capture system disturb if there is failure or outer signals, error of transmission just occur, causes data to be not allowed
True or shortage of data.These are all bad data Producing reasons.The present invention is picked when historical summary carries out data analysis pretreatment
Except these bad datas, ensure the perfect of data, numeral is accurate, and reflection is all level under normal condition, i.e., flat
Steadyization exceptional value and addendum missing data.The processing method that the present invention uses：With the data of the same type day of adjacent a cycle
A complete daily load sequential value is averagely obtained, then each daily load and this average load are carried out to the differentiation to ratio error
Analysis, error are modified more than 10%, can be replaced with average load value.
Load forecasting model is the summary of statistics track, prediction model be it is diversified, solve finite sample,
Support vector machines, which protrudes, in the identification of non-linear and higher-dimension embodies its advantage, and influences power system load factor with non-linear
Feature, using the superior Nonlinear Learning of support vector machines and estimated performance, the present invention proposes negative based on support vector machines
Lotus Forecasting Methodology model.The theoretical frame of machine learning and logical under the finite sample for the complete set established by support vector machines
With method, the practical problem such as small sample, non-linear, high dimension drawn game portion minimal point can be preferably solved, is born for main transformer peak
Lotus is predicted, can obtain higher accuracy.
Specifically, data processing module includes：Data extracting unit, data reprint load units, source data establishes unit,
Data warehouse modeling unit.Extraction, conversion, the cleaning of data are carried out to data source by special data-interface, into data
In warehouse.The function to be realized of data processing module includes extraction and conversion loading, the foundation of former data and data bins of data
The modeling in storehouse.Data processing needs to brush the data in data warehouse by regular hour rule for different data sources
Newly with integrating again.
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 field, 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 (7)
- A kind of 1. method based on electric power big data main transformer peak load forecasting, it is characterised in that comprise the following steps：S1. statistics and analysis is carried out to network system historical data, establishes data warehouse；S2. series pattern analysis and cluster analysis data digging method are used, according to the decision-making purpose of end user to extraction Knowledge is analyzed, and the data separation of most worthy is come out, submits to user；S3. by the historical data of arrangement, main transformer peak load data are pre-processed, are established negative based on support vector machines Lotus prediction model, predicts following main transformer load condition.
- 2. the method according to claim 1 based on electric power big data main transformer peak load forecasting, it is characterised in that described S1 steps further include data processing, wherein, the extraction of data processing including data, the reprinting of data load, source data is built Vertical, data warehouse modeling.
- 3. the method according to claim 1 based on electric power big data main transformer peak load forecasting, it is characterised in that described S2 steps in data mining process include data preparation, data mining and explain assessment, wherein, the data preparation bag Include following steps：Data in multifile or multiple database running environment are merged processing, resolve Ambiguity by S201. data integration Property, handle the omission in data and cleaning hash；S202. data select, and are required according to user, data are handled using some database manipulations, from extracting data Go out to need the data acquisition system excavated；S203. data prediction, reprocesses the data in S202 steps, checks the integrality and uniformity of data, right Noise data therein is handled, and the data of loss are filled up using statistical method, is prepared for further analysis, And determine the type of dredge operation that will be carried out；S204. data conversion, according to the needs of data mining, between discrete Value Data and continuous Value Data mutually turn Change, the operation that the packet of data value is classified, the calculating between data item is combined.
- 4. the method according to claim 1 based on electric power big data main transformer peak load forecasting, it is characterised in that described S3 steps in load prediction include：S301. the selection of sample and its input and output amount, input vector x ∈ R in the load regressive prediction model of support vector machinesm For historical load, meteorologic factor and day type loading effects factor, output y is load to be predicted；Established and instructed according to own primary data Practice sample set and forecast sample collection, establish support vector machines regressive object function, solve optimal solution and bring back to recurrence decision-making letter Number equation, obtains returning decision function, finally calculates prediction result；S302. the selection of kernel function, chooses gaussian radial basis function as the kernel function in regression model, gaussian radial basis function core Function formula is as follows：<mrow> <mi>k</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <msup> <mi>x</mi> <mo>&prime;</mo> </msup> <mo>)</mo> </mrow> <mo>=</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mo>-</mo> <mfrac> <mrow> <mo>|</mo> <mo>|</mo> <mi>x</mi> <mo>-</mo> <msup> <mi>x</mi> <mo>&prime;</mo> </msup> <mo>|</mo> <msup> <mo>|</mo> <mn>2</mn> </msup> </mrow> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> </mfrac> <mo>)</mo> </mrow> <mo>;</mo> </mrow>S303. historical data is pre-processed and normalized；Smooth and normalized is carried out to historical data, will Initial data by linear transformation into [- 1,1] section, composing training sample set；Normalizing formula is：<mrow> <msubsup> <mi>x</mi> <mi>i</mi> <mo>&prime;</mo> </msubsup> <mo>=</mo> <mfrac> <mrow> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>-</mo> <msub> <mi>x</mi> <mrow> <mi>i</mi> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> </msub> </mrow> <mrow> <msub> <mi>x</mi> <mrow> <mi>i</mi> <mi>m</mi> <mi>a</mi> <mi>x</mi> </mrow> </msub> <mo>-</mo> <msub> <mi>x</mi> <mrow> <mi>i</mi> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> </msub> </mrow> </mfrac> <mo>,</mo> <mrow> <mo>(</mo> <mi>i</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mi>n</mi> <mo>)</mo> </mrow> <mo>,</mo> </mrow>Wherein, x 'iFor normalized data value, xiFor measured value, ximin=min (xi),ximax=max (xi), n is input vector Dimension, that is, influence the number of load factor；S304. SVM electric load models are established, establishing SVM according to obtained training sample set and the kernel function of selection returns mesh Scalar functions；S305. the average load of following some day is carried out using forecast sample and decision-making regression equation obtained in the previous step pre- Survey；S306. after the completion of predicting, secondary month/daily load truthful data is considered as own primary data, training sample is added and concentrates, according to this class Push away the load prediction for completing whole month/day.
- 5. a kind of be used for the data warehouse based on electric power big data main transformer peak load forecasting method, it is characterised in that institute The data warehouse stated include data source modules, data processing module, data warehouse module, data analysis with excavate module, Using with display module, the data source modules, data processing module, data warehouse module, data analysis with excavate module, It is linked in sequence using with display module.
- 6. according to claim 5 be used for the data warehouse system based on electric power big data main transformer peak load forecasting method System, it is characterised in that the data source includes：Over the years main transformer load data, real-time data of power grid, power equipment account and Geography information.
- 7. according to claim 5 be used for the data warehouse system based on electric power big data main transformer peak load forecasting method System, it is characterised in that the data processing module includes：Data extracting unit, data reprint load units, source data is established The modeling unit of unit, data warehouse.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710979712.6A CN107918639A (en) | 2017-10-19 | 2017-10-19 | Based on electric power big data main transformer peak load forecasting method and data warehouse |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710979712.6A CN107918639A (en) | 2017-10-19 | 2017-10-19 | Based on electric power big data main transformer peak load forecasting method and data warehouse |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107918639A true CN107918639A (en) | 2018-04-17 |
Family
ID=61894827
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710979712.6A Pending CN107918639A (en) | 2017-10-19 | 2017-10-19 | Based on electric power big data main transformer peak load forecasting method and data warehouse |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107918639A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108549249A (en) * | 2018-05-24 | 2018-09-18 | 栾友祥 | A kind of intelligent domestic system based on big data |
| CN109583503A (en) * | 2018-12-03 | 2019-04-05 | 国网江苏省电力有限公司扬州供电分公司 | A kind of interruptible load prediction technique |
| CN109978230A (en) * | 2019-02-15 | 2019-07-05 | 杭州博钊科技有限公司 | A kind of electric power consumption sum intelligent Forecasting based on depth convolutional neural networks |
| CN110311369A (en) * | 2019-05-31 | 2019-10-08 | 广东电网有限责任公司 | A kind of stabilization of power grids section short term curve prediction method and system |
| CN110728395A (en) * | 2019-09-05 | 2020-01-24 | 广州供电局有限公司 | Main transformer short-term power load calculation method and device, computer and storage medium |
| CN110909904A (en) * | 2018-09-17 | 2020-03-24 | 济南热力集团有限公司 | Terminal user load prediction system based on wireless interconnection and data mining technology |
| CN110929954A (en) * | 2019-12-05 | 2020-03-27 | 中国南方电网有限责任公司 | Load prediction method, load prediction device, computer equipment and storage medium |
| CN112287562A (en) * | 2020-11-18 | 2021-01-29 | 国网新疆电力有限公司经济技术研究院 | Power equipment retired data completion method and system |
| CN117272121A (en) * | 2023-11-21 | 2023-12-22 | 江苏米特物联网科技有限公司 | Hotel load influence factor quantitative analysis method based on Deep SHAP |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103559562A (en) * | 2013-11-20 | 2014-02-05 | 贵州电网公司电力调度控制中心 | Power grid intelligent operation system and achieving method thereof |
| CN104537429A (en) * | 2014-12-11 | 2015-04-22 | 国家电网公司 | Short-term load forecasting method and device based on data warehouse and data mining technology |
| CN105574612A (en) * | 2015-12-14 | 2016-05-11 | 安徽工程大学 | Photovoltaic power generation prediction method based on data mining |
-
2017
- 2017-10-19 CN CN201710979712.6A patent/CN107918639A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103559562A (en) * | 2013-11-20 | 2014-02-05 | 贵州电网公司电力调度控制中心 | Power grid intelligent operation system and achieving method thereof |
| CN104537429A (en) * | 2014-12-11 | 2015-04-22 | 国家电网公司 | Short-term load forecasting method and device based on data warehouse and data mining technology |
| CN105574612A (en) * | 2015-12-14 | 2016-05-11 | 安徽工程大学 | Photovoltaic power generation prediction method based on data mining |
Non-Patent Citations (1)
| Title |
|---|
| 罗楠 等: "支持向量机方法在电力负荷预测中的应用", 《电网技术》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108549249A (en) * | 2018-05-24 | 2018-09-18 | 栾友祥 | A kind of intelligent domestic system based on big data |
| CN110909904A (en) * | 2018-09-17 | 2020-03-24 | 济南热力集团有限公司 | Terminal user load prediction system based on wireless interconnection and data mining technology |
| CN109583503A (en) * | 2018-12-03 | 2019-04-05 | 国网江苏省电力有限公司扬州供电分公司 | A kind of interruptible load prediction technique |
| CN109978230A (en) * | 2019-02-15 | 2019-07-05 | 杭州博钊科技有限公司 | A kind of electric power consumption sum intelligent Forecasting based on depth convolutional neural networks |
| CN109978230B (en) * | 2019-02-15 | 2021-08-10 | 杭州博钊科技有限公司 | Intelligent power sale amount prediction method based on deep convolutional neural network |
| CN110311369A (en) * | 2019-05-31 | 2019-10-08 | 广东电网有限责任公司 | A kind of stabilization of power grids section short term curve prediction method and system |
| CN110311369B (en) * | 2019-05-31 | 2020-09-25 | 广东电网有限责任公司 | Power grid stable section short-term load curve prediction method and system |
| CN110728395A (en) * | 2019-09-05 | 2020-01-24 | 广州供电局有限公司 | Main transformer short-term power load calculation method and device, computer and storage medium |
| CN110929954A (en) * | 2019-12-05 | 2020-03-27 | 中国南方电网有限责任公司 | Load prediction method, load prediction device, computer equipment and storage medium |
| CN112287562A (en) * | 2020-11-18 | 2021-01-29 | 国网新疆电力有限公司经济技术研究院 | Power equipment retired data completion method and system |
| CN112287562B (en) * | 2020-11-18 | 2023-03-10 | 国网新疆电力有限公司经济技术研究院 | Power equipment retired data completion method and system |
| CN117272121A (en) * | 2023-11-21 | 2023-12-22 | 江苏米特物联网科技有限公司 | Hotel load influence factor quantitative analysis method based on Deep SHAP |
| CN117272121B (en) * | 2023-11-21 | 2024-03-12 | 江苏米特物联网科技有限公司 | Hotel load influence factor quantitative analysis method based on Deep SHAP |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107918639A (en) | Based on electric power big data main transformer peak load forecasting method and data warehouse | |
| Tan et al. | Combining multi-criteria decision making (MCDM) methods with building information modelling (BIM): A review | |
| CN102708149A (en) | Data quality management method and system | |
| CN106570778A (en) | Big data-based data integration and line loss analysis and calculation method | |
| CN102194164A (en) | Index-system-model-based agile business intelligence data construction method | |
| CN107330589A (en) | Satellite network coordinates the quantitative evaluation method and system of risk | |
| Feng et al. | [Retracted] Design and Simulation of Human Resource Allocation Model Based on Double‐Cycle Neural Network | |
| CN108694470A (en) | A kind of data predication method and device based on artificial intelligence | |
| CN115860529A (en) | Supply chain carbon accounting system based on industrial internet | |
| Zheng et al. | [Retracted] Analysis of Enterprise Human Resources Demand Forecast Model Based on SOM Neural Network | |
| Chen | [Retracted] Construction Project Cost Management and Control System Based on Big Data | |
| CN110222129A (en) | A kind of credit appraisal algorithm based on relevant database | |
| CN114331041A (en) | Smart power plant maturity assessment method and system based on information physical system architecture | |
| Qin | The construction of corporate financial management risk model based on XGBoost algorithm | |
| CN105550813A (en) | Commercial evaluation method for value network-based mechanized equipment cloud service resources | |
| Xu | Research on enterprise knowledge unified retrieval based on industrial big data | |
| Gerlach et al. | Critical Success Factors for AI-driven Smart Energy Services. | |
| Jin et al. | Prediction model of employee flow based on semi-markov chain and its application analysis | |
| Weiguo et al. | Research on the application of smart logistics system based on big data: Taking jingdong logistics as an example | |
| CN117934209B (en) | Regional power system carbon emission big data analysis method based on knowledge graph | |
| Li | [Retracted] Economic Development Analysis of China’s Coastal Industries Based on Deep Learning and SDA Model Weighting | |
| Rong | Analysis of the comprehensive evaluation model of enterprise technological innovation ability based on improved genetic algorithm | |
| Weng et al. | Performance Evaluation of High-tech Innovation Enterprises Based on Grey Chaotic Neural Network | |
| Deng et al. | PTS‐FNN‐Based Health Prediction Method for Flexible Photoelectric Film Material Processing Equipment | |
| Jia-xu et al. | Analysis and Research on Complex System of Power Grid Company Assets Management targets and functions under the Optimum of Overall Situation |
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 |
Application publication date: 20180417 |
|
| RJ01 | Rejection of invention patent application after publication |