Photovoltaic power prediction method

By combining a variety of photovoltaic power prediction methods and weight adjustment of meteorological factors, a weight table is generated, which solves the problem of low prediction accuracy of photovoltaic power generation power, and achieves higher prediction accuracy and stability.

CN119093336APending Publication Date: 2024-12-06YUNNAN POWER GRID CO LTD ELECTRIC POWER RES INST
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411127634.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The existing photovoltaic power prediction methods have limitations in generalization of a single algorithm and simple combination considerations, resulting in low prediction accuracy and unstable prediction accuracy.

Method used

A photovoltaic power prediction method is adopted. By obtaining the current forecast meteorological data within the preset time period, matching the corresponding weight table, combining with a variety of photovoltaic power prediction methods with higher accuracy, different weights are given according to meteorological factors, and iterative adjustment weights are added. The value is generated to generate a preset weight table for predicting the target power of the photovoltaic electric field.

Benefits of technology

The prediction accuracy and stability of photovoltaic power generation is improved, and the prediction accuracy instability that may occur in a single algorithm or a simple combination algorithm is avoided, ensuring the adaptability and robustness of the prediction results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119093336A_ABST
    Figure CN119093336A_ABST
Patent Text Reader

Abstract

The embodiment of the invention discloses a photovoltaic power prediction method, which comprises the following steps of: matching weights of various photovoltaic power prediction methods corresponding to meteorological factors in a preset weight table according to different meteorological factors by utilizing various existing photovoltaic power prediction methods with relatively high precision; then, various photovoltaic power prediction methods are endowed with different weights, and finally, the target prediction photovoltaic power of the photovoltaic electric field is predicted according to the current prediction photovoltaic power predicted by the various photovoltaic power prediction methods and the weights of the various photovoltaic power prediction methods. According to the method, the weight advantages of various existing photovoltaic power prediction methods with relatively high precision are combined, so that the problems of limitation of scene generalization of a single algorithm and low prediction precision of the photovoltaic power generation power caused by simple combination and consideration of one-sidedness can be effectively solved; therefore, the prediction precision of the photovoltaic power generation power can be effectively improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a photovoltaic power prediction method. Background Art

[0002] Driven by the "dual carbon" goal, the electric energy system is undergoing a profound transformation towards energy conservation and low carbon. Photovoltaic technology, with its mature technology and large-scale and commercial development potential, has become a leader in the field of new energy power generation. my country's photovoltaic industry is booming, and the installed capacity of centralized photovoltaics and distributed photovoltaics has shown a rapid growth trend. Photovoltaics has become the second largest power source after thermal power, and its position in the electric energy structure is becoming increasingly prominent.

[0003] However, large-scale photovoltaic power generation connected to the grid also brings a series of challenges. Among them, the most prominent is the real-time random fluctuation of photovoltaic power generation, which is mainly affected by meteorological factors. When a large number of photovoltaic power plants are connected to the grid, this fluctuation will have a significant impact on the power quality of the grid, posing a severe challenge to the safe and stable operation of the power system.

[0004] In order to deal with the random volatility of photovoltaic power generation, it is necessary to predict the photovoltaic power generation power of the photovoltaic field. Therefore, domestic and foreign scholars have conducted a lot of research on photovoltaic power generation prediction and achieved certain results.

[0005] However, the current photovoltaic power generation prediction is based on a single algorithm or a simple combination of several different algorithms. Due to the limitations of the generalization of a single algorithm scenario and the one-sidedness of the simple combination, the prediction accuracy of photovoltaic power generation is not high. Summary of the invention

[0006] Based on this, it is necessary to propose a photovoltaic power prediction method to address the above problems, which can effectively solve the problem of low prediction accuracy of photovoltaic power generation due to the limitations of generalization of single algorithm scenarios and the one-sidedness of simple combination considerations. That is, this method can effectively improve the prediction accuracy of photovoltaic power generation.

[0007] To achieve the above object, the present invention provides a photovoltaic power prediction method in a first aspect, the method comprising:

[0008] Get the current forecast weather data within the preset time period;

[0009] According to the current forecast meteorological data, weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data are matched in a preset weight table;

[0010] Using each preset photovoltaic power prediction method, predicting the current predicted photovoltaic power of each preset photovoltaic power prediction method at each time point within a preset time period according to the current forecast meteorological data;

[0011] The target predicted photovoltaic power of the photovoltaic field at each time point within the preset time period is determined according to the current predicted photovoltaic power of each preset photovoltaic power prediction method at each time point within the preset time period and the weight of each preset photovoltaic power prediction method.

[0012] Optionally, matching the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data in a preset weight table according to the current forecast meteorological data includes:

[0013] According to the current forecast meteorological data, the historical forecast meteorological data with the highest similarity to the current forecast meteorological data is matched in the preset weight table, and the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data with the highest similarity are used as the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data.

[0014] Optionally, the method further comprises:

[0015] Acquiring historical forecast meteorological data within a plurality of preset time periods, and historical actual photovoltaic power of the photovoltaic field at each time point within the plurality of preset time periods;

[0016] Using each preset photovoltaic power prediction method, predicting the historical predicted photovoltaic power of each preset photovoltaic power prediction method at each time point in the nth preset time period according to the historical forecast meteorological data in the nth preset time period;

[0017] Determine the prediction error of each preset photovoltaic power prediction method at each time point in the n preset time period according to the historical predicted photovoltaic power of each preset photovoltaic power prediction method at each time point in the n preset time period and the historical actual photovoltaic power of the photovoltaic electric field at each time point in the n preset time period;

[0018] Iterate the iteration weight of each preset photovoltaic power prediction method according to the prediction error of each preset photovoltaic power prediction method at each time point in the nth preset time period, and in the iterative process, determine the combined prediction value according to the iteration weights of all preset photovoltaic power prediction methods of the current iteration and the historical predicted photovoltaic power at the time point corresponding to the nth preset time period, and when the combined prediction value meets the preset iteration condition, determine the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in the nth preset time period by the iteration weights of all preset photovoltaic power prediction methods of the current iteration;

[0019] Wherein, n successively takes integers greater than 0 to obtain weights of all preset photovoltaic power prediction methods corresponding to historical forecast meteorological data within multiple preset time periods;

[0020] The preset weight table is generated according to the historical forecast meteorological data within a plurality of preset time periods and the weights of all corresponding preset photovoltaic power prediction methods.

[0021] Optionally, the acquiring of historical forecast meteorological data within a plurality of preset time periods, and the historical actual photovoltaic power of the photovoltaic field at each time point within the plurality of preset time periods, comprises:

[0022] Acquiring initial historical forecast meteorological data and initial historical actual photovoltaic power of the photovoltaic field;

[0023] Clustering the initial historical meteorological forecast data to obtain historical meteorological forecast data within a plurality of preset time periods;

[0024] The initial historical actual photovoltaic power of the photovoltaic field is divided according to historical forecast meteorological data in multiple preset time periods to obtain the historical actual photovoltaic power of the photovoltaic field at each time point in the multiple preset time periods.

[0025] Optionally, the iterating the iteration weight of each preset photovoltaic power prediction method according to the prediction error of each preset photovoltaic power prediction method at each time point in the nth preset time period includes:

[0026] Using the formula Iterating the iteration weights of each preset photovoltaic power prediction method;

[0027] in, is the initial iteration weight, M is the number of preset photovoltaic power prediction methods, to are the iteration weights of the mth preset photovoltaic power prediction method from the 1st iteration to the Dth iteration, α is the smoothing parameter, α∈[0,1], E m,n,t1 To E m,n,tT is the prediction error of the mth preset photovoltaic power prediction method from the 1st time point to the Tth time point within the nth preset time period.

[0028] Optionally, the preset iteration condition includes: a difference between the combined prediction value and the historical actual photovoltaic power at a time point corresponding to the combined prediction value within the nth preset time period is less than or equal to a maximum allowable error value.

[0029] Optionally, the determining of the combined prediction value according to the iteration weights of all preset photovoltaic power prediction methods of the current iteration and the historically predicted photovoltaic powers at the time point corresponding to the nth preset time period includes:

[0030] The iteration weights of each preset photovoltaic power prediction method of the current iteration are normalized to obtain the standard weights of each preset photovoltaic power prediction method;

[0031] Determine a combined prediction value according to the standard weight of each preset photovoltaic power prediction method and the historical predicted photovoltaic power at the time point corresponding to the nth preset time period;

[0032] The step of determining the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in the nth preset time period by using the iteration weights of all preset photovoltaic power prediction methods of the current iteration includes:

[0033] The iteration weights of each preset photovoltaic power prediction method of the current iteration are normalized to obtain the standard weights of each preset photovoltaic power prediction method;

[0034] The standard weights of all preset photovoltaic power prediction methods are used as weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in the nth preset time period.

[0035] Optionally, the iterative weights of each preset photovoltaic power prediction method of the current iteration are normalized to obtain a standard weight of each preset photovoltaic power prediction method, including:

[0036] Determine a maximum iteration weight and a minimum iteration weight among the iteration weights of all preset photovoltaic power prediction methods of the current iteration;

[0037] The standard weight of each preset photovoltaic power prediction method is determined according to the iteration weight of each preset photovoltaic power prediction method in the current iteration, the maximum iteration weight and the minimum iteration weight.

[0038] Optionally, determining a standard weight of each preset photovoltaic power prediction method according to the iteration weight of each preset photovoltaic power prediction method of the current iteration, the maximum iteration weight, and the minimum iteration weight includes:

[0039] Using the formula Determining a standard weight for each preset photovoltaic power prediction method;

[0040] The method of determining the combined prediction value according to the standard weight of each preset photovoltaic power prediction method and the historical predicted photovoltaic power at the time point corresponding to the preset photovoltaic power prediction method in the nth preset time period includes:

[0041] Using the formula Determining a combined forecast value;

[0042] in, is the standard weight of the mth preset photovoltaic power prediction method, is the iteration weight of the mth preset photovoltaic power prediction method in the current iteration, is the maximum iteration weight, is the minimum iteration weight, P d is the combined prediction value, M is the number of types of preset photovoltaic power prediction methods, is the historical predicted photovoltaic power at time point t corresponding to the standard weight of the mth preset photovoltaic power prediction method within the nth preset time period.

[0043] Optionally, determining the target predicted photovoltaic power of the photovoltaic field at each time point within the preset time period according to the current predicted photovoltaic power at each time point within the preset time period by each preset photovoltaic power prediction method and the weight of each preset photovoltaic power prediction method includes:

[0044] Using the formula Determining a target predicted photovoltaic power of the photovoltaic electric field at each time point within a preset time period;

[0045] Among them, P target,m,t is the target predicted photovoltaic power of the photovoltaic field at the tth time point within the preset time period, M is the number of types of preset photovoltaic power prediction methods, and W m ′ is the weight of the mth preset photovoltaic power prediction method, P m,t It is the current predicted photovoltaic power at the tth time point within the preset time period using the mth preset photovoltaic power prediction method.

[0046] To achieve the above object, the present invention provides a photovoltaic power prediction device in a second aspect, the device comprising:

[0047] An acquisition module is used to obtain the current forecast meteorological data within a preset time period;

[0048] A matching module, used for matching the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data in a preset weight table according to the current forecast meteorological data;

[0049] A prediction module, configured to use each preset photovoltaic power prediction method to predict the current predicted photovoltaic power of each preset photovoltaic power prediction method at each time point within a preset time period according to the current forecast meteorological data;

[0050] The determination module is used to determine the target predicted photovoltaic power of the photovoltaic field at each time point within the preset time period according to the current predicted photovoltaic power at each time point within the preset time period of each preset photovoltaic power prediction method and the weight of each preset photovoltaic power prediction method.

[0051] To achieve the above-mentioned object, the present invention provides in a third aspect a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the method as described in any one of the first aspects.

[0052] To achieve the above-mentioned purpose, the present invention provides a computer device in a fourth aspect, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the method as described in any one of the first aspects.

[0053] The embodiment of the present invention has the following beneficial effects: the above method obtains the current forecast meteorological data within a preset time period, then matches the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data in a preset weight table according to the current forecast meteorological data, and then uses each preset photovoltaic power prediction method to predict the current predicted photovoltaic power of each preset photovoltaic power prediction method at each time point within the preset time period according to the current forecast meteorological data, and finally determines the target predicted photovoltaic power of the photovoltaic field at each time point within the preset time period according to the current predicted photovoltaic power of each preset photovoltaic power prediction method at each time point within the preset time period and the weight of each preset photovoltaic power prediction method; that is, by using Various existing photovoltaic power prediction methods with high accuracy are used, and the weights of various photovoltaic power prediction methods corresponding to meteorological factors are matched in a preset weight table according to different meteorological factors, and then different weights are assigned to various photovoltaic power prediction methods. Finally, the target predicted photovoltaic power of the photovoltaic field is predicted according to the current predicted photovoltaic power predicted by various photovoltaic power prediction methods and the weights of various photovoltaic power prediction methods. Since this method combines the advantages of the weights of various existing photovoltaic power prediction methods with high accuracy, it can effectively solve the problem of low prediction accuracy of photovoltaic power generation due to the limitations of generalization of single algorithm scenarios and the one-sidedness of simple combination considerations, that is, this method can effectively improve the prediction accuracy of photovoltaic power generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0055] in:

[0056] Figure 1 A schematic diagram of a photovoltaic power prediction method in an embodiment of the present application;

[0057] Figure 2 A schematic diagram of a photovoltaic power prediction device in an embodiment of the present application;

[0058] Figure 3 1 is a diagram of the internal structure of a computer device in some embodiments. DETAILED DESCRIPTION

[0059] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] Driven by the "dual carbon" goal, the electric energy system is undergoing a profound transformation towards energy conservation and low carbon. Photovoltaic technology, with its mature technology and large-scale and commercial development potential, has become a leader in the field of new energy power generation. my country's photovoltaic industry is booming, and the installed capacity of centralized photovoltaics and distributed photovoltaics has shown a rapid growth trend. Photovoltaics has become the second largest power source after thermal power, and its position in the electric energy structure is becoming increasingly prominent.

[0061] However, large-scale photovoltaic power generation connected to the grid also brings a series of challenges. Among them, the most prominent is the real-time random fluctuation of photovoltaic power generation, which is mainly affected by meteorological factors. When a large number of photovoltaic power plants are connected to the grid, this fluctuation will have a significant impact on the power quality of the grid, posing a severe challenge to the safe and stable operation of the power system.

[0062] In order to deal with the random volatility of photovoltaic power generation, it is necessary to predict the photovoltaic power generation power of the photovoltaic field. Therefore, domestic and foreign scholars have conducted a lot of research on photovoltaic power generation prediction and achieved certain results.

[0063] However, the current photovoltaic power generation prediction is based on a single algorithm or a simple combination of several different algorithms. Due to the limitations of the generalization of a single algorithm scenario and the one-sidedness of the simple combination, the prediction accuracy of photovoltaic power generation is not high.

[0064] The photovoltaic power prediction studied before this application also involves the use of various existing photovoltaic power prediction methods with relatively high accuracy, assigning different weights to various photovoltaic power prediction methods according to their overall performance (such as accuracy), and finally predicting the target predicted photovoltaic power of the photovoltaic field according to the current predicted photovoltaic power predicted by the various photovoltaic power prediction methods and the weights of the various photovoltaic power prediction methods. However, since this method only assigns different weights to various photovoltaic power prediction methods according to their overall performance, it has instability in prediction accuracy. The prediction accuracy may be high at this moment, while the prediction accuracy may be low at the next moment, and it cannot effectively improve the prediction accuracy of photovoltaic power.

[0065] In response to the above problems, the present application proposes a photovoltaic power prediction method, which can effectively solve the problem of low prediction accuracy of photovoltaic power generation due to the limitations of generalization of a single algorithm scenario and the one-sidedness of simple combination considerations, that is, this method can effectively improve the prediction accuracy of photovoltaic power generation; in addition, the prediction accuracy of this method is highly stable, and there will be no problem of high prediction accuracy at this moment and low prediction accuracy at the next moment, that is, this method can truly effectively improve the prediction accuracy of photovoltaic power generation, and the specific implementation principle will be described in detail in the following embodiments.

[0066] See also Figure 1 , is a schematic diagram of a photovoltaic power prediction method in an embodiment of the present application, the method comprising:

[0067] Step 110: Obtain current forecast weather data within a preset time period.

[0068] Among them, the preset time period can be obtained by the operator based on a lot of experience, experiments or statistics, and of course, can also be set by the operator according to actual needs; the current forecast meteorological data refers to various meteorological factors that affect the fluctuation of photovoltaic power generation within the preset time period.

[0069] In some embodiments, the preset time period can be set to a time period of a season, a time period of a month, or a time period of a day, which is not limited here.

[0070] In some embodiments, the current forecast meteorological data includes but is not limited to temperature, relative air humidity, air pressure, wind speed, wind direction, surface horizontal radiation, normal direct radiation, scattered radiation, etc. Since there are many meteorological factors that affect the fluctuation of photovoltaic power generation power, they are no longer listed one by one here. As long as they are meteorological factors that can affect the fluctuation of photovoltaic power generation power, they can be used; of course, the current forecast meteorological data can also include only at least one of them, which is not limited here.

[0071] Step 120: matching the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data in a preset weight table according to the current forecast meteorological data.

[0072] Among them, the preset weight table refers to a data table that stores various historical forecast meteorological data and the weights of various preset photovoltaic power prediction methods corresponding to each historical forecast meteorological data; various preset photovoltaic power prediction methods refer to various existing photovoltaic power prediction methods with higher accuracy.

[0073] In some embodiments, the preset weight table may be obtained by an operator based on a large amount of experience, experiments or statistics.

[0074] In some embodiments, the operator may collect various photovoltaic power prediction methods existing in the prior art, and then determine various photovoltaic power prediction methods with higher prediction accuracy through experiments or simulations to serve as various preset photovoltaic power prediction methods of the present application.

[0075] Step 130: using each preset photovoltaic power prediction method, predicting the current predicted photovoltaic power of each preset photovoltaic power prediction method at each time point within a preset time period according to the current forecast meteorological data.

[0076] Herein, each time point refers to each moment.

[0077] In some embodiments, the time points may be continuous time points or spaced time points; for example, the time interval between the first time point and the second time point may be 0, or the time interval between the first time point and the second time point may be 15 minutes, which is not limited here.

[0078] Step 140: determining the target predicted photovoltaic power of the photovoltaic field at each time point within the preset time period according to the current predicted photovoltaic power at each time point within the preset time period by each preset photovoltaic power prediction method and the weight of each preset photovoltaic power prediction method.

[0079] In an embodiment of the present application, various existing photovoltaic power prediction methods with higher accuracy are utilized, and the weights of various photovoltaic power prediction methods corresponding to meteorological factors are matched in a preset weight table according to different meteorological factors, and then different weights are assigned to various photovoltaic power prediction methods, and finally the target predicted photovoltaic power of the photovoltaic field is predicted according to the current predicted photovoltaic power predicted by various photovoltaic power prediction methods and the weights of various photovoltaic power prediction methods. Since this method combines the advantages of the weights of various existing photovoltaic power prediction methods with higher accuracy, it can effectively solve the problem of low prediction accuracy of photovoltaic power generation due to the limitations of generalization of single algorithm scenarios and the one-sidedness of simple combination considerations, that is, this method can effectively improve the prediction accuracy of photovoltaic power generation.

[0080] In addition, the prediction accuracy of this method is highly stable, and there will not be a problem of high prediction accuracy at one moment and low prediction accuracy at the next moment. That is, this method can truly and effectively improve the prediction accuracy of photovoltaic power generation.

[0081] In a feasible implementation method, step 120 in the above embodiment matches the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data in the preset weight table according to the current forecast meteorological data, including: matching the historical forecast meteorological data with the highest similarity to the current forecast meteorological data in the preset weight table according to the current forecast meteorological data, and using the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data with the highest similarity as the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data.

[0082] Among them, the historical forecast meteorological data is similar to the current forecast meteorological data, and also refers to various meteorological factors that affect the fluctuation of photovoltaic power generation power; for the explanation of the historical forecast meteorological data, please refer to the relevant content of the current forecast meteorological data in the above embodiment, which will not be repeated here.

[0083] Regarding the method for determining the historical meteorological forecast data with the highest similarity to the current meteorological forecast data, in some embodiments, the difference between each meteorological factor in the current meteorological forecast data and each meteorological factor in the nth historical meteorological forecast data can be first determined, wherein n successively takes integers greater than 0, until the difference corresponding to each meteorological factor in all historical meteorological forecast data is obtained, and the historical meteorological forecast data with the smallest difference corresponding to each meteorological factor is taken as the historical meteorological forecast data with the highest similarity; of course, in other embodiments, it can also be slightly adjusted, and after obtaining the difference corresponding to each meteorological factor in all historical meteorological forecast data, the historical meteorological forecast data with the smallest sum of the differences corresponding to all meteorological factors is taken as the historical meteorological forecast data with the highest similarity; or, after obtaining the difference corresponding to each meteorological factor in all historical meteorological forecast data, the historical meteorological forecast data with the smallest average value of the differences corresponding to all meteorological factors is taken as the historical meteorological forecast data with the highest similarity; the present application is not limited here.

[0084] In an embodiment of the present application, by matching the historical forecast meteorological data with the highest similarity to the current forecast meteorological data in a preset weight table according to the current forecast meteorological data, and using the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data with the highest similarity as the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data, the prediction accuracy can be effectively improved, the stability can be enhanced, and it has the advantage of strong adaptability.

[0085] It can be understood that the prediction accuracy is improved: since the weights corresponding to the historical forecast meteorological data that is most similar to the current forecast meteorological data are used, the performance differences of various photovoltaic power prediction methods under the current forecast meteorological data can be more accurately reflected, and then different photovoltaic power prediction methods are given different weights, which can comprehensively consider the advantages of various photovoltaic power prediction methods, thereby improving the prediction accuracy; Enhanced stability: Since the weights are obtained based on similar historical forecast meteorological data, they have certain stability and reliability. Even when the forecast meteorological data undergoes certain changes, the weights under similar forecast meteorological data are used to maintain relatively stable prediction performance, avoiding the problem of unstable prediction accuracy that may exist in a single algorithm or a simple combination algorithm; Strong adaptability: This method does not rely on specific forecast meteorological data, but searches for similar situations in all historical forecast meteorological data based on the current forecast meteorological data, and determines the weights of various photovoltaic power prediction methods accordingly. Therefore, this method has strong adaptability and can cope with the photovoltaic power generation power prediction problem under different forecast meteorological data.

[0086] In addition, considering that the power generation efficiency of the photovoltaic panel is affected by the use time, in some embodiments, it is also necessary to regularly update the preset weight table to determine the accuracy and validity of its data.

[0087] In a feasible implementation, the method in the above embodiment also includes: obtaining historical forecast meteorological data within multiple preset time periods, and historical actual photovoltaic power of the photovoltaic field at each time point within the multiple preset time periods; using each preset photovoltaic power prediction method to predict the historical predicted photovoltaic power of each preset photovoltaic power prediction method at each time point within the n preset time period according to the historical forecast meteorological data within the nth preset time period; determining the prediction error of each preset photovoltaic power prediction method at each time point within the n preset time period according to the historical predicted photovoltaic power at each time point within the n preset time period and the historical actual photovoltaic power of the photovoltaic field at each time point within the n .... The prediction error of the intermediate point iterates the iteration weight of each preset photovoltaic power prediction method, and in the iterative process, the combined prediction value is determined according to the iteration weights of all the preset photovoltaic power prediction methods of the current iteration and the historical predicted photovoltaic power at the time point corresponding to the nth preset time period, and when the combined prediction value meets the preset iteration condition, the iteration weights of all the preset photovoltaic power prediction methods of the current iteration are used to determine the weights of all the preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in the nth preset time period; wherein n takes integers greater than 0 in turn, and the weights of all the preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in multiple preset time periods are obtained; and a preset weight table is generated according to the historical forecast meteorological data in multiple preset time periods and the weights of all the corresponding preset photovoltaic power prediction methods.

[0088] For the preset time period and the explanation of each time point, please refer to the relevant content in the above embodiment, which will not be repeated here;

[0089] The preset iteration conditions may be obtained by the operator based on a large amount of experience, experiments or statistics, and of course, may also be set by the operator based on actual needs.

[0090] In some embodiments, it is assumed that the preset time periods are divided by seasons, that is, there are 4 preset time periods, the first preset time period is spring, the second preset time period is summer, the third preset time period is autumn, and the third preset time period is winter; of course, it can also be divided by month or by day, and this application is not limited here.

[0091] In some embodiments, a weight iteration method based on the inverse of the error may be used to iterate the iteration weight of each preset photovoltaic power prediction method according to the prediction error of each preset photovoltaic power prediction method at each time point in the nth preset time period.

[0092] In an embodiment of the present application, various photovoltaic power prediction methods are used to predict historical predicted photovoltaic power based on each historical forecast meteorological data, and then the weight of each photovoltaic power prediction method is iterated based on the prediction error between the historical predicted photovoltaic power and the historical actual photovoltaic power data, so as to assign weights to each photovoltaic power prediction method, and then generate a preset weight table, which can effectively improve the prediction accuracy, enhance adaptability and robustness.

[0093] It can be understood that the prediction accuracy is improved: by iteratively evaluating the performance of each photovoltaic power prediction method based on historical forecast meteorological data, and adjusting the weight of each photovoltaic power prediction method according to the prediction error, a more accurate weight table is finally obtained, which is helpful in subsequent predictions. According to the current forecast meteorological data, the appropriate combination weight of photovoltaic power prediction methods is selected, thereby improving the accuracy of the prediction; Enhance adaptability and robustness: Since the preset weight table is constructed based on historical forecast meteorology, it can adapt to changes in various forecast meteorological data. When encountering new forecast meteorological data, the weight of the appropriate photovoltaic power prediction method can be quickly selected or combined according to the preset weight table, thereby enhancing the adaptability and robustness of the prediction;

[0094] In addition, this method can simplify the prediction process, that is, once the preset weight table is constructed and optimized, the prediction process can be greatly simplified. It is only necessary to find the corresponding weights in the preset weight table according to the current forecast meteorological data, and then select or combine the weights of the photovoltaic power prediction method according to these weights. This greatly reduces the amount of calculation and complexity in the prediction process; as well as the possibility of continuous updating and improvement: since the preset weight table is constructed based on historical forecast meteorological data, it can be continuously updated and optimized with the arrival of new forecast meteorological data, which means that the prediction performance can be continuously improved over time, thereby further improving the prediction accuracy.

[0095] In a feasible implementation method, the acquisition of historical forecast meteorological data within multiple preset time periods and the historical actual photovoltaic power of the photovoltaic field at each time point within the multiple preset time periods in the above-mentioned embodiment includes: acquiring initial historical forecast meteorological data and the initial historical actual photovoltaic power of the photovoltaic field; clustering and dividing the initial historical forecast meteorological data to obtain the historical forecast meteorological data within multiple preset time periods; dividing the initial historical actual photovoltaic power of the photovoltaic field according to the historical forecast meteorological data within multiple preset time periods to obtain the historical actual photovoltaic power of the photovoltaic field at each time point within the multiple preset time periods.

[0096] The initial historical forecast meteorological data and the initial historical actual photovoltaic power may be the data of the photovoltaic field in the previous year, or the data in the previous two years, or the data in the previous three years, etc., which are not limited here.

[0097] In some embodiments, in order to adapt to the changes in various meteorological conditions, the present application may adopt a clustering method to divide the initial historical forecast meteorological data into historical forecast meteorological data within multiple preset time periods, wherein the multiple preset time periods may include but are not limited to time periods divided by seasonality, time periods divided by months, and time periods divided by days, etc.; further, the initial historical actual photovoltaic power of the photovoltaic field can be divided into corresponding time periods according to the historical forecast meteorological data within multiple preset time periods, that is, the initial historical actual photovoltaic power can be divided correspondingly by seasonality, by month, by day, etc., so as to divide the initial historical actual photovoltaic power into the historical actual photovoltaic power at each time point within multiple preset time periods.

[0098] It should be noted that before clustering the initial historical forecast meteorological data to obtain the historical forecast meteorological data within multiple preset time periods, it is also necessary to use the random forest method to preprocess the missing values ​​and outliers of the initial historical forecast meteorological data and the initial historical actual photovoltaic power to avoid the decrease in prediction accuracy due to the influence of missing values ​​and outliers.

[0099] In an embodiment of the present application, the initial historical forecast meteorological data is clustered to obtain historical forecast meteorological data within multiple preset time periods, and then the initial historical actual photovoltaic power of the photovoltaic field is divided according to the historical forecast meteorological data within the multiple preset time periods to obtain the historical actual photovoltaic power of the photovoltaic field at each time point within the multiple preset time periods, thereby achieving the rationality of data grouping and improving prediction accuracy.

[0100] It can be understood that the rationality of data grouping: through clustering, the initial historical forecast meteorological data can be grouped according to the similarity of time periods to form historical forecast meteorological data within multiple preset time periods. This grouping method can ensure that the forecast meteorological data within each time period has similar characteristics, thereby providing more targeted data support for subsequent photovoltaic power prediction; improving prediction accuracy: according to the historical forecast meteorological data within multiple preset time periods after clustering, the initial historical actual photovoltaic power of the photovoltaic field is divided, and the historical actual photovoltaic power of the photovoltaic field at each time point within multiple preset time periods can be obtained. Since this division is based on the similarity of meteorological data, it can ensure that the historical actual photovoltaic power data in each time period is highly correlated with the meteorological data in that time period, thereby improving the accuracy of subsequent predictions.

[0101] In a feasible implementation, the iterative weight of each preset photovoltaic power prediction method is iterated according to the prediction error of each preset photovoltaic power prediction method at each time point in the nth preset time period in the above embodiment, including:

[0102] Using the formula Iterating the iteration weights of each preset photovoltaic power prediction method;

[0103] in, is the initial iteration weight, M is the number of preset photovoltaic power prediction methods, to are the iteration weights of the mth preset photovoltaic power prediction method from the 1st iteration to the Dth iteration, α is the smoothing parameter, α∈[0,1], E m,n,t1 To E m,n,tT is the prediction error of the mth preset photovoltaic power prediction method from the 1st time point to the Tth time point within the nth preset time period.

[0104] In the embodiments of the present application, a rigorous iterative calculation formula for weights based on an exponentially weighted moving average of the inverse of the error is provided from a mathematical perspective. The accuracy of the iterative weights of the various photovoltaic power prediction methods that are iterated can be ensured from the rigor of the mathematical logic, and the above-mentioned iterative calculation formula for weights is preferably shown to provide reference, understanding and calculation for technical personnel; in addition, by adopting the above-mentioned iterative calculation formula for weights, a more reasonable combination of iterative weights of various photovoltaic power prediction methods can be obtained, which can effectively improve the prediction accuracy and enhance the adaptability and robustness.

[0105] In a feasible implementation, the preset iteration condition in the above embodiment includes: the difference between the combined prediction value and the historical actual photovoltaic power at the time point corresponding to the nth preset time period is less than or equal to the maximum allowable error value.

[0106] The maximum allowable error value may be obtained by the operator based on a large amount of experience, experiments or statistics. Of course, it may also be set by the operator according to actual needs, which is not limited here.

[0107] It should be noted that if the combined prediction value is obtained by the 8th iteration of the nth preset time period, that is, the prediction error of all preset photovoltaic power prediction methods at the 8th time point within the nth preset time period, then correspondingly, the historical actual photovoltaic power at the time point corresponding to the nth preset time period refers to the historical actual photovoltaic power at the 8th time point within the nth preset time period.

[0108] In an embodiment of the present application, by using the difference between the combined prediction value and the historical actual photovoltaic power at the time point corresponding to the nth preset time period to be less than or equal to the maximum allowable error value as a preset iteration condition, it is possible to ensure prediction accuracy, optimize weight allocation, improve stability, and enhance prediction reliability.

[0109] It can be understood that ensuring prediction accuracy: the preset iteration condition sets a maximum allowable error value to ensure that the difference between the combined prediction value and the actual photovoltaic power does not exceed this threshold, which helps to ensure the accuracy of the prediction results and makes the prediction value closer to the actual photovoltaic power output; optimizing weight allocation: during the iteration process, when the combined prediction value meets the preset iteration condition, the current weight allocation is deemed valid and is used for subsequent photovoltaic power predictions. This error-based weight iteration optimization can ensure that each photovoltaic power prediction method obtains a reasonable weight allocation based on its prediction performance, thereby improving the accuracy of the prediction; improving stability: through the restrictions of the preset iteration conditions, the photovoltaic power prediction method and weight allocation can be automatically adjusted during the prediction process to adapt to different meteorological conditions and photovoltaic electric field states. This adaptability enables the prediction to maintain stable prediction performance in the face of complex and changing environments; enhancing prediction reliability: by ensuring that the error between the predicted value and the actual value is within an acceptable range, the preset iteration condition can enhance the reliability of the prediction results.

[0110] In a feasible implementation method, the above-mentioned embodiment determines the combined prediction value based on the iteration weights of all preset photovoltaic power prediction methods of the current iteration and the historical predicted photovoltaic power at the time point corresponding to the nth preset time period, including: normalizing the iteration weights of each preset photovoltaic power prediction method of the current iteration to obtain the standard weight of each preset photovoltaic power prediction method; determining the combined prediction value based on the standard weight of each preset photovoltaic power prediction method and the historical predicted photovoltaic power at the time point corresponding to the nth preset time period.

[0111] In the above embodiment, the iterative weights of all preset photovoltaic power prediction methods in the current iteration are used to determine the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in the nth preset time period, including: normalizing the iterative weights of each preset photovoltaic power prediction method in the current iteration to obtain the standard weight of each preset photovoltaic power prediction method; and using the standard weights of all preset photovoltaic power prediction methods as the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in the nth preset time period.

[0112] In an embodiment of the present application, a standard weight for each preset photovoltaic power prediction method is obtained through normalization processing, and the standard weights of all preset photovoltaic power prediction methods are used as the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data within the nth preset time period, and the standard weights of all preset photovoltaic power prediction methods are used as the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data within the nth preset time period, which can optimize the prediction accuracy, enhance the prediction stability and improve the rationality of the weight update.

[0113] It can be understood that the prediction accuracy is optimized: by normalizing the iterative weights of each preset photovoltaic power prediction method, the obtained standard weights can ensure that the relative importance of each photovoltaic power prediction method in the combined prediction is reasonably reflected, which helps to give higher weights to photovoltaic power prediction methods with better performance, thereby improving the accuracy and reliability of the combined prediction; enhancing the prediction stability: using the normalized standard weights to determine the combined prediction value can ensure that the prediction results are not affected by the extreme weights of individual photovoltaic power prediction methods. This processing method enhances the stability of the prediction and makes the prediction results more reliable; improving the rationality of weight update: the iterative weights of the current iteration are directly used as the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in the nth preset time period after normalization, ensuring the continuity and consistency of the weight update. This processing method makes the weight update more reasonable and can reflect the impact of forecast meteorological data in different time periods on photovoltaic power prediction.

[0114] In a feasible implementation method, the iteration weights of each preset photovoltaic power prediction method in the current iteration in the above embodiment are normalized to obtain the standard weight of each preset photovoltaic power prediction method, including: determining the maximum iteration weight and the minimum iteration weight among the iteration weights of all preset photovoltaic power prediction methods in the current iteration; determining the standard weight of each preset photovoltaic power prediction method according to the iteration weight, the maximum iteration weight and the minimum iteration weight of each preset photovoltaic power prediction method in the current iteration.

[0115] In an embodiment of the present application, by normalizing the iterative weights of each preset photovoltaic power prediction method in the current iteration, a standard weight for each preset photovoltaic power prediction method is obtained, which can ensure the rationality of the weights, optimize the prediction performance, and improve stability.

[0116] It can be understood that ensuring the rationality of weights: normalization processing can ensure that the sum of the weights of all preset photovoltaic power prediction methods is 1, which makes each photovoltaic power prediction method have a clear contribution ratio to the combined prediction results, which helps to understand the importance of each photovoltaic power prediction method in the prediction and can be adjusted as needed; optimizing prediction performance: the normalized standard weights can reflect the performance advantages and disadvantages of different photovoltaic power prediction methods in the current iteration. Photovoltaic power prediction methods with better performance will obtain higher weights, thereby playing a greater role in the combined prediction. This performance-based weight allocation helps to improve the accuracy of the prediction; improving stability: normalization processing can eliminate extreme values ​​or excessively large or small values ​​in the weights, making the weight distribution more uniform and stable, which helps to reduce the volatility and uncertainty of the prediction results and improve the stability of the prediction system.

[0117] In a feasible implementation, the above embodiment determines the standard weight of each preset photovoltaic power prediction method according to the iteration weight, maximum iteration weight and minimum iteration weight of each preset photovoltaic power prediction method in the current iteration, including:

[0118] Using the formula Determining a standard weight for each preset photovoltaic power prediction method;

[0119] In the above embodiment, the combined prediction value is determined according to the standard weight of each preset photovoltaic power prediction method and the historical predicted photovoltaic power at the time point corresponding to the nth preset time period, including:

[0120] Using the formula Determining a combined forecast value;

[0121] in, is the standard weight of the mth preset photovoltaic power prediction method, is the iteration weight of the mth preset photovoltaic power prediction method in the current iteration, is the maximum iteration weight, is the minimum iteration weight, P d is the combined prediction value, M is the number of types of preset photovoltaic power prediction methods, is the historical predicted photovoltaic power at time point t corresponding to the standard weight of the mth preset photovoltaic power prediction method within the nth preset time period.

[0122] In the embodiments of the present application, rigorous calculation formulas for standard weights and combined prediction values ​​are provided from a mathematical perspective. The accuracy of the calculated standard weights and combined prediction values ​​can be ensured from the rigor of mathematical logic, and the above-mentioned calculation formulas are preferably shown to provide reference, understanding and calculation for technical personnel; in addition, by adopting the above-mentioned calculation formulas to calculate standard weights and combined prediction values, it is possible to optimize prediction accuracy, enhance prediction stability, improve the rationality of weight updates, ensure weight rationality, optimize prediction performance and improve stability, etc.

[0123] In a feasible implementation, the above embodiment determines the target predicted photovoltaic power of the photovoltaic field at each time point within the preset time period according to the current predicted photovoltaic power at each time point within the preset time period by each preset photovoltaic power prediction method and the weight of each preset photovoltaic power prediction method, including:

[0124] Using the formula Determine the target predicted photovoltaic power of the photovoltaic field at each time point within a preset time period;

[0125] Among them, P target,m,t is the target predicted photovoltaic power of the photovoltaic field at the tth time point within the preset time period, M is the number of types of preset photovoltaic power prediction methods, and W m ′ is the weight of the mth preset photovoltaic power prediction method, P m,t It is the current predicted photovoltaic power at the tth time point within the preset time period using the mth preset photovoltaic power prediction method.

[0126] In the embodiments of the present application, a rigorous calculation formula for the target predicted photovoltaic power of the photovoltaic field at each time point within a preset time period is provided from a mathematical perspective. The accuracy of the calculated target predicted photovoltaic power of the photovoltaic field at each time point within the preset time period can be ensured from the rigor of mathematical logic, and the above calculation formula is preferably shown to provide reference, understanding and calculation for technical personnel; in addition, by adopting the above calculation formula to calculate the target predicted photovoltaic power of the photovoltaic field at each time point within the preset time period, it is possible to improve prediction accuracy, adapt to complex environments and enhance system stability.

[0127] It can be understood that improving prediction accuracy: by combining the prediction results of multiple preset photovoltaic power prediction methods and performing weighted calculation according to the weight of each photovoltaic power prediction method, a more accurate and reliable target predicted photovoltaic power can be obtained, which helps to reduce the errors that may be caused by a single prediction method and improve the prediction accuracy; adapting to complex environments: different photovoltaic power prediction methods may be suitable for different meteorological factors and environmental factors. By combining multiple photovoltaic power prediction methods, various factors can be comprehensively considered to make the prediction results more adaptable to complex and changeable environmental conditions; enhancing system stability: by comprehensively considering the prediction results of multiple photovoltaic power prediction methods, the uncertainty and volatility of the prediction results can be reduced, and the stability of the photovoltaic field operation can be enhanced, which helps to ensure the stable operation of the power grid and improve the reliability of power supply.

[0128] In a second aspect, the present application also provides a photovoltaic power prediction device.

[0129] See also Figure 2 , is a schematic diagram of a photovoltaic power prediction device in an embodiment of the present application, the device 210 includes:

[0130] The acquisition module 211 is used to acquire the current forecast meteorological data within a preset time period;

[0131] A matching module 212, used to match the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data in a preset weight table according to the current forecast meteorological data;

[0132] A prediction module 213, configured to use each preset photovoltaic power prediction method to predict the current predicted photovoltaic power of each preset photovoltaic power prediction method at each time point within a preset time period according to the current forecast meteorological data;

[0133] The determination module 214 is used to determine the target predicted photovoltaic power of the photovoltaic field at each time point in the preset time period according to the current predicted photovoltaic power of each preset photovoltaic power prediction method at each time point in the preset time period and the weight of each preset photovoltaic power prediction method.

[0134] In the embodiment of the present application, the relevant contents of the acquisition module 211, the matching module 212, the prediction module 213 and the determination module 214 can be referred to in Figure 1 The contents in the illustrated embodiment will not be described in detail here.

[0135] It should be noted that the device 210 of the present application also includes some other modules. It can be understood that the method of the present application and the device 210 have a one-to-one correspondence. Therefore, some other modules of the device 210 of the present application are the contents corresponding to the method of the present application in the above-mentioned embodiment.

[0136] In an embodiment of the present application, various existing photovoltaic power prediction methods with higher accuracy are utilized, and the weights of various photovoltaic power prediction methods corresponding to meteorological factors are matched in a preset weight table according to different meteorological factors, and then different weights are assigned to various photovoltaic power prediction methods, and finally the target predicted photovoltaic power of the photovoltaic field is predicted according to the current predicted photovoltaic power predicted by various photovoltaic power prediction methods and the weights of various photovoltaic power prediction methods. Since this method combines the advantages of the weights of various existing photovoltaic power prediction methods with higher accuracy, it can effectively solve the problem of low prediction accuracy of photovoltaic power generation due to the limitations of generalization of single algorithm scenarios and the one-sidedness of simple combination considerations, that is, this method can effectively improve the prediction accuracy of photovoltaic power generation.

[0137] In addition, the prediction accuracy of this method is highly stable, and there will not be a problem of high prediction accuracy at one moment and low prediction accuracy at the next moment. That is, this method can truly and effectively improve the prediction accuracy of photovoltaic power generation.

[0138] In a third aspect, the present application further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes a photovoltaic power prediction method in the above method embodiment.

[0139] In a fourth aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes a photovoltaic power prediction method in the above method embodiment.

[0140] Figure 3 The internal structure diagram of a computer device in some embodiments is shown. The computer device may be a terminal, a server, or a gateway. Figure 3 As shown, the computer device includes a processor, a memory and a network interface connected through a system bus.

[0141] The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor can implement each step in the above method embodiment. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor can implement each step in the above method embodiment. Those skilled in the art can understand that Figure 3The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0142] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing related hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods.

[0143] Among them, any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0144] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0145] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A photovoltaic power prediction method, It is characterized in that The method comprises: Get the current forecast weather data within the preset time period; According to the current forecast meteorological data, weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data are matched in a preset weight table; Using each preset photovoltaic power prediction method, predicting the current predicted photovoltaic power of each preset photovoltaic power prediction method at each time point within a preset time period according to the current forecast meteorological data; The target predicted photovoltaic power of the photovoltaic field at each time point within the preset time period is determined according to the current predicted photovoltaic power of each preset photovoltaic power prediction method at each time point within the preset time period and the weight of each preset photovoltaic power prediction method.

2. The method according to claim 1, It is characterized in that The matching of the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data in a preset weight table according to the current forecast meteorological data includes: According to the current forecast meteorological data, the historical forecast meteorological data with the highest similarity to the current forecast meteorological data is matched in the preset weight table, and the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data with the highest similarity are used as the weights of all preset photovoltaic power prediction methods corresponding to the current forecast meteorological data.

3. The method according to claim 1 or 2, It is characterized in that The method further comprises: Acquiring historical forecast meteorological data within a plurality of preset time periods, and historical actual photovoltaic power of the photovoltaic field at each time point within the plurality of preset time periods; Using each preset photovoltaic power prediction method, predicting the historical predicted photovoltaic power of each preset photovoltaic power prediction method at each time point in the nth preset time period according to the historical forecast meteorological data in the nth preset time period; Determine the prediction error of each preset photovoltaic power prediction method at each time point in the n preset time period according to the historical predicted photovoltaic power of each preset photovoltaic power prediction method at each time point in the n preset time period and the historical actual photovoltaic power of the photovoltaic electric field at each time point in the n preset time period; Iterate the iteration weight of each preset photovoltaic power prediction method according to the prediction error of each preset photovoltaic power prediction method at each time point in the nth preset time period, and in the iterative process, determine the combined prediction value according to the iteration weights of all preset photovoltaic power prediction methods of the current iteration and the historical predicted photovoltaic power at the time point corresponding to the nth preset time period, and when the combined prediction value meets the preset iteration condition, determine the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in the nth preset time period by the iteration weights of all preset photovoltaic power prediction methods of the current iteration; Wherein, n successively takes integers greater than 0 to obtain weights of all preset photovoltaic power prediction methods corresponding to historical forecast meteorological data within multiple preset time periods; The preset weight table is generated according to the historical forecast meteorological data within a plurality of preset time periods and the weights of all corresponding preset photovoltaic power prediction methods.

4. The method according to claim 3, It is characterized in that The obtaining of historical forecast meteorological data within a plurality of preset time periods and the historical actual photovoltaic power of the photovoltaic field at each time point within the plurality of preset time periods includes: Acquiring initial historical forecast meteorological data and initial historical actual photovoltaic power of the photovoltaic field; Clustering the initial historical meteorological forecast data to obtain historical meteorological forecast data within a plurality of preset time periods; The initial historical actual photovoltaic power of the photovoltaic field is divided according to historical forecast meteorological data in multiple preset time periods to obtain the historical actual photovoltaic power of the photovoltaic field at each time point in the multiple preset time periods.

5. The method according to claim 3, It is characterized in that The iterating of the iteration weight of each preset photovoltaic power prediction method according to the prediction error of each preset photovoltaic power prediction method at each time point in the nth preset time period comprises: Using the formula Iterating the iteration weights of each preset photovoltaic power prediction method; in, is the initial iteration weight, M is the number of preset photovoltaic power prediction methods, to are the iteration weights of the mth preset photovoltaic power prediction method from the 1st iteration to the Dth iteration, α is the smoothing parameter, α∈[0,1], E m,n,t1 To E m,n,tT is the prediction error of the mth preset photovoltaic power prediction method from the 1st time point to the Tth time point within the nth preset time period.

6. The method according to claim 3, It is characterized in that The preset iteration condition includes: the difference between the combined prediction value and the historical actual photovoltaic power at the time point corresponding to the combined prediction value in the nth preset time period is less than or equal to the maximum allowable error value.

7. The method according to claim 3, It is characterized in that The method of determining the combined prediction value according to the iteration weights of all preset photovoltaic power prediction methods of the current iteration and the historically predicted photovoltaic powers at the time point corresponding to the nth preset time period includes: The iteration weights of each preset photovoltaic power prediction method of the current iteration are normalized to obtain the standard weights of each preset photovoltaic power prediction method; Determine a combined prediction value according to the standard weight of each preset photovoltaic power prediction method and the historical predicted photovoltaic power at the time point corresponding to the nth preset time period; The step of determining the weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in the nth preset time period by using the iteration weights of all preset photovoltaic power prediction methods of the current iteration includes: The iteration weights of each preset photovoltaic power prediction method of the current iteration are normalized to obtain the standard weights of each preset photovoltaic power prediction method; The standard weights of all preset photovoltaic power prediction methods are used as weights of all preset photovoltaic power prediction methods corresponding to the historical forecast meteorological data in the nth preset time period.

8. The method according to claim 7, It is characterized in that The iterative weights of each preset photovoltaic power prediction method of the current iteration are normalized to obtain the standard weights of each preset photovoltaic power prediction method, including: Determine a maximum iteration weight and a minimum iteration weight among the iteration weights of all preset photovoltaic power prediction methods of the current iteration; The standard weight of each preset photovoltaic power prediction method is determined according to the iteration weight of each preset photovoltaic power prediction method in the current iteration, the maximum iteration weight and the minimum iteration weight.

9. The method according to claim 8, It is characterized in that The determining of the standard weight of each preset photovoltaic power prediction method according to the iteration weight of each preset photovoltaic power prediction method of the current iteration, the maximum iteration weight and the minimum iteration weight comprises: Using the formula Determining a standard weight for each preset photovoltaic power prediction method; The method of determining the combined prediction value according to the standard weight of each preset photovoltaic power prediction method and the historically predicted photovoltaic power at the time point corresponding to the preset photovoltaic power prediction method in the nth preset time period includes: Using the formula Determining a combined forecast value; in, is the standard weight of the mth preset photovoltaic power prediction method, is the iteration weight of the mth preset photovoltaic power prediction method in the current iteration, is the maximum iteration weight, is the minimum iteration weight, P d is the combined prediction value, M is the number of types of preset photovoltaic power prediction methods, is the historical predicted photovoltaic power at time point t corresponding to the standard weight of the mth preset photovoltaic power prediction method within the nth preset time period.

10. The method according to claim 1, It is characterized in that The method of determining the target predicted photovoltaic power of the photovoltaic field at each time point within the preset time period according to the current predicted photovoltaic power at each time point within the preset time period by each preset photovoltaic power prediction method and the weight of each preset photovoltaic power prediction method comprises: Using the formula Determining a target predicted photovoltaic power of the photovoltaic electric field at each time point within a preset time period; Among them, P target,m,t is the target predicted photovoltaic power of the photovoltaic field at the tth time point within the preset time period, M is the number of types of preset photovoltaic power prediction methods, and W m ′ is the weight of the mth preset photovoltaic power prediction method, P m,t It is the current predicted photovoltaic power at the tth time point within the preset time period using the mth preset photovoltaic power prediction method.