CN116316615B - Data enhancement-based distributed light Fu Qun short-term power prediction method and system - Google Patents

Data enhancement-based distributed light Fu Qun short-term power prediction method and system Download PDF

Info

Publication number
CN116316615B
CN116316615B CN202310596490.5A CN202310596490A CN116316615B CN 116316615 B CN116316615 B CN 116316615B CN 202310596490 A CN202310596490 A CN 202310596490A CN 116316615 B CN116316615 B CN 116316615B
Authority
CN
China
Prior art keywords
photovoltaic power
power
power station
prediction
short
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.)
Active
Application number
CN202310596490.5A
Other languages
Chinese (zh)
Other versions
CN116316615A (en
Inventor
龚伟
曾伟
王文彬
周求宽
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.)
State Grid Corp of China SGCC
Electric Power Research Institute of State Grid Jiangxi Electric Power Co Ltd
Original Assignee
State Grid Corp of China SGCC
Electric Power Research Institute of State Grid Jiangxi Electric Power Co Ltd
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 State Grid Corp of China SGCC, Electric Power Research Institute of State Grid Jiangxi Electric Power Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN202310596490.5A priority Critical patent/CN116316615B/en
Publication of CN116316615A publication Critical patent/CN116316615A/en
Application granted granted Critical
Publication of CN116316615B publication Critical patent/CN116316615B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Electricity, gas or water supply
    • 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
    • H02J2203/10—Power transmission or distribution systems management focussing at grid-level, e.g. load flow analysis, node profile computation, meshed network optimisation, active network management or spinning reserve management
    • 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
    • H02J2300/22—The renewable source being solar energy
    • H02J2300/24—The renewable source being solar energy of photovoltaic origin
    • 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

Abstract

The application belongs to the technical field of power system operation and analysis, and relates to a data enhancement-based distributed light Fu Qun short-term power prediction method and a data enhancement-based distributed light Fu Qun short-term power prediction system, wherein the method carries out per unit on a photovoltaic power generation power actual measurement value and a short-term photovoltaic power generation power prediction value of a photovoltaic power station with modeling conditions; training a comprehensive photovoltaic power station prediction model based on the historical data of each distributed photovoltaic power station after per unit, and obtaining a comprehensive prediction result by using the comprehensive photovoltaic power station prediction model; naming the comprehensive prediction result as a short-term generation power prediction value of a typical photovoltaic power station; the short-term power generation power prediction value and the historical power generation power of the typical photovoltaic power station participate in statistical upscale prediction to form a short-term power generation power prediction result of the distributed photovoltaic power station group. According to the method, a data enhancement mode is adopted, and the prediction accuracy of the short-term power generation power of the sample photovoltaic power station is improved to the greatest extent by enhancing the training of the historical data of the sample power station.

Description

Data enhancement-based distributed light Fu Qun short-term power prediction method and system
Technical Field
The application relates to the technical field of power system operation and analysis, in particular to a data enhancement-based distributed light Fu Qun short-term power prediction method and system.
Background
The photovoltaic short-term power generation power prediction is based on historical power generation output, numerical weather forecast and actual measurement meteorological data, and the prediction of future 24h photovoltaic power generation output is realized by constructing a short-term power generation power prediction model. Centralized photovoltaics generally have a better data basis, and the prediction effect can be generally guaranteed. However, the large number of distributed photovoltaics does not have the data conditions described above, resulting in distributed photovoltaics that do not have conditions for developing high-level short-term generated power predictions similar to centralized photovoltaics. To date, some prediction technologies exist, so that short-term power generation prediction of distributed photovoltaic power station groups is achieved, for example, patent document CN109978242a discloses a photovoltaic power generation cluster power prediction method based on statistical upscaling.
In the process of statistical upscaling prediction of the short-term generated power of a distributed photovoltaic power station group, 2 basic conditions are usually required for selecting a typical photovoltaic power station: (1) The output power of the distributed photovoltaic power station group has higher correlation; (2) The power prediction result of the single photovoltaic power station has higher precision, and the root mean square error is generally less than 30% and the average absolute error is generally less than 25%. For the 1 st condition, based on a time sequence matrix of the total actual power generation power of each station and the cluster, the screening of the typical photovoltaic power station is realized by analyzing the correlation coefficient between the actual power generation power of each station and the total actual power generation power of the cluster. However, for condition 2, the local and same spatial-temporal resolution data (including actual measurement and prediction, weather and electrical data) required for photovoltaic short-term generation power prediction are generally not available in distributed photovoltaic power stations, so that the range selectable by a typical photovoltaic power station is limited, and finally, the accuracy of power prediction of the typical photovoltaic power station is difficult to meet the requirement.
Therefore, in the short-term power generation power prediction process of the distributed photovoltaic power station group based on statistical upscaling, a photovoltaic power station with higher correlation with the output power of the distributed photovoltaic power station group and higher precision of the self power prediction result is generally selected as a typical photovoltaic power station to participate in the statistical upscaling prediction, so that the short-term power generation power prediction result of the distributed photovoltaic power station group is formed. However, the prediction accuracy of the short-term generated power of the distributed photovoltaic power station group cannot be guaranteed due to the influence of the prediction accuracy of the typical photovoltaic power station.
Disclosure of Invention
The application aims to provide a data enhancement-based distributed light Fu Qun short-term power prediction method and system, which take a photovoltaic power station with modeling conditions as a sample, give full play to data value, furthest improve the prediction precision of short-term power generation of the sample photovoltaic power station through a comprehensive model technology, take the sample photovoltaic power station as a typical photovoltaic power station short-term power generation power prediction result, participate in statistical upscale calculation and finally form a distributed photovoltaic power station group short-term power generation power prediction result.
The application is realized by the following technical scheme, and the data enhancement-based distributed light Fu Qun short-term power prediction method comprises the following steps:
step one, carrying out per unit on a photovoltaic power generation power actual measurement value and a short-term photovoltaic power generation power prediction value of a photovoltaic power station with modeling conditions; analyzing the condition and the geographic position of an access bus among all distributed photovoltaic power stations, dividing the access bus into a plurality of clusters, and adopting the following formula for per unit of photovoltaic power generation:
wherein:representing the capacity of the kth photovoltaic power plant with modeling conditions; />Representing the actually measured value of the photovoltaic power generation power of the kth photovoltaic power station with modeling conditions in the t period; />Representing the kth photovoltaic with modeling condition in t period after per unitActual measurement values of photovoltaic power generation power of the power station; />Representing a short-term photovoltaic power generation power predicted value of a kth photovoltaic power station with modeling conditions in a t period; />Representing a short-term photovoltaic power generation power predicted value of a photovoltaic power station with modeling conditions in a t period after per unit of the photovoltaic power station, wherein q is the number of the photovoltaic power stations;
step two, training a comprehensive photovoltaic power station prediction model based on the historical data of each distributed photovoltaic power station after per unit, and obtaining a comprehensive prediction result by using the comprehensive photovoltaic power station prediction model;
the comprehensive prediction result of the comprehensive prediction model of the photovoltaic power station based on q photovoltaic power stations with modeling conditions is shown as follows:wherein->，/>，/>Representing the comprehensive predicted value of the comprehensive predicted model of the photovoltaic power station in the t period; />The weight coefficient of the photovoltaic power station with modeling conditions is the kth photovoltaic power station;
target of comprehensive prediction model of photovoltaic power station: for q photovoltaic power plants with modeling conditionsOptimizing a group of optimal weight coefficients to minimize the fitting error of a comprehensive prediction model of the photovoltaic power station, wherein the objective function is expressed as follows:wherein->Represents the fitting error, n represents the number of time periods, +.>Representing the sum of the actual photovoltaic power generation power values of the photovoltaic power station in the t period after per unit of the photovoltaic power station, and +.>；
Step three, naming the comprehensive prediction result and taking the result as a short-term generation power prediction value of a typical photovoltaic power stationThe method comprises the steps of carrying out a first treatment on the surface of the The comprehensive prediction result is named by adopting the following formula: />；
Step four, based on a short-term power generation power predicted value of a typical photovoltaic power station and historical power generation power of the typical photovoltaic power station, participating in statistical upscale prediction, and forming a short-term power generation power predicted result of a distributed photovoltaic power station group; the historical power generation power of the typical photovoltaic power station is formed by accumulating actual measured values of the historical photovoltaic power generation power of the photovoltaic power station with modeling conditions.
Further preferably, the step four specifically includes:
calculating a correlation coefficient between the historical power generation power of a typical photovoltaic power station and the actual measurement value of the historical power generation power of a distributed photovoltaic power station group：
Wherein: y is Y 0t Historical power generation power of a typical photovoltaic power station in a t period;the method comprises the steps of obtaining a historical power generation actual measurement value of a distributed photovoltaic power station group at a t period; />Is the average value of historical power generation power of a typical photovoltaic power station in a time period; />Distributing the average value of the historical power generation actual measurement values of the photovoltaic power station group in a time period;
calculating weight coefficients of typical photovoltaic power stations in photovoltaic power generation power prediction to form a weight coefficient matrixWeight coefficient matrix->Expressed as: />，/>Is constant, matrix->The definition is as follows:
wherein:indicate->Typical photovoltaic power stations are 1->F is less than or equal to F, and F is the number of typical photovoltaic power stations; />Representing typical lightCorrelation coefficient between historical power generation power and predicted value of the photovoltaic power station; sign->Hadamard product representing matrix, +.>Represent the firstCorrelation coefficient between historical power generation power of typical photovoltaic power station and actual measurement value of historical power generation power of distributed photovoltaic power station group, +.>Indicate->The correlation coefficient of the historical power generation power and the predicted value of the typical photovoltaic power station;
according to the weight coefficient matrix, different weights are distributed for each typical photovoltaic power station, and based on a statistical upscaling method, the short-term generation power predicted value of the distributed photovoltaic power station group is expressed as follows:
wherein:a time sequence matrix of short-term generation power predicted values for the distributed photovoltaic power plant group; />A time series matrix of photovoltaic power generation power predicted values for a typical power station; the statistical parameters c and d are constants; e is an identity matrix; according to historical data, a least square method is adopted to solve a calculation equation of a short-term generation power prediction value of the distributed photovoltaic power station group, the values of c and d are calculated, and then the calculation equation is used for future short-term generation power prediction of the distributed photovoltaic power station group.
The application provides a data enhancement-based distributed photovoltaic Fu Qun short-term power prediction system, which comprises a data acquisition module, a photovoltaic power station comprehensive prediction module, a correlation coefficient calculation module and a statistical upscale prediction module, wherein the data acquisition module is used for acquiring actual measurement values of photovoltaic power generation power of all photovoltaic power stations, the photovoltaic power station comprehensive prediction module is internally provided with a photovoltaic power station comprehensive prediction model, the photovoltaic power station comprehensive prediction model obtains a comprehensive prediction result according to the short-term photovoltaic power generation power prediction value of each photovoltaic power station after per unit treatment and the actual measurement values of the photovoltaic power generation power of each photovoltaic power station after per unit treatment, the correlation coefficient calculation module is used for calculating correlation coefficients between the historical power generation power of the typical photovoltaic power stations and the historical power generation power measurement values of a distributed photovoltaic power station group and the correlation coefficients of the historical power generation power of each typical photovoltaic power station and the prediction values, the correlation coefficient calculation module is used for calculating weight coefficients of the typical photovoltaic power stations in the photovoltaic power generation power prediction, and the statistical upscale prediction module is used for distributing different weights for all typical photovoltaic power stations according to the weight coefficients and calculating the short-term power generation power prediction value of the distributed photovoltaic power station group based on the statistical upscale method.
The present application provides a non-volatile computer storage medium having stored thereon computer executable instructions for performing the data enhanced based distributed light Fu Qun short-term power prediction method of any of the above embodiments.
The present application also provides a computer program product comprising a computer program stored on a non-volatile computer storage medium, the computer program comprising program instructions which, when executed by a computer, cause the computer to perform the data enhanced distributed light Fu Qun short-term power prediction method of the above embodiments.
The present application provides an electronic device including: at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions for execution by the at least one processor to enable the at least one processor to perform a data enhancement based distributed light Fu Qun short-term power prediction method.
According to the method, a limited photovoltaic power station with modeling conditions is taken as a sample, a comprehensive prediction model of the photovoltaic power station is constructed, the training of historical data of the sample photovoltaic power station is enhanced by adopting a mode of enhancing per unit data of photovoltaic power generation power, and the prediction precision of short-term power generation power of the sample photovoltaic power station is improved to the greatest extent. The prediction results indicate that the method is effective.
Drawings
FIG. 1 is a flow chart of the method of the present application.
FIG. 2 is a graph showing the comparison of predicted values and measured values.
Detailed Description
The present application is further illustrated in detail below.
Referring to fig. 1, a data enhancement based distributed optical Fu Qun short-term power prediction method comprises the following steps:
step one, carrying out per unit on a photovoltaic power generation power actual measurement value and a short-term photovoltaic power generation power prediction value of a photovoltaic power station with modeling conditions; and analyzing the condition and the geographic position of the access bus among the distributed photovoltaic power stations, and dividing the distributed photovoltaic power stations into a plurality of clusters. Taking a certain cluster as an example, the following formula is adopted for per unit of photovoltaic power generation:
wherein:representing the capacity of the kth photovoltaic power plant with modeling conditions; />Representing the actually measured value of the photovoltaic power generation power of the kth photovoltaic power station with modeling conditions in the t period; />Representing per unit of the datathe kth period of time is provided with a photovoltaic power generation power actual measurement value of the modeling condition photovoltaic power station; />Representing a short-term photovoltaic power generation power predicted value of a kth photovoltaic power station with modeling conditions in a t period; />And (3) representing a short-term photovoltaic power generation power predicted value of the photovoltaic power station with modeling conditions in the kth period after per-unit, wherein q is the number of the photovoltaic power stations.
Step two, training a comprehensive photovoltaic power station prediction model based on the historical data of each distributed photovoltaic power station after per unit, and obtaining a comprehensive prediction result by using the comprehensive photovoltaic power station prediction model;
the comprehensive prediction result of the comprehensive prediction model of the photovoltaic power station based on q photovoltaic power stations with modeling conditions is shown as follows:wherein->，/>，/>Representing the comprehensive predicted value of the comprehensive predicted model of the photovoltaic power station in the t period;the weight coefficient of the modeling condition photovoltaic power station is provided for the kth photovoltaic power station.
Target of comprehensive prediction model of photovoltaic power station: for q photovoltaic power plants with modeling conditionsOptimizing a group of optimal weight coefficients to minimize fitting errors of a comprehensive prediction model of the photovoltaic power station, wherein an objective function is expressed as follows:wherein->Represents the fitting error, n represents the number of time periods, +.>Representing the sum of the actual photovoltaic power generation power values of the photovoltaic power station in the t period after per unit of the photovoltaic power station, and +.>。
Step three, naming the comprehensive prediction result and taking the result as a short-term generation power prediction value of a typical photovoltaic power stationThe method comprises the steps of carrying out a first treatment on the surface of the The comprehensive prediction result is named by adopting the following formula: />。
And step four, based on a short-term power generation power predicted value of a typical photovoltaic power station and a historical power generation power of the typical photovoltaic power station (formed by accumulating historical photovoltaic power generation power actual measurement values of the photovoltaic power stations with modeling conditions), participating in statistical upscaling prediction to form a short-term power generation power predicted result of the distributed photovoltaic power station group.
(1) Calculating a correlation coefficient between the historical power generation power of a typical photovoltaic power station and the actual measurement value of the historical power generation power of a distributed photovoltaic power station group。
Wherein: y is Y 0t Historical power generation power of a typical photovoltaic power station in a t period;historical power generation power practice for t-period distributed photovoltaic power station groupMeasuring a value; />Is the average value of historical power generation power of a typical photovoltaic power station in a time period; />And the average value of the historical power generation actual measurement values of the photovoltaic power station group is distributed in a time period.
(2) And calculating a weight coefficient of a typical photovoltaic power station in the photovoltaic power generation power prediction.
Weight coefficient matrixExpressed as: />，/>Is constant, matrix->The definition is as follows:
wherein:indicate->Typical photovoltaic power stations are 1->F is less than or equal to F, and F is the number of typical photovoltaic power stations; r is R FA The method comprises the steps of obtaining a correlation coefficient between the historical power generation power of a typical photovoltaic power station and the actual measurement value of the historical power generation power of a distributed photovoltaic power station group; />Representing the relation between the historical power generation power and the predicted value of a typical photovoltaic power stationIs a correlation coefficient of (2); sign->Hadamard product representing matrix, +.>Indicate->Correlation coefficient between historical power generation power of typical photovoltaic power station and actual measurement value of historical power generation power of distributed photovoltaic power station group, +.>Indicate->The correlation coefficient of the historical power generation power and the predicted value of the typical photovoltaic power station;
according to the weight coefficient matrix, different weights are distributed for each typical photovoltaic power station, and a short-term power generation power predicted value of the distributed photovoltaic power station group is solved based on a statistical upscaling method; the short-term generated power predicted value of the distributed photovoltaic power station group is expressed as:
wherein:a time sequence matrix of short-term generation power predicted values for the distributed photovoltaic power plant group; />A time series matrix of photovoltaic power generation power predicted values for a typical power station; the statistical parameters c and d are constants; e is an identity matrix; according to historical data, a least square method is adopted to solve a calculation equation of a short-term generation power prediction value of the distributed photovoltaic power station group, the values of c and d are calculated, and then the calculation equation is used for future short-term generation power prediction of the distributed photovoltaic power station group.
Based on a statistical upscaling method, the short-term power generation power predicted value of the distributed photovoltaic power station group can be obtained by calculating a photovoltaic power generation power predicted value of a typical power station and a historical power generation power actual measurement value of the distributed photovoltaic power station group.
Calculating: 110kV, 10kV and 0.4kV photovoltaic power generation systems accessed to 220kV transformer substations in a certain area are selected and respectively defined as a system A, a system B and a system C, and active output data of 5min resolution of each system can be obtained. The average prediction accuracy of system a, system B, and system C over the recent period of time was 91.4%, 89.8%, and 94.7%, respectively.
And respectively based on the photovoltaic power stations with modeling conditions in the system A, the system B and the system C, adopting a comprehensive prediction model of the photovoltaic power stations to predict, and forming comprehensive prediction results of typical photovoltaic power stations in the system A, the system B and the system C.
The power correlation coefficients of the typical photovoltaic power stations and photovoltaic power generation system clusters in the system a, the system B and the system C are 0.9875, 0.943 and 0.852 respectively through correlation analysis.
According to the solving method of the weight coefficient in the statistical upscaling method and the calculated correlation coefficient of the system A0.9981 the correlation coefficient between the historical generated power of the 1 st typical photovoltaic power station and the actual measurement value of the historical generated power of the distributed photovoltaic power station group ∈ ->0.9511 the correlation coefficient between the historical generated power of the 2 nd typical photovoltaic power station and the actual measurement value of the historical generated power of the distributed photovoltaic power station group ∈ ->0.9642 the correlation coefficient between the historical power of the 3 rd typical photovoltaic power plant and the actual measurement of the historical power of the distributed photovoltaic power plant group ∈ ->=0.881, the weight coefficient matrix ++can be obtained according to the weight coefficient calculation formula>。
Finally, based on the predicted value and the measured value of the generated power of the selected typical photovoltaic power station, as shown in fig. 2, a predicted result is obtained, and the corresponding prediction accuracy is 95.3%.
According to the method, the photovoltaic power station with modeling conditions is used as a sample, the comprehensive prediction model of the photovoltaic power station can be effectively improved in prediction precision through strengthening training of the comprehensive prediction model of the photovoltaic power station, and finally the short-term generation power of the distributed photovoltaic power station group is improved. It is easy to understand that in the conventional statistical upscale distributed photovoltaic power station group short-term power generation power prediction process, a single typical photovoltaic power station is generally selected to participate in the statistical upscale prediction, so as to form a distributed photovoltaic power station group short-term power generation power prediction result. The number of samples is single, and the prediction accuracy of the distributed photovoltaic power station group short-term generation power cannot be guaranteed due to the influence of the prediction accuracy of a typical photovoltaic power station. In the process of selecting a typical photovoltaic power station, the full information of the photovoltaic power station with modeling conditions is fully utilized, and the comprehensive prediction model prediction technology of the photovoltaic power station is assisted, so that the interference caused by the extreme condition of a single photovoltaic power station can be avoided, and the power prediction precision of a distributed photovoltaic power station group can be guaranteed.
The embodiment provides a data enhancement-based distributed photovoltaic Fu Qun short-term power prediction system, which comprises a data acquisition module, a photovoltaic power station comprehensive prediction module, a correlation coefficient calculation module and a statistical upscale prediction module, wherein the data acquisition module is used for acquiring actual measurement values of photovoltaic power generation power of all photovoltaic power stations, the photovoltaic power station comprehensive prediction module is internally provided with a photovoltaic power station comprehensive prediction model, the photovoltaic power station comprehensive prediction model obtains a comprehensive prediction result according to the short-term photovoltaic power generation power prediction value of all the photovoltaic power stations after per unit treatment and the actual measurement values of the photovoltaic power generation power of all the photovoltaic power stations after per unit treatment, the correlation coefficient calculation module is used for calculating correlation coefficients between the historical power generation power of the typical photovoltaic power stations and the historical power generation power of a distributed photovoltaic power station group and correlation coefficients of all the typical photovoltaic power station historical power generation power and the prediction values, the correlation coefficient calculation module is used for calculating weight coefficients of the typical photovoltaic power stations in the photovoltaic power generation power prediction, and the statistical upscale prediction module is used for distributing different weights for all the typical photovoltaic power stations according to the weight coefficients, and calculating the short-term upscale method is used for calculating the power generation power prediction value of the distributed photovoltaic power station group.
In another embodiment, a non-volatile computer storage medium is provided having stored thereon computer executable instructions for performing the data enhanced distributed light Fu Qun short-term power prediction method of any of the embodiments described above.
The present embodiment also provides a computer program product comprising a computer program stored on a non-volatile computer storage medium, the computer program comprising program instructions which, when executed by a computer, cause the computer to perform the data enhanced distributed light Fu Qun short-term power prediction method of the above embodiments.
The present embodiment provides an electronic device including: at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform a data enhancement based distributed light Fu Qun short-term power prediction method.
According to the method, a data enhancement mode is adopted, the historical data training of the sample photovoltaic power station is enhanced, the prediction accuracy of the short-term power generation power of the sample photovoltaic power station is improved to the greatest extent, the short-term power generation power of the distributed photovoltaic power station group is improved, and the prediction result shows that the method is effective through calculation and comparison.
The above-described specific embodiments further illustrate the objects, technical solutions and technical effects of the present application in detail. It should be understood that the foregoing is only illustrative of the present application and is not intended to limit the scope of the application, and that all equivalent changes and modifications that may be made by those skilled in the art without departing from the spirit and principles of the application shall fall within the scope of the application.

Claims (5)

1. A data enhancement-based distributed light Fu Qun short-term power prediction method, which is characterized by comprising the following steps:
step one, carrying out per unit on a photovoltaic power generation power actual measurement value and a short-term photovoltaic power generation power prediction value of a photovoltaic power station with modeling conditions; analyzing the condition and the geographic position of an access bus among all distributed photovoltaic power stations, dividing the access bus into a plurality of clusters, and adopting the following formula for per unit of photovoltaic power generation:
；
；
wherein:represent the firstkThe capacity of the photovoltaic power station with modeling conditions; />Representing the actually measured value of the photovoltaic power generation power of the kth photovoltaic power station with modeling conditions in the t period; />Representing the actually measured value of the photovoltaic power generation power of the photovoltaic power station with modeling conditions in the kth period after per unit of the power generation system; />Representing a short-term photovoltaic power generation power predicted value of a kth photovoltaic power station with modeling conditions in a t period; />Representing the short-term photovoltaic power generation power predicted value of the modeling condition photovoltaic power station in the kth period after per unit of time,q is the number of photovoltaic power stations;
step two, training a comprehensive photovoltaic power station prediction model based on the historical data of each distributed photovoltaic power station after per unit, and obtaining a comprehensive prediction result by using the comprehensive photovoltaic power station prediction model;
based onqThe comprehensive prediction result of the comprehensive prediction model of the photovoltaic power station with modeling conditions is shown as follows:wherein->，/>，/>Representing the comprehensive predicted value of the comprehensive predicted model of the photovoltaic power station in the t period;the weight coefficient of the photovoltaic power station with modeling conditions is the kth photovoltaic power station;
target of comprehensive prediction model of photovoltaic power station: for the followingqPhotovoltaic power plant with modeling conditionsOptimizing a group of optimal weight coefficients to minimize the fitting error of a comprehensive prediction model of the photovoltaic power station, wherein the objective function is expressed as follows:wherein z represents a fitting error, n represents the number of periods, +.>Representing the sum of the actual photovoltaic power generation power values of the photovoltaic power station in the t period after per unit of the photovoltaic power station, and +.>；
Step three, naming the comprehensive prediction result and taking the result as a short-term generation power prediction value of a typical photovoltaic power stationThe method comprises the steps of carrying out a first treatment on the surface of the The comprehensive prediction result is named by adopting the following formula: />；
Step four, based on a short-term power generation power predicted value of a typical photovoltaic power station and historical power generation power of the typical photovoltaic power station, participating in statistical upscale prediction, and forming a short-term power generation power predicted result of a distributed photovoltaic power station group; the historical power generation power of the typical photovoltaic power station is formed by accumulating actual measured values of the historical photovoltaic power generation power of the photovoltaic power station with modeling conditions.
2. The method for data-enhanced distributed optical Fu Qun short-term power prediction as claimed in claim 1, wherein the fourth step comprises:
calculating a correlation coefficient between the historical power generation power of a typical photovoltaic power station and the actual measurement value of the historical power generation power of a distributed photovoltaic power station group：
；
Wherein:is thattHistorical power generation power of a typical photovoltaic power station in a period; />Is thattHistorical power generation actual measurement values of the time period distributed photovoltaic power station group; />Is the average value of historical power generation power of a typical photovoltaic power station in a time period; />Distributing the average value of the historical power generation actual measurement values of the photovoltaic power station group in a time period;
calculating weight coefficients of typical photovoltaic power stations in photovoltaic power generation power prediction to form a weight coefficient matrixWeight coefficient matrix->Expressed as: />，/>Is constant, matrix->The definition is as follows:
；
wherein:indicate->Typical photovoltaic power stations are 1->≤FF is the number of typical photovoltaic power stations; />Representing a correlation coefficient between the historical generated power of a typical photovoltaic power station and a predicted value; sign->Hadamard product representing matrix, +.>Indicate->Correlation coefficient between historical power generation power of typical photovoltaic power station and actual measurement value of historical power generation power of distributed photovoltaic power station group, +.>Indicate->The correlation coefficient of the historical power generation power and the predicted value of the typical photovoltaic power station;
according to the weight coefficient matrix, different weights are distributed for each typical photovoltaic power station, and based on a statistical upscaling method, the short-term generation power predicted value of the distributed photovoltaic power station group is expressed as follows:
；
wherein:a time sequence matrix of short-term generation power predicted values for the distributed photovoltaic power plant group; />A time series matrix of photovoltaic power generation power predicted values for a typical power station; the statistical parameters c and d are constants; e is an identity matrix; according to historical data, solving a calculation equation of a short-term power generation power predicted value of the distributed photovoltaic power station group by adopting a least square method, and calculating c and cThe value of d is then used for future short-term generated power prediction of the distributed photovoltaic power plant group.
3. The distributed photovoltaic Fu Qun short-term power prediction system for implementing the data enhancement-based distributed photovoltaic Fu Qun short-term power prediction method as claimed in claim 1 or 2 is characterized by comprising a data acquisition module, a photovoltaic power station comprehensive prediction module, a correlation coefficient calculation module and a statistical upscale prediction module, wherein the data acquisition module is used for acquiring actual measurement values of photovoltaic power generation power of each photovoltaic power station, the photovoltaic power station comprehensive prediction module is internally provided with a photovoltaic power station comprehensive prediction model, the photovoltaic power station comprehensive prediction model obtains a comprehensive prediction result according to the short-term photovoltaic power generation power prediction value of each photovoltaic power station after per unit treatment and the actual measurement value of the photovoltaic power generation power of each photovoltaic power station after per unit treatment, the correlation coefficient calculation module calculates correlation coefficients between historical power generation power of a typical photovoltaic power station and historical power generation power of each typical photovoltaic power station and the prediction value, the correlation coefficient calculation module also calculates weight coefficients of the typical photovoltaic power stations in the photovoltaic power generation power prediction, and the statistical upscale prediction module distributes different weights for each photovoltaic power station according to the weight coefficients, and calculates the statistical upscale prediction value based on the distributed photovoltaic power station short-term power generation power.
4. A non-transitory computer storage medium having stored thereon computer executable instructions for performing the data-enhanced distributed optical Fu Qun short-term power prediction method of claim 1 or 2.
5. An electronic device, comprising: at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions for execution by the at least one processor, wherein the instructions are executable by the at least one processor to enable the at least one processor to perform the data-enhanced distributed light Fu Qun short-term power prediction method of claim 1 or 2.
CN202310596490.5A 2023-05-25 2023-05-25 Data enhancement-based distributed light Fu Qun short-term power prediction method and system Active CN116316615B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310596490.5A CN116316615B (en) 2023-05-25 2023-05-25 Data enhancement-based distributed light Fu Qun short-term power prediction method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310596490.5A CN116316615B (en) 2023-05-25 2023-05-25 Data enhancement-based distributed light Fu Qun short-term power prediction method and system

Publications (2)

Publication Number Publication Date
CN116316615A CN116316615A (en) 2023-06-23
CN116316615B true CN116316615B (en) 2023-09-12

Family

ID=86783704

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310596490.5A Active CN116316615B (en) 2023-05-25 2023-05-25 Data enhancement-based distributed light Fu Qun short-term power prediction method and system

Country Status (1)

Country Link
CN (1) CN116316615B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116596162B (en) * 2023-07-05 2023-10-20 浙江正泰新能源开发有限公司 Photovoltaic power generation amount measuring and calculating method, device, equipment and storage medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102722760A (en) * 2012-05-28 2012-10-10 中国电力科学研究院 Regional power prediction method for photovoltaic power station group
CN109978242A (en) * 2019-03-12 2019-07-05 国网冀北电力有限公司 The photovoltaic power generation cluster power forecasting method and device of scale are risen based on statistics
CN109978236A (en) * 2019-03-04 2019-07-05 三峡大学 A kind of small power station's short term power fining prediction technique based on feature combination
CN110929953A (en) * 2019-12-04 2020-03-27 国网山东省电力公司电力科学研究院 Photovoltaic power station ultra-short term output prediction method based on cluster analysis
US11070056B1 (en) * 2020-03-13 2021-07-20 Dalian University Of Technology Short-term interval prediction method for photovoltaic power output

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102722760A (en) * 2012-05-28 2012-10-10 中国电力科学研究院 Regional power prediction method for photovoltaic power station group
CN109978236A (en) * 2019-03-04 2019-07-05 三峡大学 A kind of small power station's short term power fining prediction technique based on feature combination
CN109978242A (en) * 2019-03-12 2019-07-05 国网冀北电力有限公司 The photovoltaic power generation cluster power forecasting method and device of scale are risen based on statistics
CN110929953A (en) * 2019-12-04 2020-03-27 国网山东省电力公司电力科学研究院 Photovoltaic power station ultra-short term output prediction method based on cluster analysis
US11070056B1 (en) * 2020-03-13 2021-07-20 Dalian University Of Technology Short-term interval prediction method for photovoltaic power output

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于CatBoost算法的短期光伏功率预测方法;陈海宏 等;《浙江电力》;第42卷(第2期);第67-75页 *

Also Published As

Publication number Publication date
CN116316615A (en) 2023-06-23

Similar Documents

Publication Publication Date Title
Wang et al. A seasonal GM (1, 1) model for forecasting the electricity consumption of the primary economic sectors
CN113937763B (en) Wind power prediction method, device, equipment and storage medium
CN106779226B (en) Fan batch power prediction method based on mixed kernel machine learning
Jafarian-Namin et al. Forecasting the wind power generation using Box–Jenkins and hybrid artificial intelligence: A case study
CN106485093B (en) Based on the solar irradiance time series synthetic method for improving Markov chain
CN116316615B (en) Data enhancement-based distributed light Fu Qun short-term power prediction method and system
CN111612244B (en) QRA-LSTM-based method for predicting nonparametric probability of photovoltaic power before day
CN110766200A (en) Method for predicting generating power of wind turbine generator based on K-means mean clustering
CN112200377A (en) Photovoltaic medium-long term power generation capacity forecasting method and device based on SARIMAX model
CN111600300B (en) Robust optimal scheduling method considering wind power multivariate correlation ellipsoid set
CN111488896A (en) Distribution line time-varying fault probability calculation method based on multi-source data mining
CN115481918A (en) Active sensing and predictive analysis system for unit state based on source network load storage
CN112884601A (en) Electric power system operation risk assessment method based on weather regionalization strategy
CN113610296A (en) Method for predicting peak value of electrical load in region and method for planning investment of power grid
CN112307672A (en) BP neural network short-term wind power prediction method based on cuckoo algorithm optimization
CN116307139A (en) Wind power ultra-short-term prediction method for optimizing and improving extreme learning machine
CN115912334A (en) Method for establishing prediction model of output guarantee rate of wind power plant and prediction method
CN111861259B (en) Load modeling method, system and storage medium considering time sequence
Tascıkaraoglu Impacts of accurate renewable power forecasting on Optimum operation of power system
CN114492923A (en) Long-time scale power prediction method
CN115310648A (en) Medium-and-long-term wind power combination prediction method based on multi-meteorological variable model identification
CN113887839A (en) Clustering-based wind power plant ultra-short-term wind power prediction method and system
CN113487062A (en) Power load prediction method based on periodic automatic encoder
CN111368257A (en) Method and device for analyzing and predicting coal-to-electricity load characteristics
CN115049443B (en) Multi-scale feature-based adaptive regression meteorological medium and long term optimization method and system

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
GR01 Patent grant
GR01 Patent grant