Photovoltaic power probability estimation method and system based on optimized copula

Abstract

The present disclosure relates to the technical field of photovoltaic power generation, and proposes a photovoltaic power probability estimation method and system based on optimized copula. First, the weather types are classified through a clustering method. Then, for each weather type, the best representative weight of the Copula function is constructed to quantitatively represent the power space correlation of centralized photovoltaic and distributed photovoltaic. Finally, the real value of centralized photovoltaic power at the future time is used to represent the prediction value of distributed photovoltaic, and the conditional prediction probability estimation of distributed photovoltaic power is realized. The present disclosure can better reflect the space-time correlation of distributed photovoltaic output by adding k-means and optimized copula function, solve the problem of insufficient single Copula fitting, and achieve good prediction effect.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of photovoltaic power generation, in particular to a photovoltaic power probability estimation method and system based on optimized copula. BACKGROUND

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute the prior art.

[0003] With the acceleration of industrialization and electrification, the demand for energy, especially for electricity, is increasing rapidly. Distributed photovoltaic has the advantages of abundant resources, small development and construction difficulty, significant energy saving and environmental protection benefits, etc., and is one of the important ways of photovoltaic development and utilization. However, due to the instability of weather conditions, photovoltaic power generation has strong intermittency and randomness, which poses challenges to the planning and operation of existing power systems. Therefore, the accuracy of photovoltaic power prediction is an important factor affecting the access of photovoltaic to the power system.

[0004] The inventors found that there are photovoltaic power prediction methods, including physical model-based methods and statistical model-based methods. The physical model-based method needs to be based on the actual photovoltaic power generation system for mathematical modeling, and the accuracy of the prediction is positively correlated with the accuracy of the modeling, and the modeling complexity is difficult to unify and generalize. Statistical models are data-driven models developed using computer performance and artificial intelligence technology. Statistical models currently include neural network models, copula models, etc. Distributed photovoltaic data is difficult to obtain, and is installed in a decentralized manner. Due to the large amount of data required by neural networks and the inability to reflect correlation, it is often difficult to apply. Copula modeling is used to better reflect the spatial correlation of distributed photovoltaic systems, and the amount of data required is relatively small. The use of probability prediction instead of point prediction can reflect more prediction information, making it have certain practical reference value. However, traditional single copula functions have limitations and cannot well fit power data, resulting in reduced accuracy of photovoltaic power prediction. SUMMARY

[0005] To solve the above problems, the present disclosure provides a photovoltaic power probability estimation method and system based on optimized copula, which effectively utilizes meteorological rules based on weather clustering analysis to realize photovoltaic power probability estimation and prediction under each weather type, optimizes the copula function model, improves the accuracy of power prediction, improves the reliability of power system operation, reduces electricity costs, reduces energy consumption, saves energy, reduces emissions, and improves economic efficiency.

[0006] To achieve the above purpose, the present disclosure adopts the following technical solutions:

[0007] One or more embodiments provide a photovoltaic power probability estimation method based on optimized copula, comprising the following steps:

[0008] According to the acquired historical photovoltaic data of centralized and distributed photovoltaic power stations, weather clustering is performed to obtain multiple weather types;

[0009] According to the cumulative distribution of photovoltaic output obtained from the photovoltaic data under different weather types, multiple copula function models quantitatively and dynamically representing the power space correlation of centralized photovoltaic and distributed photovoltaic are constructed, and the optimal model is selected for different weather;

[0010] According to the acquired data of centralized photovoltaic power stations, point prediction of distributed photovoltaic is realized through the optimal model corresponding to the weather;

[0011] A conditional probability model is constructed based on the relationship between the actual value and the point prediction value of the distributed photovoltaic, and the probability distribution of the distributed photovoltaic power and the conditional probability corresponding to the point prediction value are obtained through the conditional probability model.

[0012] One or more embodiments provide a photovoltaic power probability estimation system based on optimized copula, comprising:

[0013] The clustering module is configured to perform weather clustering according to the acquired historical photovoltaic data of centralized and distributed photovoltaic power stations to obtain multiple weather types;

[0014] The model determination module is configured to construct multiple copula function models quantitatively and dynamically representing the power space correlation of centralized photovoltaic and distributed photovoltaic according to the cumulative distribution of photovoltaic output obtained from the photovoltaic data under different weather types, and select the optimal model for different weather;

[0015] The point prediction module is configured to realize point prediction of distributed photovoltaic through the optimal model corresponding to the weather according to the acquired data of centralized photovoltaic power stations;

[0016] The conditional probability prediction module is configured to construct a conditional probability model based on the relationship between the actual value and the point prediction value of the distributed photovoltaic, and obtain the probability distribution of the distributed photovoltaic power and the conditional probability corresponding to the point prediction value through the conditional probability model.

[0017] An electronic device includes a memory and a processor, and computer instructions stored on the memory and running on the processor, when the computer instructions are run by the processor, the steps of the above method are completed.

[0018] Compared with the prior art, the beneficial effects of the present disclosure are:

[0019] The disclosure innovatively combines the optimized copula model under weather clustering and applies it to the field of distributed photovoltaic power prediction, solves the problem of insufficient single Copula fitting, and achieves good prediction effect. First, the weather clustering under three-dimensional scale is carried out for the historical meteorological data, and the copula prediction model is constructed based on the clustering results; meanwhile, the photovoltaic power prediction model and algorithm of the copula considering the historical operation data and weather classification are considered, so that the accuracy of the obtained optimized model is higher, the accuracy of predicting the distributed photovoltaic power through the centralized photovoltaic data is improved, the problem that the distributed photovoltaic data cannot be collected is solved through the centralized photovoltaic prediction of the distributed photovoltaic, and more accurate reference data is provided for power system operation regulation.

[0020] The advantages of the present disclosure and the advantages of additional aspects will be described in detail in the following specific embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0021] The accompanying drawings, which form a part of the present disclosure, are used to provide a further understanding of the present disclosure, and the illustrative embodiments of the present disclosure and their description serve the purpose of explaining the present disclosure. They do not constitute limitations on the present disclosure.

[0022] Figure 1 The flowchart of the distributed photovoltaic power probability estimation method based on the optimized copula of the present disclosure embodiment 1;

[0023] Figure 2 The weather clustering result graph of the K-means clustering of the present disclosure embodiment 1;

[0024] Fig. 3(a) is a frequency histogram of centralized and distributed photovoltaic output under cloudy type in the present disclosure embodiment 1;

[0025] Fig. 3(b) is a frequency histogram of centralized and distributed photovoltaic output under sunny type in the present disclosure embodiment 1;

[0026] Fig. 3(c) is a frequency histogram of centralized and distributed photovoltaic output under overcast type in the present disclosure embodiment 1;

[0027] Fig. 4(a) is an optimized copula density graph under cloudy type in the present disclosure embodiment 1;

[0028] Fig. 4(b) is an optimized copula density graph under sunny type in the present disclosure embodiment 1;

[0029] Fig. 4(c) is an optimized copula density graph under overcast type in the present disclosure embodiment 1;

[0030] Fig. 5(a) is a point prediction and actual value of photovoltaic output under cloudy type in the present disclosure embodiment 1;

[0031] Figure 5(b) is the point prediction, actual value of the photovoltaic output under the sunny type of the embodiment 1 of the present disclosure;

[0032] Figure 5(c) is the point prediction, actual value of the photovoltaic output under the overcast type of the embodiment 1 of the present disclosure;

[0033] Figure 6(a) is the conditional probability distribution of the photovoltaic output under the overcast type when the point prediction is 0.7 of the embodiment 1 of the present disclosure;

[0034] Figure 6(b) is the conditional probability distribution of the photovoltaic output under the sunny type when the point prediction is 0.7 of the embodiment 1 of the present disclosure;

[0035] Figure 6(c) is the conditional probability distribution of the photovoltaic output under the overcast type when the point prediction is 0.7 of the embodiment 1 of the present disclosure;

[0036] Figure 7(a) is the point prediction, actual value, confidence interval of the probability prediction of the photovoltaic output under the overcast type of the embodiment 1 of the present disclosure;

[0037] Figure 7(b) is the point prediction, actual value, confidence interval of the probability prediction of the photovoltaic output under the sunny type of the embodiment 1 of the present disclosure;

[0038] Figure 7(c) is the point prediction, actual value, confidence interval of the probability prediction of the photovoltaic output under the overcast type of the embodiment 1 of the present disclosure. DETAILED DESCRIPTION

[0039] The present disclosure will be further described below in conjunction with the accompanying drawings and embodiments.

[0040] It should be noted that the following detailed description is exemplary in nature and is intended to provide further description of the present disclosure. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.

[0041] It should be noted that the terms used herein are only intended to describe specific embodiments and are not intended to limit exemplary embodiments according to the present disclosure. As used herein, the singular form is intended to include the plural form unless the context clearly indicates otherwise, and it should also be understood that when the terms "comprise" and / or "include" are used in the specification, they indicate the presence of the features, steps, operations, devices, components and / or combinations thereof. It should be noted that the various embodiments and features in the present disclosure can be combined with each other without conflict, and the embodiments will be described in detail below in conjunction with the accompanying drawings.

[0042] Technical term explanation:

[0043] To solve the problems in the background art, the present disclosure proposes a photovoltaic power probability estimation method based on optimized copula. First, the weather types are classified by clustering method. Then, according to the characteristics of each weather attribute, the best representative weight of the Copula function is constructed, and the optimized-copular algorithm is proposed to quantitatively represent the power space correlation of centralized photovoltaic and distributed photovoltaic. Finally, the real value of centralized photovoltaic power at the future time is used to represent the distributed photovoltaic prediction value, and the conditional prediction probability estimation of distributed photovoltaic power is realized. By adding k-means and optimized copula function, the spatio-temporal correlation of distributed photovoltaic output can be better reflected, and the prediction accuracy can be improved to a certain extent. The following will be described with specific examples.

[0044] Embodiment 1

[0045] In the technical solutions disclosed in one or more embodiments, as Figure 1 -As shown in Figure 7, the photovoltaic power probability estimation method based on optimized copula includes the following steps:

[0046] Step 1: According to the obtained historical photovoltaic data of centralized and distributed photovoltaic power stations, weather clustering is performed to obtain multiple weather types;

[0047] Step 2: According to the cumulative distribution of photovoltaic output obtained from photovoltaic data under different weather types, multiple copula function models quantitatively representing the power space correlation of centralized photovoltaic and distributed photovoltaic are constructed respectively;

[0048] Step 3: For different weathers, multiple copula function models are evaluated respectively to obtain the copula function model with the highest photovoltaic power prediction accuracy under different weathers as the optimal model;

[0049] Step 4: According to the data of centralized photovoltaic power station, the point prediction of distributed photovoltaic is realized through the optimal model corresponding to the weather;

[0050] Step 5: Based on the relationship between the actual value and the point prediction value of the distributed photovoltaic, a conditional probability model is constructed, and the probability distribution of the distributed photovoltaic power and the conditional probability corresponding to the point prediction value are obtained through the conditional probability model.

[0051] The Copula function describes the correlation between variables, which is actually a class of functions that connect joint distribution functions with their respective marginal distribution functions, also known as connection functions.

[0052] In this embodiment, the innovative combination of clustering algorithm and optimized copula model under weather classification is applied to the field of distributed photovoltaic power prediction. The weather is clustered in three-dimensional scale based on historical meteorological data, and the copula prediction model is constructed based on the clustering results. The photovoltaic power prediction model and algorithm considering historical operation data and weather classification copula make the obtained optimization model more accurate, improve the accuracy of distributed photovoltaic power prediction through centralized photovoltaic data, solve the problem of distributed photovoltaic data collection, and provide more accurate reference data for power system operation and regulation.

[0053] Further, in step 2, the plurality of copula function models include a Frank Copula function model and a mixed Copula correlation function model, and the mixed Copula correlation function model is a weighted sum of the Frank Copula function model and other models in the Archimedean Copula function cluster model.

[0054] In step 1, specifically, the weather is clustered in three-dimensional scale based on historical meteorological data to prepare for modeling of the copula prediction model. The specific steps are as follows:

[0055] In step 1.1, historical photovoltaic data is obtained and data cleaning is performed.

[0056] Specifically, the historical photovoltaic power data includes historical power data of centralized photovoltaic power stations and distributed photovoltaic power stations. The photovoltaic historical output value is cleaned to exclude abnormal values and zero negative values.

[0057] Optionally, the amount of photovoltaic data that can be collected is the photovoltaic power station power data of 1, 2, and 12 months, and the sampling interval is 10 minutes.

[0058] In step 1.2, historical photovoltaic power data corresponding to the period of meteorological data is obtained, and based on correlation analysis, the clustering elements are determined to cluster the weather, and different weather types are obtained.

[0059] Specifically, in this embodiment, k-means clustering is used, and the process includes the following:

[0060] In step 1.2-1, correlation analysis is used to determine the meteorological elements affecting photovoltaic output as clustering elements.

[0061] The meteorological data of the date corresponding to the historical photovoltaic output value is selected, and the correlation analysis is shown in Table 1, and finally the meteorological elements including atmospheric pressure, relative humidity and radiation are determined as clustering elements. The correlation analysis can be comprehensively measured by statistical correlation coefficients including Pearson, Spearman and Kendall.

[0062] Table 1

[0063]

[0064] Pearson is the Pearson correlation coefficient, Spearman is the Spearman rank correlation coefficient, and Kendall is the Kendall rank correlation coefficient.

[0065] In this embodiment, three factors with strong correlation are selected from five meteorological factors as classification data, i.e. relative humidity, air pressure and shortwave radiation.

[0066] Step 1.2-2, according to the determined clustering elements, the weather is clustered by using the k-means algorithm, and the result is shown in Figure 2 According to the range of each meteorological element corresponding to the clustering result, three weather types are finally determined, i.e. cloudy, sunny and overcast.

[0067] In step 2, for the obtained weather types, the cumulative distribution of photovoltaic output under each weather type is calculated to respectively establish Frank Copula correlation function model and mixed Copula correlation function model.

[0068] Step 2-1, according to the cumulative distribution of photovoltaic output, the correlation coefficient λ value under each weather type is obtained, and the Frank Copula correlation function model is established;

[0069] By observing the frequency distribution, as shown in Figures 3(a)(b)(c). Under different weathers, the frequency distribution will be different, but overall it satisfies the symmetric tail correlation, and Frank Copula can be used to model respectively, and the Frank Copula function model is shown as follows:

[0070]

[0071] In the formula, n and v are two edge distribution variables; λ is the correlation coefficient, and the λ value under each weather type can be obtained according to the cumulative distribution of photovoltaic output, and the copula model under each weather type can be obtained.

[0072] Step 2-2, based on other functions in the Archimedean Copula function cluster except the Frank Copula function, a Copula function cluster model corresponding to each weather is constructed, and the Copula function cluster model and the Frank Copula correlation function model are weighted and summed to obtain an optimized mixed Copula correlation function model.

[0073] Based on the method of step 2-1, it is found that only the Frank Copula can be basically used. In this embodiment, based on the Archimedean Copula function cluster, the parameters are solved and optimized by an optimization algorithm, which can improve the prediction accuracy.

[0074] Specifically, in addition to the Frank Copula function, this embodiment also uses two other functions in the Archimedean Copula, including: Gumble Copula function and Clayton Copula function, as shown in formulas (2) and (3) respectively,

[0075]

[0076]

[0077] The optimized mixed Copula correlation function model obtained after weighting is shown in formula (4):

[0078] C H (u,v; λ1, λ2, λ3) = A*C F (u,v; λ1) + B*C C (u,v; λ2) + C*C G (u,v; λ3) (4)

[0079] In the formula, A, B, and C are weight coefficients of Frank Copula, Clayton Copula, and Gumble Copula respectively; λ1, λ2, and λ3 are corresponding correlation coefficients.

[0080] The solving method of the weight coefficients and the correlation coefficients λ1, λ2, and λ3 is: based on the above-mentioned formulas (1), (2), and (3) into formula (4), the copula value to be solved is obtained, and the error between the solved copula value and the empirical copula value is taken as the objective function. Genetic algorithm is used to solve the parameters, and all the parameter values to be solved are obtained, that is, the weight coefficients and the correlation coefficients λ1, λ2, and λ3.

[0081] In this embodiment, the optimized copula function based on the Archimedean Copula function cluster as a model improves the prediction accuracy to some extent and can better fit the power data; the parameter values and weight values in the function cluster are determined according to the optimization algorithm, which overcomes the limitations of the traditional single copula function, can be suitable for the fitting of photovoltaic power data, and is more flexible in optimization to establish a model.

[0082] In step 3, the optimal copula model corresponding to each weather type can be selected from the Frank Copula model and the optimized mixed Copula correlation function model by comparing the correlation coefficients and error evaluation indexes under different weathers.

[0083] In this embodiment, the selected correlation coefficients include the Pearson correlation coefficient, the determination coefficient R 2 , and the error evaluation index is the root mean square error RMSE.

[0084] Specifically, the correlation coefficients and error indexes are shown in Table 2, and the optimal copula model under each weather type is selected. The selected error index is RMSE, the correlation coefficient is R 2 , and the Pearson. According to the error relationship and the degree of similarity between the true value and the model prediction value, the model with the best index under each weather type is selected as the optimal copula model.

[0085] Table 2

[0086]

[0087] In Table 2, the single model refers to the Frank Copula model, and the mixed model refers to the mixed model of formula 4.

[0088] Pearson (Pearson Correlation Coefficient) correlation coefficient, which represents the trend and degree of change between two variables, the value range is between-1 and +1, 0 represents no correlation, positive value represents positive correlation, negative value represents negative correlation, and the greater the value, the stronger the correlation.

[0089] R 2 is the determination coefficient, also known as the goodness of fit, which is the square of the correlation coefficient r. It represents the part of the dependent variable variation that can be explained according to the independent variable variation. The size of the determination coefficient determines the closeness of the correlation. Meaning: the larger the goodness of fit, the higher the degree of explanation of the dependent variable by the independent variable, and the higher the percentage of the total variation caused by the independent variable. The observation points are more concentrated near the regression straight line. RMSE is the root mean square error.

[0090] The error indicators of the mixed Copula model for cloudy and overcast days are better than those of the Frank Copula model, and the optimized mixed Copula model is preferred under this weather condition; for sunny days, the R 2 has improved, but the RMSE is weaker than that of the Frank Copula model, the Pearson of the two models is equal, and the effects of the two models are similar under this condition, and the mixed Copula model can also be selected for sunny days. In this embodiment, clustering is performed on three weather types, and other weather types can use the same method to select the optimal model. The Copula density function graphs of cloudy, overcast and sunny days are shown in FIGS. 4(a), (b) and (c).

[0091] In step 4, the centralized photovoltaic power prediction results are used as inputs under different weather conditions, and the point prediction results of the distributed photovoltaic power are obtained from the corresponding model. Part of the point prediction results under each weather condition are selected, as shown in FIGS. 5(a), (b) and (c), and the effect graph of the point prediction under different weather types is obtained through the mixed model established in this embodiment, as shown in FIG. 5. It can be seen from FIG. 5 that the photovoltaic power predicted by the method of this embodiment can achieve the predetermined effect.

[0092] In step six, finally, a conditional probability model is constructed based on the actual value and the point prediction value of the distributed photovoltaic, and the actual probability distribution and the conditional probability corresponding to the point prediction value are obtained through the conditional probability model. The conditional probability model is shown in formulas (5) and (6).

[0093] Let the actual value be x and the point prediction value be y, then the joint probability density function of x and y is shown in formula (5).

[0094]

[0095] In the formula, f X (x) and f Y (y) are the probability density functions of the marginal distribution of x and y, respectively; c(F X (x), F Y (y)) is the density function of the Copula.

[0096] Among them, the Copula probability density function is obtained according to the established model (formula 4). The probability density of the marginal distribution is obtained from the data of centralized and distributed photovoltaic output.

[0097] Given the point prediction y = p, the conditional probability density function of the actual value is shown in formula (6).

[0098]

[0099] The formula (6) shows that the conditional probability density includes two parts of the copula density function with a variable multiplier and the probability density function of the actual value. The conditional probability distribution of the photovoltaic output under different weather types when the point prediction is 0.7 is shown in FIG. 6 (a) (b) (c). As can be seen from FIG. 6, the conditional probability of different weather types is different under the same point prediction, and thus the overall probability prediction diagram is established, as shown in FIG. 7.

[0100] To illustrate the effect of the optimization model of the embodiment, a simulation experiment is performed to evaluate and test the predicted conditional probability. The prediction result is shown in FIG. 7, and the prediction evaluation result is shown in Table 3.

[0101] Table 3: Average width of prediction interval

[0102]

[0103] FIG. 7 is a probability prediction effect diagram under different weather types. As can be seen from FIG. 7, the conditional probability prediction on all point predictions can be realized, and the prediction information on different confidence intervals can be obtained. Table 3 is the probability prediction result of the optimized model which is better than the single model only using the interval average width.

[0104] Embodiment 2

[0105] Based on the embodiment 1, the embodiment 2 provides a photovoltaic power probability estimation system based on an optimized copula, which comprises:

[0106] A clustering module configured to perform weather clustering to obtain a plurality of weather types according to the acquired historical photovoltaic data of the centralized and distributed photovoltaic power stations.

[0107] A model determination module configured to respectively construct a plurality of copula function models quantitatively representing the power space correlation of the centralized photovoltaic and the distributed photovoltaic according to the cumulative distribution of the photovoltaic output obtained from the photovoltaic data under different weather types, and select the optimal model for different weather.

[0108] A point prediction module configured to realize the point prediction of the distributed photovoltaic through the optimal model of the corresponding weather according to the acquired data of the centralized photovoltaic power station.

[0109] A conditional probability prediction module configured to construct a conditional probability model based on the relationship between the actual value and the point prediction value of the distributed photovoltaic, and obtain the probability distribution of the distributed photovoltaic power and the conditional probability corresponding to the point prediction value through the conditional probability model.

[0110] It should be noted that each module in the embodiment corresponds to each step in the embodiment 1, and the specific implementation process is the same, which will not be repeated here.

[0111] Example 3

[0112] The embodiment provides an electronic device, comprising a memory and a processor and computer instructions stored on the memory and running on the processor, when the computer instructions are run by the processor, the steps of the method in the embodiment 1 are completed.

[0113] The above merely provides the preferred embodiments of the present disclosure but is not intended to limit the present disclosure. For those skilled in the art, the present disclosure can have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

[0114] The above describes the specific embodiments of the present disclosure in combination with the drawings, but is not intended to limit the protection scope of the present disclosure, and those skilled in the art should understand that various modifications or changes made on the basis of the technical solutions of the present disclosure are still within the protection scope of the present disclosure without creative labor.

Claims

1. A photovoltaic power probability estimation method based on optimized copula, characterized in that... Includes the following steps: Based on the historical photovoltaic data of centralized and distributed photovoltaic power stations, weather clustering is performed to obtain multiple weather types; Based on the cumulative distribution of photovoltaic power output obtained from photovoltaic data under different weather types, multiple copula function models are constructed to quantitatively and dynamically represent the power spatial correlation of centralized photovoltaic and distributed photovoltaic. Among them, the multiple copula function models include the Frank Copula correlation function model and the hybrid Copula correlation function model. The hybrid Copula correlation function model is a weighted sum of the Frank Copula correlation function model and other models in the Archimedesian Copula function family model. Based on the cumulative distribution of photovoltaic output, the correlation coefficients for each weather type were obtained, and a Frank Copula correlation function model was established. The correlation coefficients included the Pearson correlation coefficient and the coefficient of determination R0. 2 The error evaluation index is the root mean square error. Based on the Gumble Copula and Clayton Copula function models in the Archimedesian Copula function family, a Copula function family model corresponding to each weather condition is constructed. The optimized hybrid Copula correlation function model is obtained by weighted summation of the Copula function family model and the Frank Copula correlation function model, as shown in the following formula: In the formula, A, B, and C are the weight coefficients of the Frank Copula function model, the Clayton Copula function model, and the Gumble Copula function model, respectively. The corresponding correlation coefficients; weighting coefficients and correlation coefficients. The solution method is as follows: Based on the Frank Copula correlation function model, Gumble Copula function model, and Clayton Copula function model, the copula value to be solved is obtained by substituting it into the formula, and the empirical copula value is used as the setpoint. The error between the solved copula value and the empirical copula value is used as the objective function. A genetic algorithm is used to solve for the parameters, and the weight coefficients and correlation coefficients are obtained. ; Based on the correlation coefficient and error evaluation index, multiple copula function models were evaluated for different weather conditions, and the copula function model with the highest photovoltaic power prediction accuracy under different weather conditions was selected as the optimal model. Based on the data obtained from centralized photovoltaic power plants, point prediction for distributed photovoltaic power is achieved through the optimal model corresponding to the weather. A conditional probability model is constructed based on the relationship between the actual value and the point prediction value of distributed photovoltaic power. The probability distribution of distributed photovoltaic power and the conditional probability corresponding to the point prediction value are obtained through the conditional probability model. The conditional probability density includes the Copula density function of the variable multiplier and the probability density function of the actual value.

2. The photovoltaic power probability estimation method based on optimized copula as described in claim 1, characterized in that... A method for obtaining multiple weather types by performing weather clustering based on historical photovoltaic data from centralized and distributed photovoltaic power plants includes the following steps: Acquire historical photovoltaic data and perform data cleaning; Meteorological data corresponding to historical photovoltaic power data are obtained, and clustering elements are determined based on correlation analysis to cluster the weather and obtain different weather types.

3. The photovoltaic power probability estimation method based on optimized copula as described in claim 2, characterized in that: The clustering elements are three-dimensional features of atmospheric pressure, relative humidity, and radiance.

4. The photovoltaic power probability estimation method based on optimized copula as described in claim 2, characterized in that: The clustering method used is the k-means clustering algorithm.

5. The photovoltaic power probability estimation method based on optimized copula as described in claim 1, characterized in that: A conditional probability model is constructed based on the actual values ​​and point prediction values ​​of distributed photovoltaic power generation to calculate the actual probability distribution and the conditional probability corresponding to the point prediction values.

6. A photovoltaic power probability estimation system based on optimized copula, used to implement the photovoltaic power probability estimation method based on optimized copula as described in any one of claims 1 to 5, characterized in that... include: Clustering module: Configured to perform weather clustering based on historical photovoltaic data from centralized and distributed photovoltaic power plants to obtain multiple weather types; Model determination module: It is configured to construct multiple copula function models that quantify the spatial correlation of power of centralized photovoltaic and distributed photovoltaic based on the cumulative distribution of photovoltaic power output obtained from photovoltaic data under different weather types, and select the optimal model for different weather conditions; Point prediction module: It is configured to make point predictions for distributed photovoltaics based on the acquired data from centralized photovoltaic power plants and the optimal model corresponding to the weather. Conditional probability prediction module: It is configured to build a conditional probability model based on the relationship between the actual value and the point prediction value of distributed photovoltaic power, and obtain the probability distribution of distributed photovoltaic power and the conditional probability corresponding to the point prediction value through the conditional probability model.

7. An electronic device, characterized in that... It includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, which, when executed by the processor, perform the steps of the method according to any one of claims 1 to 5.

Images