CN105631483A - Method and device for predicting short-term power load - Google Patents
Method and device for predicting short-term power load Download PDFInfo
- Publication number
- CN105631483A CN105631483A CN201610130474.7A CN201610130474A CN105631483A CN 105631483 A CN105631483 A CN 105631483A CN 201610130474 A CN201610130474 A CN 201610130474A CN 105631483 A CN105631483 A CN 105631483A
- Authority
- CN
- China
- Prior art keywords
- historical data
- power load
- data
- load
- weather information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
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
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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"
Abstract
The invention provides a method and a device for predicting short-term power load, which relate to the technical field of power load prediction. The method comprises steps: historical data of the power load and historical data of weather information are acquired, and fuzzy cluster analysis pretreatment is carried out; according to a fuzzy cluster analysis algorithm, the historical data of the power load after pretreatment are subjected to fuzzy cluster, fuzzy cluster parameters are adjusted, and the optimal classification for the historical data of the power load is acquired; according to a sliding window, historical training data of the power load are acquired from the historical data of the power load in the optimal classification, and the historical training data of the power load and the historical training data of the weather information are acquired; and according to a preset prediction algorithm, the historical training data of the power load and the historical training data of the weather information are treated to acquire the short-term power load prediction data. The problems that the current short-term power load method is poor in prediction precision and tedious and complicated in treatment can be solved.
Description
Technical field
The present invention relates to Electric Load Forecasting survey technology field, particularly relate to a kind of short-term electro-load forecast method and device.
Background technology
Currently, owing to electric energy can not be stored in a large number, power system needs to keep at any time the equilibrium of supply and demand. For ensureing the safety of power system, it is necessary to grasp the Changing Pattern of load, and the variation tendency in future, namely need power load effectively to be predicted. Load prediction is the important component part of Power System Planning, is again one of important factor of improving electric power enterprise economic benefit, promoting national economic development. And short-term electro-load forecast will refer to several hours, 1 day future, until the load prediction of a week. It is basis essential in economy operation of power grid and safety control, the security of operation of power networks, reliability and economy is played an important role. The short-term electro-load forecast of high precision contributes to reasonably arranging grid equipment scheduling and turnaround plans, it is to increase the stability of Operation of Electric Systems, reduces the cost of electricity-generating of electrical network, it is to increase the economic benefit of power system and social benefit.
The outstanding feature of short-term power load take day as the similarity cycle presenting change, and obviously by the impact of weather conditions. Therefore to realize effective short-term power consumption prediction, it is necessary to fully research load variations rule, the relation of analysis load changed factor, particularly weather conditions and the change of short-term power load. At present, for short-term electro-load forecast, mainly in the following way: one, time series forecasting method; Two, linear regression method; Three, neural network prediction method; Four, gray evaluation. Wherein, time series forecasting method is most widely used in power industry, develops more ripe method, load data is regarded as the time series by week, sky and hours period change by it, according to historical summary, set up the statistics rule that a mathematical model describes this stochastic variable of power load, the power load in future is forecast. But the existence of time series forecasting method is subject to the interference of noise data, prediction precision increases and the problems such as reduction with step-length. Linear regression method is by the observed data of variable is carried out statistical study, it is determined that the correlationship between variable, thus realizes the object of prediction. But due to load prediction input and output are non-linear and model lacks self-learning capability and prediction accuracy is not high, in causing linear regression method to be suitable for, long term load forecasting. Neural network prediction method inputs artificial neural network using power load affects in historical data maximum several factors as input, in input layer, hidden layer and output layer, each neuronic effect finally generates work output, carrying out network weight constantly revising until error reaches requirement to export error as objective function again, the network after training just can predict the outcome. Neural network prediction method demand data amount is big, speed of convergence slow, and lacks a kind of effective method and solve the problem such as over-fitting and poor fitting in training process. Gray evaluation be undertaken the historical data changed in certain limit adding up, regressive or level be than generating so that it is become the raised shape with index increasing law, then the ordered series of numbers differential equation generated set up grey model. Do not need during gray evaluation modeling to calculate statistical characteristic value, have desired data amount few, variation tendency need not be considered, computing is convenient, be easy to the features such as inspection, but there is the problem poor for the data prediction precision that dispersion degree is bigger.
Visible, it is poor to there is prediction precision in the method for current short-term power load, and processes the problem of comparatively very complicated.
Summary of the invention
Embodiments of the invention provide a kind of short-term electro-load forecast method and device, poor to solve the method existence prediction precision of current short-term power load, and process the problem of comparatively very complicated.
For achieving the above object, the present invention adopts following technical scheme:
A kind of short-term electro-load forecast method, comprising:
Obtain the historical data of Weather information corresponding to the historical data of the historical data of power load and described power load;
The historical data of described power load and the historical data of described Weather information are carried out fuzzy clustering analysis pre-treatment;
According to Fuzzy Cluster Analysis Algorithm, the historical data of pretreated power load being carried out fuzzy clustering, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load; The historical data of the power load in described optimal classification and the deviation of fuzzy clustering central value are all less than predetermined threshold value, and the historical data number of the power load in described optimal classification is greater than the historical data number of the power load carried out in other classification that fuzzy clustering produces;
The historical data of the power load of the moving window according to a preset length in described optimal classification obtains the history training data of power load, and the historical data of Weather information after the pre-treatment obtains the Weather information history training data of the history training data of described power load;
According to the prediction algorithm pre-set, history training data and Weather information history training data to described power load process, and obtain short-term electro-load forecast data.
Concrete, the historical data of described Weather information comprises: temperature historical data, wind speed historical data.
Concrete, the historical data of described power load and the historical data of described Weather information are carried out fuzzy clustering analysis pre-treatment, comprising:
Negative value in the historical data of power load is set to 0.
Concrete, the historical data of described power load and the historical data of described Weather information are carried out fuzzy clustering analysis pre-treatment, comprising:
The missing data in the historical data of power load and the historical data of Weather information is determined according to trend prediction model;
Described trend prediction model is: xk=ak-t+bk-t�� t;
Wherein, k is current time sequence number; T is the moment quantity of missing data before the k moment; xkFor the historical data of power load or the historical data of Weather information in k moment; For Single moving average value;For Double moving average value; N is the length of each moving average.
Concrete, according to Fuzzy Cluster Analysis Algorithm, the historical data of pretreated power load being carried out fuzzy clustering, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load, comprising:
Historical data according to pretreated power load generates the historical data vector X of power load; The data amount check of the historical data vector X of described power load is N, treats that grouping number is C;
Determine the fuzzy grouping matrix U of the historical data vector X of power load;
Wherein, ��ijIt is i-th data xiIt is under the jurisdiction of the degree of being subordinate to of jth grouping; And ��ij�� [0,1]; 1��i��N; 1��j��C;The cost function of described fuzzy grouping matrix U isWherein m is weights; DijIt is i-th data xiWith the feature v of jth groupingjBetween weighted euclidean distance;A is the variance matrix of the historical data vector X of power load;CjFor jth grouping; NjFor the data amount check of jth grouping;
Optimum degree of being subordinate to and the optimal group feature of the historical data vector X of described power load is determined according to fuzzy grouping matrix U;
Wherein, described optimum degree of being subordinate to is
Described optimal group is characterized as
Determine optimum degree of being subordinate to and optimal group feature corresponding be grouped into described optimal classification.
In addition, the prediction algorithm pre-set described in comprises: time series forecasting method, linear regression method, neural network prediction method, gray evaluation; Described time series forecasting method comprises arma modeling algorithm.
A kind of short-term electro-load forecast device, comprising:
Historical data acquiring unit, for the historical data of Weather information corresponding to the historical data of the historical data and described power load that obtain power load;
Pretreatment unit, for carrying out fuzzy clustering analysis pre-treatment to the historical data of described power load and the historical data of described Weather information;
Fuzzy clustering unit, for the historical data of pretreated power load being carried out fuzzy clustering according to Fuzzy Cluster Analysis Algorithm, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load; The historical data of the power load in described optimal classification and the deviation of fuzzy clustering central value are all less than predetermined threshold value, and the historical data number of the power load in described optimal classification is greater than the historical data number of the power load carried out in other classification that fuzzy clustering produces;
Sliding window data acquiring unit, historical data for the power load of the moving window according to a preset length in described optimal classification obtains the history training data of power load, and the historical data of Weather information after the pre-treatment obtains the Weather information history training data of the history training data of described power load;
Short-term electro-load forecast unit, for according to the prediction algorithm pre-set, history training data and Weather information history training data to described power load process, obtains short-term electro-load forecast data.
Concrete, the historical data of the Weather information that described historical data acquiring unit obtains comprises: temperature historical data, wind speed historical data.
In addition, described pretreatment unit, specifically for:
Negative value in the historical data of power load is set to 0.
Further, described pretreatment unit, also for:
The missing data in the historical data of power load and the historical data of Weather information is determined according to trend prediction model;
Described trend prediction model is: xk=ak-t+bk-t�� t;
Wherein, k is current time sequence number; T is the moment quantity of missing data before the k moment; xkFor the historical data of power load or the historical data of Weather information in k moment; For Single moving average value;For Double moving average value; N is the length of each moving average.
Concrete, described fuzzy clustering unit, comprising:
Historical data vector generation module, generates the historical data vector X of power load for the historical data according to pretreated power load; The data amount check of the historical data vector X of described power load is N, treats that grouping number is C;
Fuzzy grouping matrix determination module, for determining the fuzzy grouping matrix U of the historical data vector X of power load;
Wherein, ��ijIt is i-th data xiIt is under the jurisdiction of the degree of being subordinate to of jth grouping; And ��ij�� [0,1]; 1��i��N; 1��j��C;The cost function of described fuzzy grouping matrix U isWherein m is weights; DijIt is i-th data xiWith the feature v of jth groupingjBetween weighted euclidean distance;A is the variance matrix of the historical data vector X of power load;CjFor jth grouping; NjFor the data amount check of jth grouping;
Optimal classification determination module, for determining the historical data vector optimum degree of being subordinate to of X and the optimal group feature of described power load according to fuzzy grouping matrix U, and determine optimum degree of being subordinate to and optimal group feature corresponding be grouped into described optimal classification;
Wherein, described optimum degree of being subordinate to is
Described optimal group is characterized as
Concrete, the prediction algorithm pre-set of described short-term electro-load forecast unit application comprises: time series forecasting method, linear regression method, neural network prediction method, gray evaluation; Described time series forecasting method comprises arma modeling algorithm.
A kind of short-term electro-load forecast method that the embodiment of the present invention provides and device, it is possible to the historical data of the power load got and the historical data of Weather information are carried out fuzzy clustering analysis pre-treatment; And according to Fuzzy Cluster Analysis Algorithm, the historical data of pretreated power load being carried out fuzzy clustering, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load; And then historical data according to power load in optimal classification of the moving window of a preset length obtains the history training data of power load, it is thus possible to according to the prediction algorithm pre-set, history training data and Weather information history training data to power load process, and obtain short-term electro-load forecast data. The data of the optimal classification that the present invention is obtained by Fuzzy Cluster Analysis Algorithm, excavate to the full extent and are used for electricity consumption rule, eliminate the impact of a small amount of deflection loads curve, it is to increase the precision of short-term electro-load forecast.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, it is briefly described to the accompanying drawing used required in embodiment or description of the prior art below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings.
The schema one of a kind of short-term electro-load forecast method that Fig. 1 provides for the embodiment of the present invention;
The flowchart 2 of a kind of short-term electro-load forecast method that Fig. 2 provides for the embodiment of the present invention;
Fig. 3 is the schematic diagram predicted the outcome by the matching of Fuzzy C-means cluster algorithm and ARMA algorithm in the embodiment of the present invention;
The structural representation one of a kind of short-term electro-load forecast device that Fig. 4 provides for the embodiment of the present invention;
The structural representation two of a kind of short-term electro-load forecast device that Fig. 5 provides for the embodiment of the present invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is only the present invention's part embodiment, instead of whole embodiments. Based on the embodiment in the present invention, those of ordinary skill in the art, not making other embodiments all obtained under creative work prerequisite, belong to the scope of protection of the invention.
As shown in Figure 1, the embodiment of the present invention provides a kind of short-term electro-load forecast method, comprising:
The historical data of the Weather information that the historical data of step 101, the historical data obtaining power load and power load is corresponding.
Step 102, the historical data of power load and the historical data of Weather information are carried out fuzzy clustering analysis pre-treatment.
Step 103, according to Fuzzy Cluster Analysis Algorithm, the historical data of pretreated power load being carried out fuzzy clustering, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load.
Wherein, the historical data of the power load in optimal classification and the deviation of fuzzy clustering central value are all less than predetermined threshold value, and the historical data number of the power load in optimal classification is greater than the historical data number of the power load carried out in other classification that fuzzy clustering produces.
Step 104, historical data according to power load in optimal classification of the moving window of a preset length obtain the history training data of power load, and the historical data of Weather information after the pre-treatment obtains the Weather information history training data of the history training data of power load.
The prediction algorithm that step 105, basis pre-set, history training data and Weather information history training data to power load process, and obtain short-term electro-load forecast data.
A kind of short-term electro-load forecast method that the embodiment of the present invention provides, it is possible to the historical data of the power load got and the historical data of Weather information are carried out fuzzy clustering analysis pre-treatment; And according to Fuzzy Cluster Analysis Algorithm, the historical data of pretreated power load being carried out fuzzy clustering, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load; And then historical data according to power load in optimal classification of the moving window of a preset length obtains the history training data of power load, it is thus possible to according to the prediction algorithm pre-set, history training data and Weather information history training data to power load process, and obtain short-term electro-load forecast data. The data of the optimal classification that the present invention is obtained by Fuzzy Cluster Analysis Algorithm, excavate to the full extent and are used for electricity consumption rule, eliminate the impact of a small amount of deflection loads curve, it is to increase the precision of short-term electro-load forecast.
In order to make the technician of this area better understand the present invention, enumerating an embodiment specifically below, as shown in Figure 2, a kind of short-term electro-load forecast method that the embodiment of the present invention provides, comprising:
The historical data of the Weather information that the historical data of step 201, the historical data obtaining power load and power load is corresponding.
Wherein, the historical data of Weather information comprises: temperature historical data, wind speed historical data. This temperature historical data can be the maximum temperature in the middle of one day or minimum temperature. In order to ensure that follow-up learning sample is enough big, the load data in 24 hours integral point moment of every day in 3 months can be gathered in embodiments of the present invention, and the air speed data (m/s) in 24 hours integral point moment of every day, temperature record (DEG C) etc.
Step 202, the negative value in the historical data of power load is set to 0, and determines the missing data in the historical data of power load and the historical data of Weather information according to trend prediction model.
Owing to the acquisition mode of power load data is determined by the difference of adjacent two moment ammeter registrations, if power load data are negative value, namely determine that it ammeter reversing phenomenon occurs, introduce in negative value to follow-up calculating and can cause the problems such as result is inaccurate, it is necessary to the negative value in the historical data of power load is set to 0.
And the electricity consumption trend of user can be analyzed by trend prediction model, and analyze the trend etc. of Weather information.
Wherein, trend prediction model is: xk=ak-t+bk-t��t��
Wherein, k is current time sequence number; T is the moment quantity of missing data before the k moment; xkFor the historical data of power load or the historical data of Weather information in k moment; For Single moving average value;For Double moving average value; N is the length of each moving average.
Herein, after step 202 processes, it is possible to pretreated data are carried out sample analysis, sample set is formed. Sample analysis herein refers to the means utilizing statistics, and as sought expectation and variance etc., the statistical character presented for data is understood, as being uniformly distributed, Gaussian distribution etc. Again according to its statistical character, carry out the division of sample set. Concrete sample analysis mode repeats no more herein.
Step 203, the historical data vector X generating power load according to the historical data of pretreated power load.
Herein, the data amount check of the historical data vector X of power load is N, treats that grouping number is C. Wherein, treat that namely grouping number belongs in fuzzy clustering parameter.
The fuzzy clustering mode can chosen in embodiments of the present invention has a lot, Fuzzy C-means cluster the algorithm that such as subsequent step adopts, but not only it is confined to this, also can complete the historical data to pretreated power load by other fuzzy clustering modes and carry out fuzzy clustering, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load.
The fuzzy grouping matrix U of step 204, the historical data determining power load vector X.
Wherein, ��ijIt is i-th data xiIt is under the jurisdiction of the degree of being subordinate to of jth grouping; And ��ij�� [0,1]; 1��i��N; 1��j��C;The cost function of fuzzy grouping matrix U isWherein m is weights, if when these weights are 1, then represent be grouped into rigid grouping, and if these weights are close to infinity, then represent be grouped into fuzzy grouping; DijIt is i-th data xiWith the feature v of jth groupingjBetween weighted euclidean distance;A is the variance matrix of the historical data vector X of power load;CjFor jth grouping; NjFor the data amount check of jth grouping.
Herein, the electric power load fuzzy clustering result acquired can be as shown in table 1 below:
Table 1:
The optimum degree of being subordinate to of step 205, the historical data determining power load according to fuzzy grouping matrix U vector X and optimal group feature.
Wherein, optimum degree of being subordinate to is
Optimal group is characterized as
Step 206, determine optimum degree of being subordinate to and optimal group feature is corresponding is grouped into optimal classification.
Wherein, the historical data of the power load in optimal classification and the deviation of fuzzy clustering central value are all less than predetermined threshold value, and the historical data number of the power load in optimal classification is greater than the historical data number of the power load carried out in other classification that fuzzy clustering produces. Such as, in above-mentioned table 1, it is possible to select the data of No. 2 fuzzy clusterings.
Step 207, historical data according to power load in optimal classification of the moving window of a preset length obtain the history training data of power load, and the historical data of Weather information after the pre-treatment obtains the Weather information history training data of the history training data of power load.
Such as, described preset length can be 30 days, but is not only confined to this.
The prediction algorithm that step 208, basis pre-set, history training data and Weather information history training data to power load process, and obtain short-term electro-load forecast data.
What deserves to be explained is, the prediction algorithm pre-set can be: time series forecasting method, linear regression method, neural network prediction method, gray evaluation; Time series forecasting method comprises arma modeling algorithm.
Herein, it is described to obtain short-term electro-load forecast data instance according to arma modeling algorithm;
Wherein, arma modeling is:
yt=��0+��1yt-1+��2yt-2+...+��pyt-p+Zt; Zt=��t+��1��t-1+��2��t-2+...+��q��t-q��
If t power load charge values is yt, the load value y in it and t-p momentt-pRelation degree be ��p, ZtFor the impact of other factors, be here every day maximum temperature, minimum temperature, the meteorological factor such as wind speed. ��tFor the independent identically distributed stochastic variable sequence of t.
In addition, can be determined in the average relative error of moving window in 30 days by ARMA algorithm.
In embodiments of the present invention, utilizing the matching of Fuzzy C-means cluster algorithm and ARMA algorithm to predict the outcome as shown in Figure 3, data prediction error is as shown in table 2 below. Predicting the outcome of finally obtaining is as shown in table 3.
Table 2:
Wherein, RMSE is root-mean-square error; MAPE is average absolute percentage error; MaxAPE is maximum absolute percentage error; MAE is mean absolute error; MaxAE is maximum mean absolute error.
Table 3:
Wherein, UCL is the supremum of 95% fiducial interval; LCL is the infimum of 95% fiducial interval.
A kind of short-term electro-load forecast method that the embodiment of the present invention provides, it is possible to the historical data of the power load got and the historical data of Weather information are carried out fuzzy clustering analysis pre-treatment; And according to Fuzzy Cluster Analysis Algorithm, the historical data of pretreated power load being carried out fuzzy clustering, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load; And then historical data according to power load in optimal classification of the moving window of a preset length obtains the history training data of power load, it is thus possible to according to the prediction algorithm pre-set, history training data and Weather information history training data to power load process, and obtain short-term electro-load forecast data. The data of the optimal classification that the present invention is obtained by Fuzzy Cluster Analysis Algorithm, excavate to the full extent and are used for electricity consumption rule, eliminate the impact of a small amount of deflection loads curve, it is to increase the precision of short-term electro-load forecast.
Corresponding to the embodiment of the method shown in above-mentioned Fig. 1 and Fig. 2, as shown in Figure 4, the embodiment of the present invention provides a kind of short-term electro-load forecast device, comprising:
Historical data acquiring unit 31, for the historical data of Weather information corresponding to the historical data of the historical data and described power load that obtain power load.
Pretreatment unit 32, for carrying out fuzzy clustering analysis pre-treatment to the historical data of described power load and the historical data of described Weather information.
Fuzzy clustering unit 33, for the historical data of pretreated power load being carried out fuzzy clustering according to Fuzzy Cluster Analysis Algorithm, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load; The historical data of the power load in described optimal classification and the deviation of fuzzy clustering central value are all less than predetermined threshold value, and the historical data number of the power load in described optimal classification is greater than the historical data number of the power load carried out in other classification that fuzzy clustering produces.
Sliding window data acquiring unit 34, historical data for the power load of the moving window according to a preset length in described optimal classification obtains the history training data of power load, and the historical data of Weather information after the pre-treatment obtains the Weather information history training data of the history training data of described power load.
Short-term electro-load forecast unit 35, for according to the prediction algorithm pre-set, history training data and Weather information history training data to described power load process, obtains short-term electro-load forecast data.
Concrete, the historical data of the Weather information that this historical data acquiring unit 31 obtains comprises: temperature historical data, wind speed historical data.
In addition, this pretreatment unit 32, specifically can be set to 0 by the negative value in the historical data of power load.
Further, this pretreatment unit 32, it is also possible to determine the missing data in the historical data of power load and the historical data of Weather information according to trend prediction model.
This trend prediction model is: xk=ak-t+bk-t�� t;
Wherein, k is current time sequence number; T is the moment quantity of missing data before the k moment; xkFor the historical data of power load or the historical data of Weather information in k moment; For Single moving average value;For Double moving average value; N is the length of each moving average.
Concrete, as shown in Figure 5, this fuzzy clustering unit 33, it is possible to comprising:
Historical data vector generation module 331, generates the historical data vector X of power load for the historical data according to pretreated power load; The data amount check of the historical data vector X of this power load is N, treats that grouping number is C.
Fuzzy grouping matrix determination module 332, for determining the fuzzy grouping matrix U of the historical data vector X of power load.
Wherein, ��ijIt is i-th data xiIt is under the jurisdiction of the degree of being subordinate to of jth grouping; And �� ij�� [0,1]; 1��i��N; 1��j��C;The cost function of described fuzzy grouping matrix U isWherein m is weights; DijIt is i-th data xiWith the feature v of jth groupingjBetween weighted euclidean distance;A is the variance matrix of the historical data vector X of power load;CjFor jth grouping; NjFor the data amount check of jth grouping.
Optimal classification determination module 333, for determining the historical data vector optimum degree of being subordinate to of X and the optimal group feature of described power load according to fuzzy grouping matrix U, and determine optimum degree of being subordinate to and optimal group feature corresponding be grouped into described optimal classification.
Wherein, this optimum degree of being subordinate to is
This optimal group is characterized as
Concrete, the prediction algorithm pre-set that this short-term electro-load forecast unit 35 is applied can comprise: time series forecasting method, linear regression method, neural network prediction method, gray evaluation; Described time series forecasting method comprises arma modeling algorithm.
What deserves to be explained is, the specific implementation of a kind of short-term electro-load forecast device that the embodiment of the present invention provides see embodiment of the method corresponding to above-mentioned Fig. 1 and Fig. 2, can repeat no more herein.
A kind of short-term electro-load forecast device that the embodiment of the present invention provides, it is possible to the historical data of the power load got and the historical data of Weather information are carried out fuzzy clustering analysis pre-treatment; And according to Fuzzy Cluster Analysis Algorithm, the historical data of pretreated power load being carried out fuzzy clustering, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load; And then historical data according to power load in optimal classification of the moving window of a preset length obtains the history training data of power load, it is thus possible to according to the prediction algorithm pre-set, history training data and Weather information history training data to power load process, and obtain short-term electro-load forecast data. The data of the optimal classification that the present invention is obtained by Fuzzy Cluster Analysis Algorithm, excavate to the full extent and are used for electricity consumption rule, eliminate the impact of a small amount of deflection loads curve, it is to increase the precision of short-term electro-load forecast.
Those skilled in the art are it should be appreciated that embodiments of the invention can be provided as method, system or computer program. Therefore, the present invention can adopt the form of complete hardware embodiment, completely software implementation or the embodiment in conjunction with software and hardware aspect. And, the present invention can adopt the form at one or more upper computer program implemented of computer-usable storage medium (including but not limited to multiple head unit, CD-ROM, optical memory etc.) wherein including computer usable program code.
The present invention is that schema and/or skeleton diagram with reference to method according to embodiments of the present invention, equipment (system) and computer program describe. Should understand can by the combination of the flow process in each flow process in computer program instructions flowchart and/or skeleton diagram and/or square frame and schema and/or skeleton diagram and/or square frame. These computer program instructions can be provided to the treater of multi-purpose computer, special purpose computer, Embedded Processor or other programmable data processing device to produce a machine so that the instruction performed by the treater of computer or other programmable data processing device is produced for realizing the device of function specified in schema flow process or multiple flow process and/or skeleton diagram square frame or multiple square frame.
These computer program instructions also can be stored in and can guide in computer-readable memory that computer or other programmable data processing device work in a specific way, making the instruction that is stored in this computer-readable memory produce the manufacture comprising instruction device, this instruction device realizes the function specified in schema flow process or multiple flow process and/or skeleton diagram square frame or multiple square frame.
These computer program instructions also can be loaded in computer or other programmable data processing device, make on computer or other programmable devices, to perform a series of operation steps to produce computer implemented process, thus the instruction performed on computer or other programmable devices is provided for realizing the step of the function specified in schema flow process or multiple flow process and/or skeleton diagram square frame or multiple square frame.
The present invention applies specific embodiment the principle of the present invention and enforcement mode have been set forth, illustrating just for helping the method understanding the present invention and core concept thereof of above embodiment; Meanwhile, for one of ordinary skill in the art, according to the thought of the present invention, all will change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (12)
1. a short-term electro-load forecast method, it is characterised in that, comprising:
Obtain the historical data of Weather information corresponding to the historical data of the historical data of power load and described power load;
The historical data of described power load and the historical data of described Weather information are carried out fuzzy clustering analysis pre-treatment;
According to Fuzzy Cluster Analysis Algorithm, the historical data of pretreated power load being carried out fuzzy clustering, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load; The historical data of the power load in described optimal classification and the deviation of fuzzy clustering central value are all less than predetermined threshold value, and the historical data number of the power load in described optimal classification is greater than the historical data number of the power load carried out in other classification that fuzzy clustering produces;
The historical data of the power load of the moving window according to a preset length in described optimal classification obtains the history training data of power load, and the historical data of Weather information after the pre-treatment obtains the Weather information history training data of the history training data of described power load;
According to the prediction algorithm pre-set, history training data and Weather information history training data to described power load process, and obtain short-term electro-load forecast data.
2. short-term electro-load forecast method according to claim 1, it is characterised in that, the historical data of described Weather information comprises: temperature historical data, wind speed historical data.
3. short-term electro-load forecast method according to claim 2, it is characterised in that, the historical data of described power load and the historical data of described Weather information are carried out fuzzy clustering analysis pre-treatment, comprising:
Negative value in the historical data of power load is set to 0.
4. short-term electro-load forecast method according to claim 3, it is characterised in that, the historical data of described power load and the historical data of described Weather information are carried out fuzzy clustering analysis pre-treatment, comprising:
The missing data in the historical data of power load and the historical data of Weather information is determined according to trend prediction model;
Described trend prediction model is: xk=ak-t+bk-t�� t;
Wherein, k is current time sequence number; T is the moment quantity of missing data before the k moment; xkFor the historical data of power load or the historical data of Weather information in k moment; For Single moving average value;For Double moving average value; N is the length of each moving average.
5. short-term electro-load forecast method according to claim 4, it is characterized in that, according to Fuzzy Cluster Analysis Algorithm, the historical data of pretreated power load is carried out fuzzy clustering, adjustment fuzzy clustering parameter, obtain the optimal classification of the historical data of power load, comprising:
Historical data according to pretreated power load generates the historical data vector X of power load; The data amount check of the historical data vector X of described power load is N, treats that grouping number is C;
Determine the fuzzy grouping matrix U of the historical data vector X of power load;
Wherein,��ijIt is i-th data xiIt is under the jurisdiction of the degree of being subordinate to of jth grouping; And ��ij�� [0,1]; 1��i��N; 1��j��C;The cost function of described fuzzy grouping matrix U isWherein m is weights; DijIt is i-th data xiWith the feature v of jth groupingjBetween weighted euclidean distance;A is the variance matrix of the historical data vector X of power load;xk��Cj; CjFor jth grouping; NjFor the data amount check of jth grouping;
Optimum degree of being subordinate to and the optimal group feature of the historical data vector X of described power load is determined according to fuzzy grouping matrix U;
Wherein, described optimum degree of being subordinate to is
Described optimal group is characterized as
Determine optimum degree of being subordinate to and optimal group feature corresponding be grouped into described optimal classification.
6. short-term electro-load forecast method according to claim 5, it is characterised in that, described in the prediction algorithm that pre-sets comprise: time series forecasting method, linear regression method, neural network prediction method, gray evaluation; Described time series forecasting method comprises arma modeling algorithm.
7. a short-term electro-load forecast device, it is characterised in that, comprising:
Historical data acquiring unit, for the historical data of Weather information corresponding to the historical data of the historical data and described power load that obtain power load;
Pretreatment unit, for carrying out fuzzy clustering analysis pre-treatment to the historical data of described power load and the historical data of described Weather information;
Fuzzy clustering unit, for the historical data of pretreated power load being carried out fuzzy clustering according to Fuzzy Cluster Analysis Algorithm, adjustment fuzzy clustering parameter, obtains the optimal classification of the historical data of power load; The historical data of the power load in described optimal classification and the deviation of fuzzy clustering central value are all less than predetermined threshold value, and the historical data number of the power load in described optimal classification is greater than the historical data number of the power load carried out in other classification that fuzzy clustering produces;
Sliding window data acquiring unit, historical data for the power load of the moving window according to a preset length in described optimal classification obtains the history training data of power load, and the historical data of Weather information after the pre-treatment obtains the Weather information history training data of the history training data of described power load;
Short-term electro-load forecast unit, for according to the prediction algorithm pre-set, history training data and Weather information history training data to described power load process, obtains short-term electro-load forecast data.
8. short-term electro-load forecast device according to claim 7, it is characterised in that, the historical data of the Weather information that described historical data acquiring unit obtains comprises: temperature historical data, wind speed historical data.
9. short-term electro-load forecast device according to claim 8, it is characterised in that, described pretreatment unit, specifically for:
Negative value in the historical data of power load is set to 0.
10. short-term electro-load forecast device according to claim 9, it is characterised in that, described pretreatment unit, also for:
The missing data in the historical data of power load and the historical data of Weather information is determined according to trend prediction model;
Described trend prediction model is: xk=ak-t+bk-t�� t;
Wherein, k is current time sequence number; T is the moment quantity of missing data before the k moment; xkFor the historical data of power load or the historical data of Weather information in k moment; For Single moving average value;For Double moving average value; N is the length of each moving average.
11. short-term electro-load forecast devices according to claim 10, it is characterised in that, described fuzzy clustering unit, comprising:
Historical data vector generation module, generates the historical data vector X of power load for the historical data according to pretreated power load; The data amount check of the historical data vector X of described power load is N, treats that grouping number is C;
Fuzzy grouping matrix determination module, for determining the fuzzy grouping matrix U of the historical data vector X of power load;
Wherein,��ijIt is i-th data xiIt is under the jurisdiction of the degree of being subordinate to of jth grouping; And ��ij�� [0,1]; 1��i��N; 1��j��C;The cost function of described fuzzy grouping matrix U isWherein m is weights; DijIt is i-th data xiAnd the weighted euclidean distance between the feature vj of jth grouping;A is the variance matrix of the historical data vector X of power load;xk��Cj; CjFor jth grouping; NjFor the data amount check of jth grouping;
Optimal classification determination module, for determining the historical data vector optimum degree of being subordinate to of X and the optimal group feature of described power load according to fuzzy grouping matrix U, and determine optimum degree of being subordinate to and optimal group feature corresponding be grouped into described optimal classification;
Wherein, described optimum degree of being subordinate to is
Described optimal group is characterized as
12. short-term electro-load forecast devices according to claim 11, it is characterized in that, the prediction algorithm pre-set of described short-term electro-load forecast unit application comprises: time series forecasting method, linear regression method, neural network prediction method, gray evaluation; Described time series forecasting method comprises arma modeling algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610130474.7A CN105631483B (en) | 2016-03-08 | 2016-03-08 | A kind of short-term electro-load forecast method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610130474.7A CN105631483B (en) | 2016-03-08 | 2016-03-08 | A kind of short-term electro-load forecast method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105631483A true CN105631483A (en) | 2016-06-01 |
| CN105631483B CN105631483B (en) | 2019-03-12 |
Family
ID=56046395
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610130474.7A Active CN105631483B (en) | 2016-03-08 | 2016-03-08 | A kind of short-term electro-load forecast method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105631483B (en) |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106447133A (en) * | 2016-11-03 | 2017-02-22 | 上海交通大学 | Short-term electric load prediction method based on deep self-encoding network |
| CN106485262A (en) * | 2016-09-09 | 2017-03-08 | 国网山西省电力公司晋城供电公司 | A kind of bus load Forecasting Methodology |
| CN106844594A (en) * | 2017-01-12 | 2017-06-13 | 南京大学 | A kind of electric power Optimal Configuration Method based on big data |
| CN106920014A (en) * | 2017-03-16 | 2017-07-04 | 广东工业大学 | A kind of short-term load forecasting method and device |
| CN106936127A (en) * | 2017-02-28 | 2017-07-07 | 国家电网公司 | A kind of line load regression analysis and Forecasting Methodology and system |
| CN107491812A (en) * | 2016-06-13 | 2017-12-19 | 中国农业大学 | Short-term load forecasting method based on Spot Price |
| CN107563575A (en) * | 2017-10-11 | 2018-01-09 | 国网湖南省电力公司 | Long-term wind power power forecasting method under more meteorological variables |
| CN107578124A (en) * | 2017-08-28 | 2018-01-12 | 国网山东省电力公司电力科学研究院 | The Short-Term Load Forecasting Method of GRU neutral nets is improved based on multilayer |
| CN107590567A (en) * | 2017-09-13 | 2018-01-16 | 南京航空航天大学 | A kind of Recognition with Recurrent Neural Network short-term load forecasting method based on comentropy cluster and notice mechanism |
| CN107918779A (en) * | 2017-08-02 | 2018-04-17 | 北京国电通网络技术有限公司 | One kind builds polynary load characteristics clustering model method and system |
| CN108197773A (en) * | 2017-12-08 | 2018-06-22 | 囯网河北省电力有限公司电力科学研究院 | Methods of electric load forecasting, load forecast device and terminal device |
| CN108241625A (en) * | 2016-12-23 | 2018-07-03 | 科大讯飞股份有限公司 | Predict the method and system of student performance variation tendency |
| CN108334988A (en) * | 2018-02-08 | 2018-07-27 | 吕欣 | A kind of short-term Load Forecasting based on SVM |
| CN108596654A (en) * | 2018-04-10 | 2018-09-28 | 广州供电局有限公司 | Power information acquisition method, device, computer equipment and storage medium |
| CN108985490A (en) * | 2018-06-14 | 2018-12-11 | 内蒙古大学 | Wind-driven generator power curve modification method based on the fitting of sliding window characteristic |
| CN109214424A (en) * | 2018-08-03 | 2019-01-15 | 大连理工大学 | A method of the new-energy automobile charging time is predicted using regression analysis and clustering method |
| CN109948820A (en) * | 2017-12-20 | 2019-06-28 | 新智数字科技有限公司 | The optimization method and device of user's energy |
| CN110266002A (en) * | 2019-06-20 | 2019-09-20 | 北京百度网讯科技有限公司 | Method and apparatus for predicting electric load |
| CN110929771A (en) * | 2019-11-15 | 2020-03-27 | 北京达佳互联信息技术有限公司 | Image sample classification method and device, electronic equipment and readable storage medium |
| CN111325315A (en) * | 2019-11-22 | 2020-06-23 | 国网新疆电力有限公司电力科学研究院 | Distribution transformer power failure and power loss prediction method based on deep learning |
| CN112348702A (en) * | 2020-11-06 | 2021-02-09 | 国网黑龙江省电力有限公司电力科学研究院 | Power load prediction method based on window mobile machine learning |
| CN112381266A (en) * | 2020-10-22 | 2021-02-19 | 国网湖北省电力有限公司武汉供电公司 | System and method for predicting future power supply amount based on historical power supply and weather data |
| CN112928752A (en) * | 2021-02-03 | 2021-06-08 | 广州市电力工程设计院有限公司 | Electricity load prediction method and device and storage medium |
| CN115936184A (en) * | 2022-11-10 | 2023-04-07 | 国网冀北电力有限公司计量中心 | Load prediction matching method suitable for multi-user types |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521674A (en) * | 2011-12-07 | 2012-06-27 | 福建省电力有限公司 | Computer generating method for short-term load forecasting in summer in power consumer electricity-utilization information acquisition system |
| WO2013066793A2 (en) * | 2011-10-31 | 2013-05-10 | Siemens Corporation | Short-term load forecast using support vector regression and feature learning |
| CN103123665A (en) * | 2012-07-31 | 2013-05-29 | 上海交通大学 | Short-term power load forecasting method based on fuzzy clustering similar day |
| CN104392274A (en) * | 2014-10-29 | 2015-03-04 | 南京南瑞集团公司 | Urban short-term electrical load prediction method based on trend of electrical load and temperature |
| CN105303263A (en) * | 2015-11-16 | 2016-02-03 | 深圳合纵能源技术有限公司 | Load prediction system of regional power grid and method thereof |
-
2016
- 2016-03-08 CN CN201610130474.7A patent/CN105631483B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013066793A2 (en) * | 2011-10-31 | 2013-05-10 | Siemens Corporation | Short-term load forecast using support vector regression and feature learning |
| CN102521674A (en) * | 2011-12-07 | 2012-06-27 | 福建省电力有限公司 | Computer generating method for short-term load forecasting in summer in power consumer electricity-utilization information acquisition system |
| CN103123665A (en) * | 2012-07-31 | 2013-05-29 | 上海交通大学 | Short-term power load forecasting method based on fuzzy clustering similar day |
| CN104392274A (en) * | 2014-10-29 | 2015-03-04 | 南京南瑞集团公司 | Urban short-term electrical load prediction method based on trend of electrical load and temperature |
| CN105303263A (en) * | 2015-11-16 | 2016-02-03 | 深圳合纵能源技术有限公司 | Load prediction system of regional power grid and method thereof |
Non-Patent Citations (1)
| Title |
|---|
| 刘耀年等: "基于模糊聚类理论的电力系统短期负荷预测的方法", 《东北电力学院学报》 * |
Cited By (30)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107491812A (en) * | 2016-06-13 | 2017-12-19 | 中国农业大学 | Short-term load forecasting method based on Spot Price |
| WO2018045642A1 (en) * | 2016-09-09 | 2018-03-15 | 国网山西省电力公司晋城供电公司 | A bus bar load forecasting method |
| CN106485262A (en) * | 2016-09-09 | 2017-03-08 | 国网山西省电力公司晋城供电公司 | A kind of bus load Forecasting Methodology |
| CN106485262B (en) * | 2016-09-09 | 2020-02-07 | 国网山西省电力公司晋城供电公司 | Bus load prediction method |
| CN106447133A (en) * | 2016-11-03 | 2017-02-22 | 上海交通大学 | Short-term electric load prediction method based on deep self-encoding network |
| CN108241625A (en) * | 2016-12-23 | 2018-07-03 | 科大讯飞股份有限公司 | Predict the method and system of student performance variation tendency |
| CN106844594A (en) * | 2017-01-12 | 2017-06-13 | 南京大学 | A kind of electric power Optimal Configuration Method based on big data |
| CN106936127A (en) * | 2017-02-28 | 2017-07-07 | 国家电网公司 | A kind of line load regression analysis and Forecasting Methodology and system |
| CN106920014A (en) * | 2017-03-16 | 2017-07-04 | 广东工业大学 | A kind of short-term load forecasting method and device |
| CN107918779A (en) * | 2017-08-02 | 2018-04-17 | 北京国电通网络技术有限公司 | One kind builds polynary load characteristics clustering model method and system |
| CN107578124A (en) * | 2017-08-28 | 2018-01-12 | 国网山东省电力公司电力科学研究院 | The Short-Term Load Forecasting Method of GRU neutral nets is improved based on multilayer |
| CN107590567A (en) * | 2017-09-13 | 2018-01-16 | 南京航空航天大学 | A kind of Recognition with Recurrent Neural Network short-term load forecasting method based on comentropy cluster and notice mechanism |
| CN107563575A (en) * | 2017-10-11 | 2018-01-09 | 国网湖南省电力公司 | Long-term wind power power forecasting method under more meteorological variables |
| CN108197773A (en) * | 2017-12-08 | 2018-06-22 | 囯网河北省电力有限公司电力科学研究院 | Methods of electric load forecasting, load forecast device and terminal device |
| CN109948820A (en) * | 2017-12-20 | 2019-06-28 | 新智数字科技有限公司 | The optimization method and device of user's energy |
| CN108334988A (en) * | 2018-02-08 | 2018-07-27 | 吕欣 | A kind of short-term Load Forecasting based on SVM |
| CN108596654A (en) * | 2018-04-10 | 2018-09-28 | 广州供电局有限公司 | Power information acquisition method, device, computer equipment and storage medium |
| CN108596654B (en) * | 2018-04-10 | 2020-04-10 | 广州供电局有限公司 | Power utilization information acquisition method and device, computer equipment and storage medium |
| CN108985490A (en) * | 2018-06-14 | 2018-12-11 | 内蒙古大学 | Wind-driven generator power curve modification method based on the fitting of sliding window characteristic |
| CN109214424A (en) * | 2018-08-03 | 2019-01-15 | 大连理工大学 | A method of the new-energy automobile charging time is predicted using regression analysis and clustering method |
| CN109214424B (en) * | 2018-08-03 | 2020-02-18 | 大连理工大学 | Method for predicting charging time of new energy automobile by using regression analysis and clustering method |
| CN110266002A (en) * | 2019-06-20 | 2019-09-20 | 北京百度网讯科技有限公司 | Method and apparatus for predicting electric load |
| CN110929771A (en) * | 2019-11-15 | 2020-03-27 | 北京达佳互联信息技术有限公司 | Image sample classification method and device, electronic equipment and readable storage medium |
| CN111325315A (en) * | 2019-11-22 | 2020-06-23 | 国网新疆电力有限公司电力科学研究院 | Distribution transformer power failure and power loss prediction method based on deep learning |
| CN112381266A (en) * | 2020-10-22 | 2021-02-19 | 国网湖北省电力有限公司武汉供电公司 | System and method for predicting future power supply amount based on historical power supply and weather data |
| CN112381266B (en) * | 2020-10-22 | 2024-01-09 | 国网湖北省电力有限公司武汉供电公司 | System and method for predicting future power supply quantity based on historical power supply and weather data |
| CN112348702A (en) * | 2020-11-06 | 2021-02-09 | 国网黑龙江省电力有限公司电力科学研究院 | Power load prediction method based on window mobile machine learning |
| CN112348702B (en) * | 2020-11-06 | 2022-11-15 | 国网黑龙江省电力有限公司电力科学研究院 | Power load prediction method based on window moving machine learning |
| CN112928752A (en) * | 2021-02-03 | 2021-06-08 | 广州市电力工程设计院有限公司 | Electricity load prediction method and device and storage medium |
| CN115936184A (en) * | 2022-11-10 | 2023-04-07 | 国网冀北电力有限公司计量中心 | Load prediction matching method suitable for multi-user types |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105631483B (en) | 2019-03-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105631483A (en) | Method and device for predicting short-term power load | |
| Tian | Short-term wind speed prediction based on LMD and improved FA optimized combined kernel function LSSVM | |
| Lin et al. | An improved moth-flame optimization algorithm for support vector machine prediction of photovoltaic power generation | |
| Gao et al. | Day-ahead power forecasting in a large-scale photovoltaic plant based on weather classification using LSTM | |
| Tan et al. | Ultra-short-term wind power prediction by salp swarm algorithm-based optimizing extreme learning machine | |
| Oudjana et al. | Short term photovoltaic power generation forecasting using neural network | |
| CN108921339B (en) | Quantile regression-based photovoltaic power interval prediction method for genetic support vector machine | |
| CN111915092B (en) | Ultra-short-term wind power prediction method based on long-short-term memory neural network | |
| CN110909912A (en) | Park electric power system net load combination prediction method based on self-adaptive error feedback | |
| CN104636985A (en) | Method for predicting radio disturbance of electric transmission line by using improved BP (back propagation) neural network | |
| CN110942205A (en) | Short-term photovoltaic power generation power prediction method based on HIMVO-SVM | |
| Zhang et al. | A new solar power output prediction based on hybrid forecast engine and decomposition model | |
| CN104820877A (en) | Photovoltaic system generation power prediction method based on cloud adaptive PSO-SNN | |
| CN110070228A (en) | BP neural network wind speed prediction method for neuron branch evolution | |
| CN106503833A (en) | Photovoltaic generation short term power rolling forecast method based on algorithm of support vector machine | |
| Luo et al. | Short-term photovoltaic generation forecasting based on similar day selection and extreme learning machine | |
| Zhang et al. | Transfer learning featured short-term combining forecasting model for residential loads with small sample sets | |
| CN112288157A (en) | Wind power plant power prediction method based on fuzzy clustering and deep reinforcement learning | |
| CN116526473A (en) | Particle swarm optimization LSTM-based electrothermal load prediction method | |
| Tavares et al. | Comparison of PV power generation forecasting in a residential building using ANN and DNN | |
| Fotso et al. | A novel hybrid model based on weather variables relationships improving applied for wind speed forecasting | |
| Danish | A framework for modeling and optimization of data-driven energy systems using machine learning | |
| Tao et al. | On comparing six optimization algorithms for network-based wind speed forecasting | |
| Wang et al. | Cloud computing and extreme learning machine for a distributed energy consumption forecasting in equipment-manufacturing enterprises | |
| CN105205560A (en) | Photovoltaic power supply power prediction method based on positive and negative error variable weights |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |