CN114971051A - Wind power prediction method - Google Patents

Wind power prediction method Download PDF

Info

Publication number
CN114971051A
CN114971051A CN202210634177.1A CN202210634177A CN114971051A CN 114971051 A CN114971051 A CN 114971051A CN 202210634177 A CN202210634177 A CN 202210634177A CN 114971051 A CN114971051 A CN 114971051A
Authority
CN
China
Prior art keywords
time
wind power
real
error
wind
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
Application number
CN202210634177.1A
Other languages
Chinese (zh)
Inventor
张勇
张澜
代学冬
杨培宏
田海
吕东澔
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Inner Mongolia University of Science and Technology
Original Assignee
Inner Mongolia University of Science and Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Inner Mongolia University of Science and Technology filed Critical Inner Mongolia University of Science and Technology
Priority to CN202210634177.1A priority Critical patent/CN114971051A/en
Publication of CN114971051A publication Critical patent/CN114971051A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/381—Dispersed generators
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
    • H02J2300/20—The dispersed energy generation being of renewable origin
    • H02J2300/28—The renewable source being wind energy
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • General Physics & Mathematics (AREA)
  • Power Engineering (AREA)
  • Marketing (AREA)
  • Tourism & Hospitality (AREA)
  • General Health & Medical Sciences (AREA)
  • General Business, Economics & Management (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biophysics (AREA)
  • Game Theory and Decision Science (AREA)
  • Development Economics (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • Operations Research (AREA)
  • Primary Health Care (AREA)
  • Quality & Reliability (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention provides a wind power prediction method, which comprises the following steps: aiming at any fan in the wind power plant, the error between the acquired real-time wind power and the predicted wind power obtained by adopting the neural network model for prediction is obtained; selecting a current moment, and determining distribution characteristics according to a real-time wind power error of a historical time period before the current moment; establishing a Bayesian constant mean dynamic prediction model for error prediction according to the distribution characteristics and the real-time error of the wind power in the backward period after the current moment; and compensating the prediction result of the Bayes constant mean dynamic prediction model to the predicted wind power obtained by the neural network model prediction to obtain the compensation power. The wind power prediction method has the characteristics of high precision, quick response and the like, and can be widely applied to the field of wind power plant power error compensation.

Description

Wind power prediction method
Technical Field
The invention relates to a wind power detection technology, in particular to a wind power prediction method.
Background
At present, the main forms of new energy power generation comprise wind power, photovoltaic, nuclear power and hydroelectric power, and compared with other energy power generation forms, the wind power generation develops rapidly in new energy with the unique advantages of the wind power generation. In the wind power development process, the uncertainty of wind energy brings great trouble to power grid dispatching. The problem of how to improve the wind power prediction accuracy is extremely necessary and urgent.
The traditional wind power prediction method is generally used for predicting based on a mathematical model established by historical data of a wind power plant, but the problems of low prediction precision, slow response and the like are caused correspondingly because the data is inaccurate and the mathematical model is too complex.
Therefore, in the prior art, a wind power prediction method with high prediction accuracy and high response speed is not available.
Disclosure of Invention
In view of the above, the main objective of the present invention is to provide a wind power prediction method with high prediction accuracy and fast response speed.
In order to achieve the purpose, the technical scheme provided by the invention is as follows:
a wind power prediction method comprises the following steps:
step 1, acquiring real-time wind power of any fan at any time t in a wind farm according to real-time wind speed, real-time wind direction and real-time atmospheric humidity of the wind farm
Figure BDA0003679857310000011
Wherein the real-time wind power
Figure BDA0003679857310000012
Is a real number, and t is a natural number.
Step 2, according to the real-time wind speed and the real-time wind of the wind fieldAnd predicting the wind power in real time by adopting a neural network model to obtain the predicted wind power of any fan in the wind field at any time t
Figure BDA0003679857310000013
Wherein the wind power is predicted
Figure BDA0003679857310000021
Are real numbers.
Step 3, selecting the current time m: corresponding real-time wind power of each historical moment in a time length L before the current moment m
Figure BDA0003679857310000022
Predicting wind power
Figure BDA0003679857310000023
Comparing to obtain the real-time wind power error E in the historical time period h ＝[e m ，e m-1 ，…，e m-i ，…，e m-L ](ii) a Wherein, the real-time error of the historical wind power at any moment in the time length L before the current moment m
Figure BDA0003679857310000024
m, i and L are natural numbers, and i is more than or equal to 0 and less than or equal to L; e.g. of the type m 、e m-1 、…，e m-i 、…、e m-L Are all real numbers.
Step 4, counting the wind power real-time error E in the historical time period h Obtaining the real-time wind power error E in the historical time period h Obeying a statistical mean of mu e Statistical variance of σ e Normal distribution of and real-time error of wind power E over historical time period h Is calculated.
Step 5, obtaining the real-time wind power error E of the backward time interval in the time length S after the current moment m p ＝[e m+1 ，…，e m+j ，…，e m+S ](ii) a Wherein, the real-time error of the backward wind power at any time in the time length S after the current time m
Figure BDA0003679857310000025
j is a natural number, and j is more than or equal to 1 and less than or equal to S; e.g. of the type m+1 、…、e m+j 、…、e m+S Are all real numbers.
6, determining the real-time wind power error E in the historical time period according to the step 4 h Obeying a statistical mean of mu e Statistical variance σ e Determining the wind power real-time error E in the backward period p Also obeys statistical mean of mu e Statistical variance of σ e Is normally distributed.
7, determining the real-time error E of the wind power in the backward time period according to the step 6 p According to the real-time wind power error E of the backward time interval p Establishing a Bayesian constant mean dynamic prediction model for error prediction to obtain the wind power prediction error at the current moment
Figure BDA0003679857310000026
Step 8, obtaining the wind power prediction error of the current moment in the step 7
Figure BDA0003679857310000027
Compensating the predicted wind power at the current moment obtained in the step 2
Figure BDA0003679857310000028
Obtaining the compensation wind power of any fan at the current moment:
Figure BDA0003679857310000029
in summary, in the wind power prediction method of the present invention, the real-time wind power at each time is first obtained
Figure BDA00036798573100000210
Predicted wind power obtained by prediction of neural network model
Figure BDA0003679857310000031
And the two are compared. Determine aPrevious time m and aiming at the real-time wind power in a certain time before the current time
Figure BDA0003679857310000032
Predicted wind power obtained by prediction of neural network model
Figure BDA0003679857310000033
Real-time error E between h And determining the statistical distribution characteristics of the data. Historical time interval wind power real-time error E h The determination of the distribution characteristics means that the real-time wind power at the current moment m is later in a certain time
Figure BDA0003679857310000034
Predicted wind power obtained by prediction of neural network model
Figure BDA0003679857310000035
Backward time interval wind power real-time error between E p The distribution characteristics of (2) are also determined. Because of the real-time error E of the wind power in the historical period h Real-time wind power error E of distribution characteristic and backward time interval p The consistency is determined by the statistics of probability and mathematical statistics well known to the public. Then, according to the determined distribution characteristic and the wind power real-time error E in the backward time period p Establishing a Bayesian constant mean dynamic prediction model, and carrying out error prediction to obtain a wind power prediction error
Figure BDA0003679857310000036
Finally, the wind power prediction error is calculated
Figure BDA0003679857310000037
Compensating for predicted wind power obtained by predicting neural network model
Figure BDA0003679857310000038
Therefore, the wind power prediction method improves the prediction precision of the wind power in a mode of further compensating the wind power predicted by the neural network model. In addition, due to the inventionThe wind power prediction method has the advantages of selectable duration, concise algorithm and higher prediction response speed.
Drawings
Fig. 1 is a general flow chart of the wind power prediction method according to the present invention.
FIG. 2 is a partial predicted wind power obtained by prediction of a radial basis function neural network
Figure BDA0003679857310000039
With part of the real-time wind power
Figure BDA00036798573100000310
Schematic diagram of the difference.
FIG. 3 is a partial predicted wind power for the corresponding backward time period of FIG. 2
Figure BDA00036798573100000311
And part of real-time wind power
Figure BDA00036798573100000312
Schematic diagram of comparison.
FIG. 4 shows 3000 groups of wind power real-time errors E in historical time period in this embodiment h Schematic diagram of statistical data of (1).
FIG. 5 is a real-time wind power error E of a backward time period within a time period of 5 minutes p Schematic representation of (a).
FIG. 6 is a wind power at each moment predicted by a Bayesian constant mean value dynamic prediction model within a time length of 5 minutes
Figure BDA0003679857310000041
With part of the real-time wind power
Figure BDA0003679857310000042
Schematic diagram of comparison.
FIG. 7 is a time length of 5 minutes for compensating wind power at the present time
Figure BDA0003679857310000043
Partial predicted wind power obtained by prediction of radial basis function neural network
Figure BDA0003679857310000044
With part of the real-time wind power
Figure BDA0003679857310000045
The results are compared schematically.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
Fig. 1 is a schematic flow chart of a wind power prediction method according to the present invention. As shown in fig. 1, the wind power prediction method of the present invention includes the following steps:
step 1, acquiring real-time wind power of any fan at any time t in a wind farm according to real-time wind speed, real-time wind direction and real-time atmospheric humidity of the wind farm
Figure BDA0003679857310000046
Wherein the real-time wind power
Figure BDA0003679857310000047
Is a real number, and t is a natural number.
In step 1, the real-time wind direction includes: real-time wind direction sine value and real-time wind direction cosine value.
Step 2, according to the real-time wind speed, the real-time wind direction and the real-time atmospheric humidity of the wind field, adopting a neural network model to carry out wind power prediction to obtain the predicted wind power of any fan in the wind field at any time t
Figure BDA0003679857310000048
Wherein the wind power is predicted
Figure BDA0003679857310000049
Are real numbers.
In the method of the present invention, the neural network model is the prior art, and is not described herein again. In practical application, because there are many types of neural network models, a specific neural network model can be adopted according to actual needs.
Step 3, selecting the current time m: corresponding real-time wind power of each historical moment in a time length L before the current moment m
Figure BDA00036798573100000410
Predicting wind power
Figure BDA00036798573100000411
Comparing to obtain the real-time wind power error E in the historical time period h ＝[e m ，e m-1 ，…，e m-i ，…，e m-L ](ii) a Wherein, the real-time error of the historical wind power at any moment in the time length L before the current moment m
Figure BDA00036798573100000412
m, i and L are natural numbers, and i is more than or equal to 0 and less than or equal to L; e.g. of the type m 、e m-1 、…，e m-i 、…、e m-L Are all real numbers.
Step 4, counting the wind power real-time error E in the historical time period h Obtaining the real-time wind power error E in the historical time period h Obey statistical mean of mu e Statistical variance of σ e Normal distribution of and real-time error of wind power E over historical time period h Is calculated.
In step 4, the mean value mu is counted e 19.2, statistical variance σ e ＝32。
Step 5, obtaining the real-time wind power error E of the backward time interval in the time length S after the current moment m p ＝[e m+1 ，…，e m+j ，…，e m+S ](ii) a Wherein, the real-time error of the backward wind power at any time in the time length S after the current time m
Figure BDA0003679857310000051
j is a natural number, and j is more than or equal to 1 and less than or equal to S; e.g. of a cylinder m+1 、…、e m+j 、…、e m+S Are all real numbers.
6, determining the real-time wind power error E in the historical time period according to the step 4 h Obey statistical meanIs mu e Statistical variance of σ e Determining the real-time wind power error E in the backward period p Also obeys statistical mean of mu e Statistical variance of σ e Is normally distributed.
7, determining the real-time wind power error E of the backward time interval according to the step 6 p According to the real-time wind power error E of the backward time interval p Establishing a Bayesian constant mean dynamic prediction model for error prediction to obtain the wind power prediction error at the current moment
Figure BDA0003679857310000052
Step 8, obtaining the wind power prediction error of the current moment in the step 7
Figure BDA0003679857310000053
Compensating the predicted wind power at the current moment obtained in the step 2
Figure BDA0003679857310000054
Obtaining the compensation wind power of any fan at the current moment:
Figure BDA0003679857310000055
in summary, in the wind power prediction method of the present invention, the real-time wind power at each time is first obtained
Figure BDA0003679857310000056
Predicted wind power obtained by prediction of neural network model
Figure BDA0003679857310000057
And the two are compared. Determining a current moment m and aiming at the real-time wind power in a certain time before the current moment
Figure BDA0003679857310000058
Predicted wind power obtained by prediction of neural network model
Figure BDA0003679857310000059
Real-time error E between h And determining the statistical distribution characteristics of the data. Historical time interval wind power real-time error E h The determination of the distribution characteristics means that the real-time wind power at the current moment m is later for a certain time
Figure BDA00036798573100000510
Predicted wind power obtained by prediction of neural network model
Figure BDA00036798573100000511
Backward time interval wind power real-time error between E p The distribution characteristics of (2) are also determined. Because of the real-time wind power error E in the historical period h Real-time wind power error E of distribution characteristic and backward time interval p The consistency is determined by the statistics of probability and mathematical statistics well known to the public. Then, according to the determined distribution characteristic and the wind power real-time error E in the backward time period p Establishing a Bayesian constant mean dynamic prediction model, and carrying out error prediction to obtain a wind power prediction error
Figure BDA0003679857310000061
Finally, the wind power prediction error is calculated
Figure BDA0003679857310000062
Compensating for predicted wind power obtained by predicting neural network model
Figure BDA0003679857310000063
Therefore, the wind power prediction method improves the prediction precision of the wind power in a mode of further compensating the wind power predicted by the neural network model. In addition, the wind power prediction method can select time length, has concise algorithm and higher prediction response speed.
Corresponding each historical moment in the duration L before the current moment m to obtain the real-time wind power error E of the historical time period h ＝[e m ，e m-1 ，…，e m-i ，…，e m-L ](ii) a Wherein, the real-time error of the historical wind power at any moment in the time length L before the current moment m
Figure BDA0003679857310000064
m, i and L are natural numbers, and i is more than or equal to 0 and less than or equal to L; e.g. of a cylinder m 、e m-1 、…，e m-i 、…、e m-L Are all real numbers.
In step 7 of the method, the Bayesian constant mean dynamic prediction model specifically comprises the following steps:
the observation equation is:
Figure BDA0003679857310000065
the state equation is: e p ＝E p-1 +W m
Wherein the observation error V m ＝[v m ，v m+1 ，…，v m+j ，…，v m+S ]State error V m ＝[v m ，v m+1 ，…，v m+j ，…，v m+S ]Are all statistical mean 0 and statistical variance σ e Is normally distributed.
In the method of the invention, the wind power real-time error E in the backward time interval p Mean value at current time m
Figure BDA0003679857310000066
Real-time wind power error E in backward time period p Mean value at the next time instant (m +1)
Figure BDA0003679857310000067
Real-time wind power error E in backward time period p Variance at current time m
Figure BDA0003679857310000068
Real-time wind power error E in backward time period p Variance at the next time instant (m +1)
Figure BDA0003679857310000069
Error in observation at current time mDifference v m State error w at current time m m The observation error v at the next time (m +1) m+1 Next time (m +1) state error w m+1 The relationship between them is as follows:
Figure BDA00036798573100000610
Figure BDA0003679857310000071
in practical application, the wind power real-time error E in the backward period p Mean value at current time m
Figure BDA0003679857310000072
Real-time wind power error E in backward time period p Mean value at the next time instant (m +1)
Figure BDA0003679857310000073
Around the statistical mean μ e Fluctuating up and down; real-time wind power error E in backward time period p Variance at current time m
Figure BDA0003679857310000074
Real-time wind power error E in backward time period p Variance at the next time instant (m +1)
Figure BDA0003679857310000075
Around the statistical variance σ e Fluctuating up and down.
Examples
Take a certain wind farm of inner Mongolia as an example. In the wind power plant, for any one wind turbine, the specific operation steps in the wind power prediction method are as follows.
In the step 2, wind power prediction is performed by adopting a neural network model according to the real-time wind speed, the real-time wind direction and the real-time atmospheric humidity of the wind field, so as to obtain the predicted wind power of any fan in the wind field at any time t
Figure BDA0003679857310000076
The method specifically comprises the following steps: in the wind power plant, 7900 data groups corresponding to real-time wind speed, real-time wind direction sine value, real-time wind direction cosine value and real-time atmospheric humidity are selected according to a time sequence, a radial basis function neural network model in an MATLAB experimental box is adopted to carry out wind power prediction, and the predicted wind power of any fan at the moment t is obtained
Figure BDA0003679857310000077
FIG. 2 is a partial predicted wind power obtained by prediction of a radial basis function neural network
Figure BDA0003679857310000078
With part of the real-time wind power
Figure BDA0003679857310000079
Schematic diagram of the difference. FIG. 3 is a partial predicted wind power for the corresponding backward time period of FIG. 2
Figure BDA00036798573100000710
With part of the real-time wind power
Figure BDA00036798573100000711
Schematic diagram of comparison.
The step 4 specifically includes the following steps:
step 41, obtaining 3000 groups of wind power real-time errors E in historical time period by adopting the method h ＝[e m ，e m-1 ，…，e m-3000 ]And its statistical histogram. FIG. 4 shows 3000 groups of wind power real-time errors E in historical time period in this embodiment h Schematic diagram of statistical data of (1).
Step 42, adopting a Distribution fitting (Distribution Fitter) tool box in MATLAB, and carrying out real-time wind power error E on 3000 groups of historical time periods shown in FIG. 4 h Fitting the statistical data to obtain the real-time wind power error E in the historical period h Satisfies the mean value mu e 19.2, variance σ e Normal distribution of 32.
FIG. 5 is a schematic view ofReal-time wind power error E of backward time interval within 5 minutes p Schematic representation of (a). FIG. 6 is a wind power at each moment predicted by a Bayesian constant mean value dynamic prediction model within a time length of 5 minutes
Figure BDA0003679857310000081
And part of real-time wind power
Figure BDA0003679857310000082
Schematic diagram of comparison. By adopting the method, the real-time wind power error E of the backward time interval within the time length of 5 minutes p And forecasting errors of wind power at each moment predicted by the Bayes constant mean value dynamic forecasting model within 5 minutes
Figure BDA0003679857310000083
As shown in fig. 5 and 6, it can be seen that: moment wind power prediction error obtained by Bayes constant mean dynamic prediction model prediction
Figure BDA0003679857310000084
Is obviously reduced.
The wind power prediction error at the current moment in the embodiment is used
Figure BDA0003679857310000085
Compensating for predicted wind power at the current moment
Figure BDA0003679857310000086
Obtaining the compensation wind power of any fan at the current moment:
Figure BDA0003679857310000087
FIG. 7 is a time length of 5 minutes for compensating wind power at the present time
Figure BDA0003679857310000088
Partial predicted wind power obtained by prediction of radial basis function neural network
Figure BDA0003679857310000089
With part of the real-time wind power
Figure BDA00036798573100000810
The results are compared schematically. As can be seen from fig. 7, the wind power is compensated at time t ═ m + j
Figure BDA00036798573100000811
With corresponding real-time wind power
Figure BDA00036798573100000812
The error value between is smaller than the predicted wind power of the radial basis function neural network on the whole
Figure BDA00036798573100000813
With corresponding real-time wind power
Figure BDA00036798573100000814
Error values between, indicating: the wind power compensation method has the advantage that the precision of the obtained compensation wind power is higher.
In the invention, the verification indexes of the wind power prediction method comprise a prediction hit rate, an average relative error, a root mean square error and a maximum absolute error. For the above indexes, in this embodiment, the effect pair ratio between the prediction result of the radial basis function neural network model and the result of compensating the wind power is shown in table 1:
TABLE 1 comparison of Effect between radial basis function neural network model prediction results and Compensation wind Power results
Figure BDA00036798573100000815
In Table 1, hit rates are predicted
Figure BDA00036798573100000816
Wherein G is n Indicates whether a hit occurs, and:
Figure BDA00036798573100000817
wherein, the first and the second end of the pipe are connected with each other,
Figure BDA0003679857310000091
representing a neural network to predict wind power
Figure BDA0003679857310000092
With real-time wind power
Figure BDA0003679857310000093
The maximum error in between.
In table 1, the average relative error of the prediction result of the radial basis function neural network model is:
Figure BDA0003679857310000094
the average relative error of the wind power compensation result of the invention is as follows:
Figure BDA0003679857310000095
wherein C represents the rated installed capacity of any one fan in the wind power plant; n represents the predicted wind power of the radial basis function neural network model participating in verification
Figure BDA0003679857310000096
Or compensating the wind power P m The number of samples of (a). Predicted wind power of radial basis function neural network model
Figure BDA0003679857310000097
The number of samples and the wind power P of the compensation m Are equal in number.
In table 1, the root mean square error of the prediction result of the radial basis function neural network model is:
Figure BDA0003679857310000098
the average relative error of the wind power compensation result of the invention is as follows:
Figure BDA0003679857310000099
in table 1, the maximum absolute error of the prediction result of the radial basis function neural network model is:
Figure BDA00036798573100000910
the maximum absolute error of the wind power compensation result of the invention is as follows:
Figure BDA00036798573100000911
in summary, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (5)

1. The invention relates to a wind power prediction method, which is characterized by comprising the following steps:
step 1, acquiring real-time wind power of any fan at any time t in a wind farm according to real-time wind speed, real-time wind direction and real-time atmospheric humidity of the wind farm
Figure FDA0003679857300000011
Wherein the real-time wind power
Figure FDA0003679857300000012
Is a real number, t is a natural number;
step 2, according to the real-time wind speed and real-time of the wind fieldWind direction and real-time atmospheric humidity, and adopting a neural network model to predict wind power to obtain the predicted wind power of any fan in a wind field at any time t
Figure FDA0003679857300000013
Wherein the wind power is predicted
Figure FDA0003679857300000014
Is a real number;
step 3, selecting the current time m: corresponding real-time wind power of each historical moment in a time length L before the current moment m
Figure FDA0003679857300000015
Predicting wind power
Figure FDA0003679857300000016
Comparing to obtain the real-time wind power error E in the historical time period h ＝[e m ，e m-1 ，…，e m-i ，…，e m-L ](ii) a Wherein, the real-time error of the historical wind power at any moment in the time length L before the current moment m
Figure FDA0003679857300000017
m, i and L are natural numbers, and i is more than or equal to 0 and less than or equal to L; e.g. of the type m 、e m-1 、…，e m-i 、…、e m-L Are all real numbers;
step 4, counting the wind power real-time error E in the historical time period h The distribution characteristic of the wind power real-time error in the historical time interval E is obtained h Obeying a statistical mean of mu e Statistical variance σ e Normal distribution of and real-time error of wind power E over historical time period h The statistical histogram of (1);
step 5, obtaining the real-time wind power error E of the backward time interval in the time length S after the current moment m p ＝[e m+1 ，…，e m+j ，…，e m+S ](ii) a Wherein, the real-time error of the backward wind power at any time in the time length S after the current time m
Figure FDA0003679857300000018
j is a natural number, and j is more than or equal to 1 and less than or equal to S; e.g. of the type m+1 、…、e m+j 、…、e m+S Are all real numbers;
6, determining the real-time wind power error E in the historical time period according to the step 4 h Obeying a statistical mean of mu e Statistical variance of σ e Determining the wind power real-time error E in the backward period p Also obeys statistical mean of mu e Statistical variance of σ e Normal distribution of (2);
7, determining the real-time wind power error E of the backward time interval according to the step 6 p According to the real-time wind power error E of the backward time interval p Establishing a Bayesian constant mean dynamic prediction model for error prediction to obtain the wind power prediction error at the current moment
Figure FDA0003679857300000021
Step 8, obtaining the wind power prediction error of the current moment in the step 7
Figure FDA0003679857300000022
Compensating the predicted wind power at the current moment obtained in the step 2
Figure FDA0003679857300000023
Obtaining the compensation wind power of any fan at the current moment:
Figure FDA0003679857300000024
2. the method according to claim 1, wherein in step 7, the bayesian constant mean value dynamic prediction model specifically comprises the following steps:
the observation equation is:
Figure FDA0003679857300000025
the state equation is: e p ＝E p-1 +W m
Wherein the observation error V m ＝[v m ，v m+1 ，…，v m+j ，…，v m+S ]State error V m ＝[v m ，v m+1 ，…，v m+j ，…，v m+S ]Are all statistical mean 0 and statistical variance σ e Is normally distributed.
3. A method according to claim 1 or 2, wherein in step 4, the statistical mean μ e 19.2, the statistical variance σ e ＝32。
4. The method of claim 2, wherein the real-time wind power error E is determined by the backward time interval p Mean value at current time m
Figure FDA0003679857300000026
The wind power real-time error E of the backward time interval p Mean value at the next time instant (m +1)
Figure FDA0003679857300000027
The wind power real-time error E of the backward time interval p Variance at current time m
Figure FDA0003679857300000028
Observation error v of current time m m State error w at current time m m The observation error v at the next time (m +1) m+1 Next time (m +1) state error w m+1 The relationship between them is as follows:
Figure FDA0003679857300000029
Figure FDA00036798573000000210
5. the method according to claim 1, wherein in step 1, the real-time wind direction comprises: real-time wind direction sine value and real-time wind direction cosine value.
CN202210634177.1A 2022-06-06 2022-06-06 Wind power prediction method Pending CN114971051A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210634177.1A CN114971051A (en) 2022-06-06 2022-06-06 Wind power prediction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210634177.1A CN114971051A (en) 2022-06-06 2022-06-06 Wind power prediction method

Publications (1)

Publication Number Publication Date
CN114971051A true CN114971051A (en) 2022-08-30

Family

ID=82959283

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210634177.1A Pending CN114971051A (en) 2022-06-06 2022-06-06 Wind power prediction method

Country Status (1)

Country Link
CN (1) CN114971051A (en)

Similar Documents

Publication Publication Date Title
CN110648249B (en) Annual power balance measuring and calculating method, device and equipment
CN110766200A (en) Method for predicting generating power of wind turbine generator based on K-means mean clustering
CN110633864B (en) Wind speed numerical forecast correction method and system based on distance flat deviation
CN113657662B (en) Downscaling wind power prediction method based on data fusion
CN110460038A (en) It is a kind of to be related to more scene method for expansion planning of power transmission network of new-energy grid-connected
CN114819374A (en) Regional new energy ultra-short term power prediction method and system
CN113991711A (en) Capacity configuration method for energy storage system of photovoltaic power station
CN114971051A (en) Wind power prediction method
CN116701877A (en) Fan abnormal value repairing method based on conditional diffusion model
CN115456286A (en) Short-term photovoltaic power prediction method
CN113887839A (en) Clustering-based wind power plant ultra-short-term wind power prediction method and system
CN114595989A (en) Wind turbine generator performance evaluation method and device
CN109524983B (en) Photovoltaic output modeling method based on typical state
CN113112085A (en) New energy station power generation load prediction method based on BP neural network
CN112803403A (en) Offshore wind power clustering effect multi-level evaluation method based on time-frequency characteristics
JP7177238B1 (en) FAILURE DETERMINATION SYSTEM, FAILURE DETERMINATION PROGRAM AND FAILURE DETERMINATION METHOD
CN116822370B (en) Ultra-short-term output prediction method for wind power cluster under data driving
CN116757101B (en) Cabin wind speed correction method and system based on mechanism model and neural network
CN110659776B (en) New energy power generation power prediction method and system with self-adaptive time scale
CN109325879B (en) Space-time polymerization method for calculating variable comprehensive output coefficient of long-term scheduling in hydropower station
CN108154268B (en) Method for rapidly estimating group power generation quantity of small hydropower stations
CN116452040A (en) Evaluation method and evaluation system for generating capacity increment of wind turbine generator
CN115425645A (en) Implementation method for improving wind power prediction precision
CN114626631A (en) Short-time wind speed prediction method based on non-stationary and non-Gaussian smooth interpolation preprocessing technology
CN115271287A (en) Method, device, equipment and storage medium for determining increased power generation quantity

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