CN108665108A - A kind of big region electricity demand forecasting method and system based on big data - Google Patents
A kind of big region electricity demand forecasting method and system based on big data Download PDFInfo
- Publication number
- CN108665108A CN108665108A CN201810462414.4A CN201810462414A CN108665108A CN 108665108 A CN108665108 A CN 108665108A CN 201810462414 A CN201810462414 A CN 201810462414A CN 108665108 A CN108665108 A CN 108665108A
- Authority
- CN
- China
- Prior art keywords
- electricity consumption
- vector
- big
- data
- parameter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The big region electricity demand forecasting method and system based on big data that the present invention relates to a kind of, including predictive server, interface equipment and database;Processor in the predictive server obtains the set of parameter vector first, then the optimal leading time arrow of parameter vector is obtained, then the interconnection vector between optimal leading parameter and electricity consumption is obtained, the prediction electricity consumption in big region in next time cycle is finally obtained.The present invention is real-time, and the big metastable feature of region electricity consumption is fully taken into account, electricity consumption is regarded as including both fundamental invariant and variable, use electricity consumption change rate indirect predictions electricity consumption, so that the error of electricity consumption change rate is similar to the error of electricity consumption, interfered caused by substantially avoided invariant.
Description
Technical field
The present invention relates to big data analysis and electricity demand forecasting field, especially a kind of big region based on big data is used
Power predicating method and system.
Background technology
The electricity demand forecasting in big region (such as big region and the whole nation of provincial region including multiple provinces) is always
The hot spot of power industry research, for example, Xu Lina《The foundation and requirement forecasting of Chinese electricity consumption demand model》, Zhang Minwei
's《The provincial monthly power supply volume prediction of Utilities Electric Co. is inquired under market condition》, stone snow plum etc.《Peace based on K-L information Contents Methods
Emblem province industry electricity demand forecasting》Etc. plurality of articles be directed to big region electricity demand forecasting.
These prediction techniques can substantially be divided into two classes:The first kind be according to history electricity consumption data foundation electric model, it is right
Following electricity consumption is predicted;Second class is to determine ginseng in advance according to the leading parameter of one or more electricity consumptions of selection
Several leading periods passes through the variation of the variation prediction electricity consumption of leading parameter;Such as《Anhui Province based on K-L information Contents Methods
Industrial power quantity predicting》Wen Zhong, the leading electricity consumption of budgetary expenditure of local government 9 months, therefore can be according to budgetary expenditure of local government
Electricity consumption variation of the variation prediction after 9 months.
There are following technical problems for above-mentioned prediction technique:
The first, real-time is not strong, i.e., the leading period is predetermined quiescent period rather than dynamic is newer, pre-
Directly used when surveying electricity consumption, after leading to a period of time, the leading period it is possible that obvious error, to give
Electricity consumption brings error.Such as the current leading electricity consumption of budgetary expenditure of local government 9 months, it may be after several years, budgetary expenditure of local government is
Through becoming leading electricity consumption 7 months, in this case, still predicted apparent error will occur with 9 months.
The second, directly prediction electricity consumption causes the error of prediction result larger, rejects seasonal effect, electricity consumption is each
The variation of time cycle (such as moon) will not be it is obvious that for example《The industrial power quantity predicting in Anhui Province based on K-L information Contents Methods》
In Fig. 4 of one text, January, March, the electricity consumption in April are almost the same.Moreover, the electricity consumption basis in big region itself is very big, it is this
In the case of, the error of 4% or so predicted value can't fully meet required precision.
Invention content
In view of this, the big region electricity demand forecasting method that the purpose of the present invention is to propose to a kind of based on big data and being
System, it is real-time, and the big metastable feature of region electricity consumption has been fully taken into account, electricity consumption is regarded as including basic
Both invariant and variable use electricity consumption change rate indirect predictions electricity consumption so that the error of electricity consumption change rate is similar
In the error of electricity consumption, interfered caused by substantially avoided invariant.
The present invention is realized using following scheme:A kind of big region electricity demand forecasting method based on big data, including it is following
Step:
Step S1:Acquire the set X={ X of the relevant N groups lookahead data vector of big region electricity consumption1,X2,...,XN,
Each vector Xi=(xi(-ti),xi(-ti+1),...,xi(-1)), the value range of i is from 1 to N;Acquire electricity consumption vector E=(e-te,
e-te+1,...,e-1);The vector XiIt is identical as the measurement period of E, the vector XiEach of value be corresponding timing statistics
Lookahead data value in period, value is the electricity consumption in the corresponding timing statistics period each of in the vector E;
Step S2:Set P={ the P of parameter vector are obtained according to set X1,P2,...Pi,...,PN, each parameter to
AmountWherein,The value model of m
It encloses and arrives t for 1i-1;
Step S3:According to the set P of electricity consumption vector E and parameter vector, obtain the optimal look ahead time of parameter vector to
Measure T=(T1,T2,...Ti,...,TN), wherein TiFor parameter vector PiIn look ahead time corresponding to optimal leading parameter;
Step S4:Obtain the interconnection vector K=(k between optimal leading parameter and electricity consumption change rate1,k2,...,
ki...,kN);
Step S5:Obtain the prediction electricity consumption in big region in next time cycle
Wherein, e-1For the practical electricity consumption in big region in current time period.
Further, in step S1, the vector XiMeasurement period with E is the moon.
Further, in step S1, the lookahead data is the data after being modified to the data in data source, institute
It is the number of days in data divided by this month before correcting to state revised data;The electricity consumption is after being modified to practical electricity consumption
Electricity consumption, the revised electricity consumption be correct before electricity consumption divided by this month number of days.
Further, in the step S3, optimal leading time arrow T is obtained using K-L information Contents Methods.
Further, the step S3 specifically includes following steps:
Step S31:Electricity consumption change rate vector is calculated according to electricity consumption vector E
Wherein,The value range of m is 1 to te-1;
Step S32:For each parameter vector P in parameter vector set Pi, extract PiIn teA subvector, each
There is t in subvectoreThe collection of -1 parameter, the subvector is combined intoAny subvector PijForThe value range of j is 1 to te;
Step S33:Calculate the related coefficient of each subvector of electricity consumption change rate vector sumIts
Middle Cov () is covariance function, and σ () is variance function, and the value range of j is 1 to te;
Step S34:Obtain ρijIn maximum value ρimaxWith the j values of the correspondence maximum value, and it is denoted as jmax;
Step S35:By jmaxAs parameter vector PiIn look ahead time T corresponding to optimal leading parameteri, i.e. Ti=
jmax。
Further, the step S34 further includes:If ρimaxMore than or equal to predetermined threshold value, then executing step
S35;Otherwise, step S35 is no longer executed, and by the ρimaxCorresponding PiIt is rejected from set P.
Further, in step S4, the interconnection vector K=(k1,k2,...,ki,...,kN) acquisition by phase relation
Several ratios are worth to, i.e.,:
Further, in step S4, the interconnection vector K=(k1,k2,...,ki,...,kN) acquisition be using most
Small square law analytic equation Z=KQ, wherein:
Z=(δ-1,δ-2,...,δ-M)T;
Wherein, M < N.
The big region electricity demand forecasting system based on big data that the present invention also provides a kind of, including predictive server, connect
Jaws equipment and database;
The interface equipment is for being connected to more than one data source, to obtain the relevant N groups of big region electricity consumption
The set of lookahead data vector;The set of electricity consumption vector E and lookahead data vector are stored in the database;
The predictive server includes processor and storage medium, and it is any that the storage medium is stored with claim 1-8
The computer program of the method, whenever predicting the electricity consumption of next time cycle, the processor executes above-mentioned meter
Calculation machine program.
Compared with prior art, the present invention has following advantageous effect:
1, real-time, i.e., whenever predicting the electricity consumption of next time cycle, the present invention is performed both by step S1 to step
Rapid S5 also avoids the error caused by real-time reason to ensure that real-time.
2, fully take into account the big metastable feature of region electricity consumption, electricity consumption is regarded as including fundamental invariant and
Both variables use electricity consumption change rate indirect predictions electricity consumption so that the error of electricity consumption change rate is similar to electricity consumption
Error, substantially avoided and interfere caused by invariant.
Description of the drawings
Fig. 1 is the method flow schematic diagram of the present invention.
Specific implementation mode
The present invention will be further described with reference to the accompanying drawings and embodiments.
As shown in Figure 1, a kind of big region electricity demand forecasting method based on big data is present embodiments provided, including with
Lower step:
Step S1:Acquire the set X={ X of the relevant N groups lookahead data vector of big region electricity consumption1,X2,...,XN,
Each vector Xi=(xi(-ti),xi(-ti+1),...,xi(-1)), the value range of i is from 1 to N;Acquire electricity consumption vector E=
(e-te,e-te+1,...,e-1);The vector XiIt is identical as the measurement period of E, the vector XiEach of value be corresponding statistics
Lookahead data value in time cycle, value is the electricity consumption in the corresponding timing statistics period each of in the vector E;
Step S2:Set P={ the P of parameter vector are obtained according to set X1,P2,...Pi,...,PN, each parameter to
AmountWherein,The value model of m
It encloses and arrives t for 1i-1;PiIn number of parameters than corresponding XiIn lookahead data few 1 of quantity;
Step S3:According to the set P of electricity consumption vector E and parameter vector, obtain the optimal look ahead time of parameter vector to
Measure T=(T1,T2,...Ti,...,TN), wherein TiFor parameter vector PiIn look ahead time corresponding to optimal leading parameter;Root
According to one embodiment of the present of invention, can utilize《The industrial power quantity predicting in Anhui Province based on K-L information Contents Methods》It is introduced in text
Method obtain optimal leading time arrow T, or using other methods in the prior art obtain optimal look ahead times to
Measure T.
Step S4:Obtain the interconnection vector K=(k between optimal leading parameter and electricity consumption change rate1,k2,...,
ki...,kN);
Step S5:Obtain the prediction electricity consumption (such as predicting average daily electricity consumption) in big region in next time cycleWherein, e-1For practical electricity consumption (such as practical day in big region in current time period
Equal electricity consumption).
The present embodiment additionally provides a kind of big region electricity demand forecasting system based on big data, including predictive server,
Interface equipment and database;The interface equipment is for being connected to more than one data source, to obtain big region electricity consumption
The set of relevant N groups lookahead data vector;Be stored in the database electricity consumption vector E and the lookahead data to
The set of amount;The predictive server includes processor and storage medium, and the storage medium is stored with claim 1-8 and appoints
The computer program of one the method, whenever predicting the electricity consumption of next time cycle, the processor executes above-mentioned
Computer program.
Preferably, in the present embodiment, according to the present invention, lookahead data is the data that can react the following electricity consumption, first
Row data can include but is not limited to《The industrial power quantity predicting in Anhui Province based on K-L information Contents Methods》Cited warp in text
It helps data, the physical data known by sensor that can also be including precipitation, temperature etc., in this hair embodiment not
Any restriction is done to the type of lookahead data.In the present embodiment, the data source for obtaining lookahead data can be multiple, such as
The data source of economic data be statistics part, employer's organization and/or customs database, the data source of precipitation is meteorology portion
The database of door, the data source of temperature is the sensor etc. being arranged in particular sample point, likewise, also not right in the present embodiment
The type of data source does any restriction.In the present embodiment, set X includes the N groups lookahead data vector of user's selection, example
Such as, if the lookahead data selected is monthly mean rainfall and monthly mean temperature, N=2 (under normal circumstances, the values of N
The order of magnitude is 100-1000), set X includes X1(monthly mean rainfall) and X2(monthly mean temperature) totally two lookahead datas to
Amount, the i.e. value of i are 1 and 2.It is vectorial (or subvector of electricity consumption vector) that electricity consumption is also stored in database, vector
XiTiming statistics cycle phase with E is the same as (such as the timing statistics period is " moon "), vectorial XiEach of value x when being corresponding statistics
Between lookahead data value in the period, value e is the electricity consumption in the corresponding timing statistics period each of in vectorial E.Further,
The subscript of vector indicates the number in the timing statistics period of range prediction electricity consumption, in the feelings that the timing statistics period is " moon "
Under condition, if to predict the electricity consumption of certain year September part, vector X1Or X2Subscript (- 1) indicate the elder generation of this year August part
Row data (such as average precipitation or mean temperature of August part), the subscript (- 1) of vectorial E indicate the use of concept August part
Electricity data, and so on, (- ti) and (- te) indicate tiAnd teLookahead data between a month.
In the present embodiment, interface equipment and database can be implemented as software, can also be embodied as hardware, existing skill
It is any in art to realize that the equipment of corresponding function or software module can be applied in the present invention.
In the present embodiment, in step S1, the vector XiMeasurement period with E is the moon.
In the present embodiment, in step S1, the lookahead data is the number after being modified to the data in data source
According to the revised data are the number of days in data divided by this month before correcting;The electricity consumption is to be repaiied to practical electricity consumption
Electricity consumption after just, the revised electricity consumption are electricity consumption divided by the number of days in this month before correcting.By being repaiied to data
Just, enable to lookahead data and electricity consumption data more accurate, when measurement period is " moon ", on the one hand, day monthly
Number differed for 28 days to 31 days, and in the similar data such as GDP or steel monthly output of processing, revised data can evade
The error that monthly number of days is brought;On the other hand, when being predicted for (- 1) moon (i.e. the previous moon of predicted month), can avoid compared with
Big error, for example, wishing to predict the electricity consumption data of September part when August 10 days, then the data of August (i.e.-January) only have
First 10 days electricity consumption data, it is clear that there are large errors, this is the number of days (10) for using the data/this month for correcting first 10 days,
I.e. average daily electricity consumption data replace of that month total electricity consumption data, can effectively avoid the generation of error.
In the present embodiment, in the step S3, optimal leading time arrow T is obtained using K-L information Contents Methods.
In the present embodiment, the step S3 specifically includes following steps:
Step S31:Electricity consumption change rate vector is calculated according to electricity consumption vector E
Wherein,The value range of m is 1 to te-1;Obviously, the quantity ratio of the electricity consumption change rate in Δ
The quantity of electricity consumption in corresponding E is 1 few.When measurement period is " moon ", preferred te=13, that is, provide one completely
Electricity consumption change rate in annual period.
Step S32:For each parameter vector P in parameter vector set Pi, extract PiIn teA (such as it is aforementioned excellent
13 chosen) subvector, there is in each subvector teThe collection of -1 parameter, the subvector is combined into Pisub={ Pi1,
Pi2,...,Pi(te), any subvector PijForThe value range of j is 1 to te;
Step S33:Calculate the related coefficient of each subvector of electricity consumption change rate vector sumIts
Middle Cov () is covariance function, and σ () is variance function, and the value range of j is 1 to te;Obviously, in teWhen=13, Neng Gouji
It calculates and obtains 13 correlation coefficient ρs1,ρ2,...,ρ13。
Step S34:Obtain ρijIn maximum value ρimaxWith the j values of the correspondence maximum value, and it is denoted as jmax;
Step S35:By jmaxAs parameter vector PiIn look ahead time T corresponding to optimal leading parameteri, i.e. Ti=
jmax。
In the present embodiment, the step S34 further includes:If ρimaxMore than or equal to predetermined threshold value (such as 0.6),
Illustrate parameter vector PiIt is higher with the degree of correlation of Δ, then executing step S35;Otherwise, illustrate parameter vector PiIt is related to Δ
It spends relatively low, no longer executes step S35, and by the ρimaxCorresponding PiIt is rejected from set P.
Particularly, in step s3, also optimal leading time arrow T is saved in database, i.e., will predicts institute every time
The optimal leading time arrow T obtained is calculated to be saved in database.Further, it is " moon " in the timing statistics period
In preferred embodiment, optimal leading time arrow T is preserved according to month in database.The step S3 further comprises:
Step S36:Upper one month optimal leading time arrow T is extracted from databasem=(Tm1,Tm2,...,TmN),
And the optimal leading time arrow T of this month of upper one yeary=(Ty1,Ty2,...,TyN);For example, it is desired to which prediction is 2018 1
The electricity consumption data of the moon, the then optimal look ahead time used when extraction prediction in December, 2017 electricity consumption data from database
Vector is used as Tm, the optimal leading time arrow used when prediction in January, 2017 electricity consumption data is extracted as Ty;
Step S37:For optimal leading time arrow T=(T1,T2,...,TN) in any optimal look ahead time Ti,
If | Ti-Tmi| > D1, and, | Ti-Tyi| > D1So to user's early warning TiValue may be abnormal, wherein D1For predetermined threshold value, example
Such as D1=2 months.
Step S38:For optimal leading time arrow T=(T1,T2,...,TN) in any optimal look ahead time Ti,
IfAndSo may be whole abnormal to user's pre-warning time vector T,
Wherein D2For threshold, such as D2=0.4.
By step S36 to step S38, data in optimal leading time arrow T that can be to calculating acquisition in real time into
Row verification, user is notified when data generate abnormal in T in time, avoids making pre- since the exception of data in T is not detectable
Measured value generates large error.
Through the above steps, each optimal look ahead time in optimal leading time arrow T can be obtained.
In the present embodiment, in step S4, the interconnection vector K=(k1,k2,...,ki,...,kN) acquisition by phase
The ratio of relationship number is worth to, i.e.,:
In the present embodiment, in step S4, the interconnection vector K=(k1,k2,...,ki,...,kN) acquisition be profit
Equation Z=KQ is parsed with least square method, wherein:
Z=(δ-1,δ-2,...,δ-M)T;
Wherein, M < N.When using the moon as measurement period, it is preferred that M values are 12.
Particularly, in the present embodiment, te=13.
The present embodiment has following advantages:
1, real-time, i.e., whenever predicting the electricity consumption of next time cycle, the present invention is performed both by step S1 to step
Rapid S5 also avoids the error caused by real-time reason to ensure that real-time.
2, fully take into account the big metastable feature of region electricity consumption, electricity consumption is regarded as including fundamental invariant and
Both variables use electricity consumption change rate indirect predictions electricity consumption so that the error of electricity consumption change rate is similar to electricity consumption
Error, substantially avoided and interfere caused by invariant.For example, in certain group contrast test prediction, carried out using the prior art
Electricity consumption error is 3.2%;Using the present invention, the predicted value of electricity consumption change rate is 2.1%, actual value 2.28%,
The prediction error of electricity consumption change rate is 7.9%, but the electricity demand forecasting value calculated in step S5 and actual electricity consumption it
Between error be 0.18%, far below in the prior art 3.2% prediction error.
The foregoing is merely presently preferred embodiments of the present invention, all equivalent changes done according to scope of the present invention patent
With modification, it should all belong to the covering scope of the present invention.
Claims (9)
1. a kind of big region electricity demand forecasting method based on big data, it is characterised in that:Include the following steps:
Step S1:Acquire the set X={ X of the relevant N groups lookahead data vector of big region electricity consumption1,X2,...,XN, Mei Gexiang
Measure Xi=(xi(-ti),xi(-ti+1),...,xi(-1)), the value range of i is from 1 to N;Acquire electricity consumption vector E=(e-te,
e-te+1,...,e-1);The vector XiIt is identical as the measurement period of E, the vector XiEach of value be corresponding timing statistics week
Lookahead data value in phase, value is the electricity consumption in the corresponding timing statistics period each of in the vector E;
Step S2:Set P={ the P of parameter vector are obtained according to set X1,P2,...Pi,...,PN, each parameter vectorWherein,The value range of m is
1 arrives ti-1;
Step S3:According to the set P of electricity consumption vector E and parameter vector, the optimal leading time arrow T=of parameter vector is obtained
(T1,T2,...Ti,...,TN), wherein TiFor parameter vector PiIn look ahead time corresponding to optimal leading parameter;
Step S4:Obtain the interconnection vector K=(k between optimal leading parameter and electricity consumption change rate1,k2,...,ki...,kN);
Step S5:Obtain the prediction electricity consumption in big region in next time cycleWherein,
e-1For the practical electricity consumption in big region in current time period.
2. a kind of big region electricity demand forecasting method based on big data according to claim 1, it is characterised in that:Step
In S1, the vector XiMeasurement period with E is the moon.
3. a kind of big region electricity demand forecasting method based on big data according to claim 1, it is characterised in that:Step
In S1, the lookahead data is the data after being modified to the data in data source, and the revised data are before correcting
The number of days in data divided by this month;The electricity consumption is the electricity consumption after being modified to practical electricity consumption, the revised use
Electricity is electricity consumption divided by the number of days in this month before correcting.
4. a kind of big region electricity demand forecasting method based on big data according to claim 1, it is characterised in that:It is described
In step S3, optimal leading time arrow T is obtained using K-L information Contents Methods.
5. a kind of big region electricity demand forecasting method based on big data according to claim 1, it is characterised in that:It is described
Step S3 specifically includes following steps:
Step S31:Electricity consumption change rate vector is calculated according to electricity consumption vector EIts
In,The value range of m is 1 to te-1;
Step S32:For each parameter vector P in parameter vector set Pi, extract PiIn teA subvector, per height to
There is t in amounteThe collection of -1 parameter, the subvector is combined intoAny subvector PijForThe value range of j is 1 to te;
Step S33:Calculate the related coefficient of each subvector of electricity consumption change rate vector sumWherein Cov
() is covariance function, and σ () is variance function, and the value range of j is 1 to te;
Step S34:Obtain ρijIn maximum value ρimaxWith the j values of the correspondence maximum value, and it is denoted as jmax;
Step S35:By jmaxAs parameter vector PiIn look ahead time T corresponding to optimal leading parameteri, i.e. Ti=jmax。
6. a kind of big region electricity demand forecasting method based on big data according to claim 5, it is characterised in that:It is described
Step S34 further includes:If ρimaxMore than or equal to predetermined threshold value, then executing step S35;Otherwise, step is no longer executed
S35, and by the ρimaxCorresponding PiIt is rejected from set P.
7. a kind of big region electricity demand forecasting method based on big data according to claim 5, it is characterised in that:Step
In S4, the interconnection vector K=(k1,k2,...,ki,...,kN) acquisition be worth to by the ratio of related coefficient, i.e.,:
8. a kind of big region electricity demand forecasting method based on big data according to claim 1, it is characterised in that:Step
In S4, the interconnection vector K=(k1,k2,...,ki,...,kN) acquisition be using least square method parse equation Z=KQ,
Wherein:
Z=(δ-1,δ-2,...,δ-M)T;
Wherein, M < N.
9. a kind of big region electricity demand forecasting system based on big data, it is characterised in that:Including predictive server, interface equipment
With database;
The interface equipment is leading to obtain the relevant N groups of big region electricity consumption for being connected to more than one data source
The set of data vector;The set of electricity consumption vector E and lookahead data vector are stored in the database;
The predictive server includes processor and storage medium, and the storage medium is stored with claim 1-8 any bars institute
The computer program for stating method, whenever predicting the electricity consumption of next time cycle, the processor executes above computer
Program.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810462414.4A CN108665108A (en) | 2018-05-15 | 2018-05-15 | A kind of big region electricity demand forecasting method and system based on big data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810462414.4A CN108665108A (en) | 2018-05-15 | 2018-05-15 | A kind of big region electricity demand forecasting method and system based on big data |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108665108A true CN108665108A (en) | 2018-10-16 |
Family
ID=63779671
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810462414.4A Pending CN108665108A (en) | 2018-05-15 | 2018-05-15 | A kind of big region electricity demand forecasting method and system based on big data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108665108A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110009165A (en) * | 2019-05-08 | 2019-07-12 | 国网福建省电力有限公司经济技术研究院 | Prediction technique and system based on multiaddress fence area history electricity consumption |
CN113449919A (en) * | 2021-06-29 | 2021-09-28 | 国网山东省电力公司菏泽供电公司 | Power consumption prediction method and system based on feature and trend perception |
CN115392803A (en) * | 2022-10-28 | 2022-11-25 | 北京国电通网络技术有限公司 | Method and apparatus for adjusting power supply amount in area, electronic device, and medium |
-
2018
- 2018-05-15 CN CN201810462414.4A patent/CN108665108A/en active Pending
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110009165A (en) * | 2019-05-08 | 2019-07-12 | 国网福建省电力有限公司经济技术研究院 | Prediction technique and system based on multiaddress fence area history electricity consumption |
CN113449919A (en) * | 2021-06-29 | 2021-09-28 | 国网山东省电力公司菏泽供电公司 | Power consumption prediction method and system based on feature and trend perception |
CN115392803A (en) * | 2022-10-28 | 2022-11-25 | 北京国电通网络技术有限公司 | Method and apparatus for adjusting power supply amount in area, electronic device, and medium |
CN115392803B (en) * | 2022-10-28 | 2022-12-23 | 北京国电通网络技术有限公司 | Method, device, electronic device, and medium for adjusting power supply amount in area |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Ali et al. | Review of urban building energy modeling (UBEM) approaches, methods and tools using qualitative and quantitative analysis | |
Naser | Estimating and forecasting the real prices of crude oil: A data rich model using a dynamic model averaging (DMA) approach | |
Han et al. | Drought forecasting based on the remote sensing data using ARIMA models | |
Charnavoki et al. | The effects of global shocks on small commodity-exporting economies: lessons from Canada | |
Shamshad et al. | First and second order Markov chain models for synthetic generation of wind speed time series | |
Juárez et al. | Model-based clustering of non-Gaussian panel data based on skew-t distributions | |
Cuestas et al. | Oil price shocks and unemployment in Central and Eastern Europe | |
Corsi et al. | Threshold bipower variation and the impact of jumps on volatility forecasting | |
AU2001255994B8 (en) | Method of Business Analysis | |
CN103383445A (en) | System and method for forecasting service life and reliability of intelligent electric meter | |
CN108665108A (en) | A kind of big region electricity demand forecasting method and system based on big data | |
CN110457867A (en) | A kind of time series data based on machine learning is filled up and restoring method | |
Ramanarayanan | Imported inputs, irreversibility, and international trade dynamics | |
Zheng et al. | Statistical regularities of Carbon emission trading market: Evidence from European Union allowances | |
CN109492783A (en) | A kind of Application of Power Metering Instruments failure risk prediction technique based on GBDT | |
CN105825286A (en) | System and method of estimating full life cycle cost of weapon equipment | |
Leandro et al. | Forecasting upper and lower uncertainty bands of river flood discharges with high predictive skill | |
Arsenault et al. | An efficient method to correct under-dispersion in ensemble streamflow prediction of inflow volumes for reservoir optimization | |
CN102156641A (en) | Prediction method and system for confidence interval of software cost | |
Karanasos et al. | On the order of integration of monthly US ex-ante and ex-post real interest rates: New evidence from over a century of data | |
CN111178605A (en) | Distribution network engineering project construction period prediction method based on feature selection | |
Zhao et al. | Evaluating influences of the Manwan Dam and climate variability on the hydrology of the Lancang-Mekong River, Yunnan Province, southwest China | |
Tang et al. | Comparison of information-based measures of forecast uncertainty in ensemble ENSO prediction | |
CN112184495B (en) | Stock low-efficiency land monitoring system and analysis platform using same | |
Ignatiadis et al. | Forecasting residential monthly electricity consumption using smart meter data |
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: 20181016 |
|
RJ01 | Rejection of invention patent application after publication |