Photovoltaic power generation power prediction method and system, storage medium and electronic equipment

Abstract

The invention provides a photovoltaic power generation power prediction method and system, a storage medium and electronic equipment, and the method comprises the steps: inputting weather forecast data into a pre-training model, and obtaining corrected power generation power prediction data; the model pre-training comprises the steps of obtaining historical multi-mode weather forecast data and historical photovoltaic station observation data of a photovoltaic power station to be predicted, and constructing an earth surface solar radiation correction model, a photovoltaic power generation power prediction model and a photovoltaic power generation power correction model based on the historical multi-mode weather forecast data and the historical photovoltaic station observation data. According to the prediction method provided by the invention, the defect that the accuracy rate of surface solar radiation predicted by a single numerical prediction mode is low is overcome, and the prediction uncertainty is reduced through the method for predicting the short-term photovoltaic power generation power through the meteorological parameters output by multiple numerical prediction modes; and meanwhile, the photovoltaic power generation power forecasting accuracy can be further improved through the secondary correction method.

Description

Technical Field

[0001] The present disclosure relates to the field of new energy technologies, and in particular, to a photovoltaic power prediction method and system, a storage medium, and an electronic device. Background Art

[0002] As one of the rich renewable energy sources, photovoltaic power generation has the advantages of being clean, pollution-free, and inexhaustible. It can effectively alleviate the combustion of fossil fuels such as oil and coal and slow down global warming. However, photovoltaic power generation has strong randomness, intermittency, and volatility, and its power generation has great uncertainty, thus posing severe challenges to the planning, operation, and control of the power system. The short-term prediction of photovoltaic power not only relates to the safe operation of the power grid but also plays an important role in the power generation planning of the power department. Accurate short-term power prediction is one of the key technologies to solve grid connection obstacles.

[0003] The existing photovoltaic power prediction methods mainly include physical modeling methods and statistical methods. The physical modeling method obtains historical irradiance based on historical meteorological data and then establishes a mathematical model that can convert light energy into electrical energy to indirectly predict the photovoltaic power. Based on the surface solar radiation output by a single numerical prediction model and substituting it into an empirical formula or using statistical methods to obtain the photovoltaic power, the accuracy is low and the uncertainty is large. Summary of the Invention

[0004] The purpose of the present disclosure is to provide a photovoltaic power prediction method and system, a storage medium, and an electronic device, which can solve at least one of the above-mentioned technical problems. The specific solutions are as follows:

[0005] According to specific embodiments of the present disclosure, on the one hand, the present disclosure provides a photovoltaic power generation prediction method, including: obtaining multi-mode meteorological forecast data of a photovoltaic power station to be predicted; inputting the multi-mode meteorological forecast data into a pre-trained model to obtain corrected power generation prediction data; wherein, the pre-trained model includes: obtaining historical multi-mode meteorological forecast data and historical photovoltaic power station observation data of the photovoltaic power station to be predicted, wherein the historical photovoltaic power station observation data includes historical surface solar radiation data and historical photovoltaic power generation data; based on the historical multi-mode meteorological forecast data and the historical surface solar radiation data, constructing a surface solar radiation correction model; inputting the historical multi-mode meteorological forecast data into the surface solar radiation correction model to obtain corrected surface solar radiation data; based on the historical multi-mode meteorological forecast data, the historical photovoltaic power generation data and the corrected surface solar radiation data, constructing a photovoltaic power generation prediction model; inputting the historical multi-mode meteorological forecast data into the photovoltaic power generation prediction model to obtain predicted photovoltaic power generation data; and constructing a photovoltaic power generation correction model based on the historical multi-mode meteorological forecast data and the predicted photovoltaic power generation data.

[0006] In an optional embodiment, the historical photovoltaic power generation data includes: the longitude and latitude information, surface solar radiation data and corresponding photovoltaic power generation data of the photovoltaic power station at preset time intervals; the preset time interval is 15 minutes.

[0007] In an optional embodiment, the obtaining of the historical multi-mode meteorological forecast data and the historical photovoltaic power station observation data of the photovoltaic power station to be predicted includes: cleaning the observation data, and removing data in the surface solar radiation data that is less than 0 W / m 2 and greater than 2000 W / m 2 ; and removing data with photovoltaic power generation less than 0 MW and greater than 50 MW.

[0008] In an optional embodiment, the obtaining of the historical multi-mode meteorological forecast data and the historical photovoltaic power station observation data of the photovoltaic power station to be predicted further includes: when there are continuously unchanged values for more than 3 hours, performing data quality control according to the actual situation.

[0009] In an optional embodiment, the pre-trained model further includes: constructing the surface solar radiation correction model by using the XGBoost algorithm.

[0010] In an optional embodiment, the pre-trained model further includes: constructing the photovoltaic power generation prediction model by using the XGBoost algorithm.

[0011] In an optional embodiment, the pre-trained model further includes: constructing the photovoltaic power generation correction model by using the XGBoost algorithm.

[0012] In an alternative embodiment, the obtaining of the historical multi-mode meteorological forecast data and the historical photovoltaic power station observation data of the photovoltaic power station to be predicted includes: obtaining the historical data of the European Centre for Medium-Range Weather Forecasts numerical weather prediction model; obtaining the historical data of the Global Forecast System model of the National Centers for Environmental Prediction of the United States; obtaining the data of the Wind Energy and Solar Energy Meteorological Forecast System of the China Meteorological Administration; wherein, the position of the photovoltaic power station to be predicted is determined in the historical data of the European Centre for Medium-Range Weather Forecasts numerical weather prediction model, the historical data of the Global Forecast System model of the National Centers for Environmental Prediction of the United States, and the data of the Wind Energy and Solar Energy Meteorological Forecast System of the China Meteorological Administration respectively by using the spatial interpolation method.

[0013] In an alternative embodiment, the data of a preset time period are obtained in the historical data of the European Centre for Medium-Range Weather Forecasts numerical weather prediction model, the historical data of the Global Forecast System model of the National Centers for Environmental Prediction of the United States, and the data of the Wind Energy and Solar Energy Meteorological Forecast System of the China Meteorological Administration respectively by using the time interpolation method.

[0014] In an alternative embodiment, the spatial interpolation method is a four-point interpolation method, which is a method for extracting the meteorological data of the four nearest grid points corresponding to the longitude and latitude of the photovoltaic power station.

[0015] In an alternative embodiment, for variables with obvious time variation characteristics such as surface solar radiation and air temperature in the time interpolation method, the cubic spline interpolation method is adopted, and for variables such as wind speed and wind direction, the linear interpolation method is adopted.

[0016] In an alternative embodiment, the pre-trained model further includes: dividing the historical multi-mode meteorological forecast data and the historical photovoltaic power generation power forecast data into a training set and a validation set according to a ratio of 8:2, establishing a pre-trained model based on the XGBoost algorithm, and optimizing the hyperparameters of the constructed pre-trained model by using the grid search method to optimize the photovoltaic power correction model.

[0017] According to a specific embodiment of the present disclosure, on the other hand, the present disclosure provides a photovoltaic power generation power prediction system configured to perform photovoltaic power generation power prediction by executing the method described in any one of the above technical solutions.

[0018] According to a specific embodiment of the present disclosure, on the other hand, the present disclosure provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the method described in any one of the above technical solutions is implemented.

[0019] According to specific embodiments of the present disclosure, on the other hand, the present disclosure provides an electronic device, including: one or more processors; a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the method described in any one of the above technical solutions.

[0020] The above solutions of the embodiments of the present disclosure have at least the following beneficial effects compared with the prior art:

[0021] The photovoltaic power prediction method provided by the present disclosure makes up for the deficiency of the low accuracy of predicting surface solar radiation by a single numerical prediction model. The method of predicting short-term photovoltaic power by meteorological parameters output by multiple numerical prediction models reduces the uncertainty of the prediction. At the same time, the secondary correction method can further improve the accuracy of the photovoltaic power prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The flowchart of the photovoltaic power prediction method according to an embodiment of the present disclosure is shown.

[0023] Figure 2 The flowchart of the pre-trained model according to an embodiment of the present disclosure is shown.

[0024] Figure 3 The schematic diagram of extracting data of four grid points around a photovoltaic power station in a numerical model by using the four-point interpolation method according to an embodiment of the present disclosure is shown.

[0025] Figure 4 The schematic diagram of the connection structure of the electronic device according to an embodiment of the present disclosure is shown.

[0026] Reference Signs:

[0027] 301: Processing system; 302: ROM; 303: RAM; 304: Bus; 305: I / O interface; 306: Input system; 307: Output system; 308: Storage system; 309: Communication system. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] In order to make the objectives, technical solutions, and advantages of the present disclosure clearer, the present disclosure will be further described in detail below with reference to the accompanying drawings. Apparently, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present disclosure without creative efforts shall fall within the protection scope of the present disclosure.

[0029] The terms used in the embodiments of the present disclosure are for the purpose of describing specific embodiments only and are not intended to limit the present disclosure. The singular forms "a", "the" and "said" used in the embodiments of the present disclosure and the appended claims are also intended to include the plural forms unless the context clearly dictates otherwise. "Plurality" generally includes at least two.

[0030] It should be understood that the term "and / or" used herein is merely a description of the relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after.

[0031] It should be understood that although terms such as first, second, and third may be used to describe structures in the embodiments of the present disclosure, these structures should not be limited to these terms. These terms are only used to distinguish different structures. For example, without departing from the scope of the embodiments of the present disclosure, the first component may also be referred to as the second component, and similarly, the second component may also be referred to as the first component.

[0032] Depending on the context, the words "if", "when" as used herein may be interpreted as "when" or "while" or "in response to determining" or "in response to detecting". Similarly, depending on the context, the phrase "if determined" or "if detected (stated condition or event)" may be interpreted as "when determined" or "in response to determining" or "when detected (stated condition or event)" or "in response to detecting (stated condition or event)".

[0033] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a commodity or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such commodity or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the presence of additional identical elements in the commodity or device comprising said element.

[0034] The main methods for predicting the power generation of photovoltaic power in related technologies are physical modeling methods and statistical methods. The physical modeling method obtains historical irradiance based on historical meteorological data, and then establishes a mathematical model that can convert light energy into electrical energy, thereby indirectly predicting the photovoltaic power generation. The statistical method is to obtain the relationship between photovoltaic power generation and meteorological data based on historical data, and then establish a photovoltaic power generation prediction model. Considering different prediction methods, the accuracy of short-term photovoltaic power generation prediction can be improved from the following three aspects: (1) Numerical weather prediction model (NWP) correction; (2) Model optimization; (3) Prediction result correction. Based on two numerical weather prediction models, the predicted surface solar radiation variables are corrected through machine learning algorithms, and on this basis, a photovoltaic power generation prediction model is established, which is an effective way to improve the accuracy of photovoltaic power generation prediction. Therefore, a short-term photovoltaic power generation prediction method based on multiple numerical weather prediction models and machine learning algorithms is needed. This also lays a foundation for promoting the high-quality development of solar energy meteorological services and comprehensively promoting the in-depth development of photovoltaic power meteorological services with refined spatio-temporal resolution. Disadvantages in related technologies: Using the surface solar radiation output by a single numerical weather prediction model, substituting it into an empirical formula or using statistical methods to obtain the photovoltaic power generation, the accuracy is low and the uncertainty is large.

[0035] In order to solve at least one of the above-mentioned technical problems, the present disclosure provides a photovoltaic power generation prediction method, system, storage medium, and electronic device. The photovoltaic power generation prediction method may include: obtaining multi-mode meteorological forecast data of a photovoltaic power station to be predicted; inputting the multi-mode meteorological forecast data into a pre-trained model to obtain corrected power generation prediction data; wherein, the pre-trained model includes: obtaining historical multi-mode meteorological forecast data and historical photovoltaic power station observation data of the photovoltaic power station to be predicted, wherein the historical photovoltaic power station observation data includes historical surface solar radiation data and historical photovoltaic power generation data; based on the historical multi-mode meteorological forecast data and the historical surface solar radiation data, constructing a surface solar radiation correction model; inputting the historical multi-mode meteorological forecast data into the surface solar radiation correction model to obtain corrected surface solar radiation data; based on the historical multi-mode meteorological forecast data, the historical photovoltaic power generation data, and the corrected surface solar radiation data, constructing a photovoltaic power generation prediction model; inputting the historical multi-mode meteorological forecast data into the photovoltaic power generation prediction model to obtain predicted photovoltaic power generation data; constructing a photovoltaic power generation correction model based on the historical multi-mode meteorological forecast data and the predicted photovoltaic power generation data. The photovoltaic power generation prediction method provided by the present disclosure makes up for the deficiency of the low accuracy of predicting surface solar radiation by a single numerical weather prediction model. The method of predicting short-term photovoltaic power generation through meteorological parameters output by multiple numerical weather prediction models reduces the uncertainty of the prediction. At the same time, the secondary correction method can further improve the accuracy of the photovoltaic power generation prediction.

[0036] The optional embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0037] Figure 1 The flowchart of the photovoltaic power prediction method according to an embodiment of the present disclosure is shown. As Figure 1 shown, according to the specific implementation manner of the present disclosure, on the one hand, a photovoltaic power prediction method is provided, and the photovoltaic power prediction method may at least include the following steps:

[0038] S100. Obtain multi-mode meteorological forecast data of the photovoltaic power station to be predicted.

[0039] S200. Input the multi-mode meteorological forecast data into the pre-trained model to obtain corrected power generation prediction data.

[0040] Figure 2 The flowchart of the pre-trained model according to an embodiment of the present disclosure is shown. As Figure 2 shown, specifically, the pre-trained model may include:

[0041] S300. Obtain historical multi-mode meteorological forecast data and historical photovoltaic power station observation data of the photovoltaic power station to be predicted, where the historical photovoltaic power station observation data includes historical surface solar radiation data and historical photovoltaic power generation data.

[0042] S400. Based on the historical multi-mode meteorological forecast data and the historical surface solar radiation data, construct a surface solar radiation correction model.

[0043] S500. Input the historical multi-mode meteorological forecast data into the surface solar radiation correction model to obtain corrected surface solar radiation data.

[0044] S600. Based on the historical multi-mode meteorological forecast data, the historical photovoltaic power generation data, and the corrected surface solar radiation data, construct a photovoltaic power generation prediction model.

[0045] S700. Input the historical multi-mode meteorological forecast data into the photovoltaic power generation prediction model to obtain predicted photovoltaic power generation data.

[0046] S800. Based on the historical multi-mode meteorological forecast data and the predicted photovoltaic power generation data, construct a photovoltaic power generation correction model.

[0047] The photovoltaic power prediction method provided by the present disclosure makes up for the deficiency of the low accuracy rate of predicting surface solar radiation by a single numerical prediction model. The method of predicting short-term photovoltaic power by meteorological parameters output by multiple numerical prediction models reduces the uncertainty of the prediction. At the same time, the secondary correction method can further improve the accuracy rate of photovoltaic power prediction.

[0048] Among them, in step S100 and step S200, the meteorological data of the photovoltaic power station to be predicted is input into the pre-trained model. After the calculation and correction of the pre-trained model, the power prediction data with a certain accuracy rate can be predicted.

[0049] Among them, in step S300 and step S400, historical observation data of photovoltaic power stations are collected. The variables included in the data are: site longitude and latitude information, surface solar radiation, and photovoltaic power generation. The time resolution is 15 minutes. The methods for quality control of the observation data may include: removing data in the surface solar radiation data that is less than 0 W / m 2 and greater than 2000 W / m 2 ; removing data with photovoltaic power less than 0 MW and greater than 50 MW; removing data with relatively large (small) surface solar radiation values and relatively small (large) photovoltaic power values. In addition, when there are continuous unchanged values for more than 3 hours, data quality control is carried out according to the actual situation. Historical multi-model meteorological data is obtained. The multi-models may include: the European Centre for Medium-Range Weather Forecasts (ECMWF), the Global Forecast System model of the National Centers for Environmental Prediction of the United States (GFS), and the data of the China Meteorological Administration's Wind Energy and Solar Energy Meteorological Forecast System (CMA-WSP); among them, the time resolution of the European Centre for Medium-Range Weather Forecasts model is 3 hours, and the spatial resolution is 0.1°; the time resolution of the Global Forecast System model of the National Centers for Environmental Prediction of the United States is 1 hour, and the spatial resolution is 0.25°; the data of the China Meteorological Administration's Wind Energy and Solar Energy Meteorological Forecast System has a time resolution of 1 hour and a spatial resolution of 9 km. Since the resolutions of the two models are relatively coarse, in order to avoid large deviations when extracting model data, a four-point interpolation method is used to extract the four nearest grid point data around the longitude and latitude of the photovoltaic power station, as Figure 3As shown, the meteorological parameters of the four points are all input as eigenvalues. To align with the observed data, cubic spline interpolation and linear interpolation methods are used to interpolate the 3-hourly and 1-hourly data into 15-minute intervals. The variables extracted by ECMWF can include: global horizontal irradiance at the surface received (GHI_EC), direct normal irradiance at the surface received (DNI_EC), low cloud cover (LCC_EC), middle cloud cover (MCC_EC), high cloud cover (HCC_EC), 2m temperature (T2_EC), 2m dew point (D2_EC), 10m U-component wind speed (U10_EC), 10m V-component wind speed (V10_EC), 100m U-component wind speed (U100_EC), 100m V-component wind speed (V100_EC), visibility (VIS_EC), surface pressure (P_EC). The variables extracted by GFS can include: global horizontal irradiance at the surface received (GHI_GFS), 2m temperature (T2_GFS), 2m specific humidity (Q2_GFS), 2m relative humidity (RH2_GFS), 10m U-component wind speed (U10_GFS), 10m V-component wind speed (V10_GFS), 100m U-component wind speed (U100_GFS), 100m V-component wind speed (V100_GFS), visibility (VIS_GFS), total cloud cover (TCC_GFS), low cloud cover (LCC_GFS), middle cloud cover (MCC_GFS), high cloud cover (HCC_GFS). The variables extracted by CMA-WSP can include: global horizontal irradiance at the surface received (GHI_CMA), direct normal irradiance at the surface received (DNI_CMA), diffuse horizontal irradiance at the surface received (DHI_CMA), 2m temperature (T2_CMA), 2m specific humidity (Q2_CMA), 2m relative humidity (RH2_CMA), 10m U-component wind speed (U10_CMA), 10m V-component wind speed (V10_CMA), 100m U-component wind speed (U100_CMA), 100m V-component wind speed (V100_CMA), visibility (VIS_CMA), total cloud cover (TCC_CMA), low cloud cover (LCC_CMA), middle cloud cover (MCC_CMA), high cloud cover (HCC_CMA). Among them, in step S500 and step S600, the observed data and the meteorological data of multi-model forecasts are combined in time series, and the corresponding observed or model data of the missing data are excluded after combination. The target value is the observed surface solar radiation value, and the eigenvalue is the meteorological parameter output by the multi-model. Eigenvalue selection is performed according to the filtering method, that is, each eigenvalue is scored using correlation, and the top ten eigenvalues are as follows: GHI_EC, GHI_CMA, DNI_EC, DNI_CMA, DHI_CMA, TCC_CMA, LCC_EC, LCC_CMA, MCC_EC, MCC_CMA.Randomly select 90% of the data as the training set and 20% of the data as the validation set. Use the XGBoost machine learning algorithm to train the historical data, and use the grid search method to adjust the important hyperparameters in XGBoost. Taking the model score and error as the criteria, determine the optimal model when the score is the highest and the error is relatively low, and obtain the surface solar radiation correction model.

[0050] Among them, in step S700 and step S800, obtain the surface solar radiation data after historical correction according to the surface solar radiation correction model, and perform eigenvalue selection and dataset division. The target value is the observed photovoltaic power generation. Based on the eigenvalues in step 102, add the corrected surface solar radiation (GHI_hist_model) as an eigenvalue. Among them, the corrected surface solar radiation ranks first, and the top ten eigenvalues are as follows: GHI_hist_model, GHI_EC, GHI_CMA, DNI_EC, DNI_CMA, DHI_CMA, TCC_CMA, LCC_EC, LCC_CMA, MCC_EC. Randomly select 80% of the data as the training set and 20% of the data as the validation set. Use XGBoost machine learning to train the historical data, adjust the hyperparameters of XGBoost, and take the model score and error as the criteria. Determine the optimal model when the score is the highest and the error is relatively low, and obtain the photovoltaic power generation prediction model. Substitute the historical multi-mode meteorological forecast data into the photovoltaic power generation prediction model to obtain the historical photovoltaic power generation data. Taking the observed photovoltaic power generation as the target value, based on the eigenvalues in step 201, add the photovoltaic power generation data (PV_his_model) calculated by the machine learning model as an eigenvalue. Among them, the calculated photovoltaic power generation ranks first, and the top ten eigenvalues are as follows: PV_his_model, GHI_hist_model, GHI_EC, GHI_CMA, DNI_EC, DNI_CMA, DHI_CMA, TCC_CMA, LCC_EC, LCC_CMA. Randomly select 80% of the data as the training set and 20% of the data as the validation set. Use XGBoost machine learning to train the historical data, adjust the important parameters in XGBoost, and take the model score and error as the criteria. Determine the optimal model when the score is the highest and the error is relatively low, and obtain the photovoltaic power generation correction model.

[0051] In some embodiments, the historical photovoltaic power generation data includes: the longitude and latitude information of the photovoltaic power station, the surface solar radiation data, and the corresponding photovoltaic power generation data at preset time intervals; the preset time interval is 15 minutes.

[0052] In some embodiments, the obtaining of the historical multi - mode meteorological forecast data and the historical photovoltaic power station observation data of the photovoltaic power station to be predicted includes: cleaning the observation data, and removing the data in the surface solar radiation data that is less than 0 W / m 2 and greater than 2000 W / m 2 ; removing the data of photovoltaic power generation with a power less than 0 MW and greater than 50 MW; when there are continuously unchanged values for more than 3 hours, data quality control is performed according to the actual situation.

[0053] In some embodiments, the pre - trained model further includes: constructing the surface solar radiation correction model using the XGBoost algorithm; constructing the photovoltaic power generation prediction model using the XGBoost algorithm.

[0054] In some embodiments, the obtaining of the historical multi - mode meteorological forecast data and the historical photovoltaic power station observation data of the photovoltaic power station to be predicted includes: obtaining the historical data of the European Centre for Medium - Range Weather Forecasts numerical weather prediction model; obtaining the historical data of the Global Forecast System model of the National Centers for Environmental Prediction of the United States; obtaining the data of the China Meteorological Administration's Wind Energy and Solar Energy Meteorological Forecast System; wherein, the spatial interpolation method is used to determine the location of the photovoltaic power station to be predicted in the historical data of the European Centre for Medium - Range Weather Forecasts numerical weather prediction model, the historical data of the Global Forecast System model of the National Centers for Environmental Prediction of the United States, and the data of the China Meteorological Administration's Wind Energy and Solar Energy Meteorological Forecast System respectively; the temporal interpolation method is used to obtain the data for a preset time period in the historical data of the European Centre for Medium - Range Weather Forecasts numerical weather prediction model, the historical data of the Global Forecast System model of the National Centers for Environmental Prediction of the United States, and the data of the China Meteorological Administration's Wind Energy and Solar Energy Meteorological Forecast System respectively.

[0055] In some embodiments, the spatial interpolation method is the four - point interpolation method, which is a method of extracting the meteorological data of the four nearest grid points corresponding to the longitude and latitude of the photovoltaic power station; for the temporal interpolation method, for variables with obvious time - varying characteristics such as surface solar radiation and air temperature, the cubic spline interpolation method is used, and for variables such as wind speed and wind direction, the linear interpolation method is used.

[0056] The pre - trained model further includes: dividing the historical multi - mode meteorological forecast data and the historical photovoltaic power generation forecast data into a training set and a validation set in a ratio of 8:2, establishing a pre - trained model based on the XGBoost algorithm, and using the grid search method to optimize the hyperparameters of the constructed pre - trained model to optimize the photovoltaic power correction model.

[0057] In an alternative embodiment, it is necessary to dynamically test the prediction effect of the pre - trained model using real - time photovoltaic power generation data, and the evaluation indicators are the qualification rate (QR) and the accuracy rate (CR). Among them, the root - mean - square error (ERMSE) is required to calculate the accuracy rate (CR) of photovoltaic power generation, and the calculation formula is as follows:

[0058]

[0059] The calculation formula for the qualified rate (QR) of photovoltaic power prediction is as follows:

[0060]

[0061] The calculation formula for the accuracy rate (CR) of photovoltaic power prediction is as follows:

[0062] C R =(1 - E RMSE )×100%;

[0063] In the formula, B i is the judgment result of the prediction qualified rate at time i, and the calculation formula is:

[0064]

[0065] where n is the number of samples; P Pi is the actual power at time i; P Mi is the predicted power at time i; C i is the installed capacity at time i.

[0066] The accuracy rate (A) and the qualified rate (Q) are also required for the verification of the correction effect of the surface solar radiation. Among them, the calculation formula for the accuracy rate (A) is:

[0067]

[0068] The calculation formula for the qualified rate (Q) is:

[0069]

[0070] Qi is the judgment result of the qualified rate at time i, and the calculation formula is:

[0071]

[0072] where, in the formula, the predicted total surface solar radiation at the i-th moment of the day, is the measured total surface solar radiation at the i-th moment of the day (or the corresponding moment of the 240-minute forecast), is the maximum value of the predicted total surface solar radiation of the day, is the maximum value of the measured total surface solar radiation of the day, and n is the total number of samples of the total surface solar radiation.

[0073] Using the above inspection method, taking a certain photovoltaic power station in Hebei as an example, the prediction accuracy and qualification rate of photovoltaic power generation before and after correction and before and after secondary correction are compared. After correction, the prediction accuracy of the surface solar radiation forecast can be increased by 1% - 3%, and after secondary correction, the prediction accuracy of the photovoltaic power generation forecast can be increased by 2% - 4%.

[0074] Table 1 Comparison of Photovoltaic Power Generation Power Prediction Results

[0075]

[0076] Table 2 Comparison of Surface Solar Radiation Correction Results

[0077]

[0078] As can be seen from Table 1 and Table 2 above, by adopting the photovoltaic power generation power prediction method of the present disclosure, the deficiency of the low prediction accuracy of the surface solar radiation by a single numerical prediction model is made up. The method of predicting short-term photovoltaic power generation power through meteorological parameters output by multiple numerical prediction models reduces the uncertainty of the prediction. At the same time, the secondary correction method can also further improve the prediction accuracy of the photovoltaic power generation power forecast.

[0079] According to a specific embodiment of the present disclosure, on the other hand, a photovoltaic power generation power prediction system is provided, configured to perform photovoltaic power generation power prediction by executing the method described in any one of the above embodiments.

[0080] According to a specific embodiment of the present disclosure, on the other hand, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the method described in any one of the above embodiments is implemented.

[0081] According to a specific embodiment of the present disclosure, on the other hand, an electronic device is provided, which may include: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the method described in any one of the above embodiments.

[0082] According to a specific embodiment of the present disclosure, on the other hand, an electronic device is provided, and this device is used for the photovoltaic power generation power prediction method. The electronic device may include: at least one processor; and a memory communicatively connected to the at least one processor; wherein,

[0083] The memory stores instructions executable by the one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: obtain multi-mode weather forecast data of a photovoltaic power station to be predicted; input the multi-mode weather forecast data into a pre-trained model to obtain corrected power generation power prediction data; wherein, the pre-trained model includes: obtaining historical multi-mode weather forecast data of the photovoltaic power station to be predicted and historical photovoltaic power station observation data, wherein the historical photovoltaic power station observation data includes historical surface solar radiation data and historical photovoltaic power generation data; constructing a surface solar radiation correction model based on the historical multi-mode weather forecast data and the historical surface solar radiation data; inputting the historical multi-mode weather forecast data into the surface solar radiation correction model to obtain corrected surface solar radiation data; constructing a photovoltaic power generation power prediction model based on the historical multi-mode weather forecast data, the historical photovoltaic power generation data and the corrected surface solar radiation data; inputting the historical multi-mode weather forecast data into the photovoltaic power generation power prediction model to obtain predicted photovoltaic power generation data; and constructing a photovoltaic power generation power correction model based on the historical multi-mode weather forecast data and the predicted photovoltaic power generation data.

[0084] Embodiments of the present disclosure provide a non-volatile computer storage medium storing computer-executable instructions that can execute the photovoltaic power generation power prediction method in any of the above method embodiments.

[0085] Refer to the following Figure 4 , which shows a schematic structural diagram of an electronic device suitable for implementing embodiments of the present disclosure. The terminal device in the embodiments of the present disclosure may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Tablet Computers), PMPs (Portable Multimedia Players), in-vehicle terminals (such as in-vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The electronic device shown is only an example and should not impose any limitations on the functions and usage scope of the embodiments of the present disclosure.

[0086] As shown in Figure 4As shown, the electronic device may include a processing system (such as a central processing unit, a graphics processing unit, etc.) 301, which may perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage system 308 into a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device are also stored. The processing system 301, the ROM 302, and the RAM 303 are connected to each other through a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0087] Generally, the following systems may be connected to the I / O interface 305: an input system 306 that may include, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output system 307 that may include, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage system 308 that may include, for example, a magnetic tape, a hard disk, etc.; and a communication system 309. The communication system 309 may allow the electronic device to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 4 an electronic device with various systems is shown, it should be understood that it is not required to implement or have all the shown systems. Instead, more or fewer systems may be implemented or had.

[0088] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart may be implemented as a computer software program. For example, an embodiment of the present disclosure may include a computer program product, which may include a computer program carried on a computer-readable medium, and the computer program includes program codes for performing the method shown in the flowchart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication system 309, or installed from the storage system 308, or installed from the ROM 302. When the computer program is executed by the processing system 301, the above functions defined in the method of the embodiment of the present disclosure are executed.

[0089] It should be noted that the above computer-readable medium in the present disclosure can be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present disclosure, the computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, and the computer-readable signal medium can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0090] The above computer-readable medium can be included in the above electronic device; or it can exist separately without being assembled into the electronic device.

[0091] The above computer-readable medium carries one or more programs. When the above one or more programs are executed by the electronic device, the electronic device is enabled: The photovoltaic power prediction method provided in the present disclosure makes up for the deficiency of the low accuracy of predicting surface solar radiation by a single numerical prediction model. The method of predicting short-term photovoltaic power through meteorological parameters output by multiple numerical prediction models reduces the uncertainty of the prediction. At the same time, the secondary correction method can further improve the accuracy of the photovoltaic power prediction.

[0092] Alternatively, the above computer-readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to: The photovoltaic power prediction method provided by the present disclosure makes up for the deficiency of the low accuracy rate of predicting surface solar radiation by a single numerical prediction model. The method of predicting short-term photovoltaic power through meteorological parameters output by multiple numerical prediction models reduces the uncertainty of the prediction. At the same time, the secondary correction method can further improve the accuracy rate of photovoltaic power prediction.

[0093] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages or combinations thereof. The programming languages may include object-oriented programming languages such as Java, Smalltalk, C++, and may also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may execute entirely on the user's computer, partially on the user's computer, execute as a stand-alone software package, execute partially on the user's computer and partially on a remote computer, or execute entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, which may include a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0094] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that, in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system for performing the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0095] The present disclosure aims to protect a photovoltaic power prediction method and system, a storage medium, and an electronic device. The photovoltaic power prediction method may include: obtaining multi-model meteorological forecast data of a photovoltaic power station to be predicted; inputting the multi-model meteorological forecast data into a pre-trained model to obtain corrected power generation prediction data. Among them, the pre-trained model includes: obtaining historical multi-model meteorological forecast data and historical photovoltaic power station observation data of the photovoltaic power station to be predicted, where the historical photovoltaic power station observation data includes historical surface solar radiation data and historical photovoltaic power generation data; constructing a surface solar radiation correction model based on the historical multi-model meteorological forecast data and the historical surface solar radiation data; inputting the historical multi-model meteorological forecast data into the surface solar radiation correction model to obtain corrected surface solar radiation data; constructing a photovoltaic power prediction model based on the historical multi-model meteorological forecast data, the historical photovoltaic power generation data, and the corrected surface solar radiation data; inputting the historical multi-model meteorological forecast data into the photovoltaic power prediction model to obtain predicted photovoltaic power data; constructing a photovoltaic power correction model based on the historical multi-model meteorological forecast data and the predicted photovoltaic power data. The photovoltaic power prediction method provided by the present disclosure makes up for the deficiency of the low accuracy rate of predicting surface solar radiation by a single numerical prediction model. The method of predicting short-term photovoltaic power by meteorological parameters output by multiple numerical prediction models reduces the uncertainty of the prediction. At the same time, the secondary correction method can further improve the accuracy rate of the photovoltaic power prediction.

[0096] Finally, it should be noted that the embodiments in this specification are described in a progressive manner, and the key points of each embodiment are the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other. For the system or device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0097] The above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them; although the present disclosure has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A photovoltaic power generation power prediction method, characterized in that: include: Obtain multi-mode meteorological forecast data for the photovoltaic power station to be predicted; Inputting the multi-mode weather forecast data into a pre-trained model to obtain revised power generation prediction data; Wherein, the pre-training model includes: Acquire historical multi-mode meteorological forecast data and historical photovoltaic site observation data of the photovoltaic power station to be predicted, wherein the historical photovoltaic site observation data includes historical surface solar radiation data and historical photovoltaic power generation data; Constructing a surface solar radiation correction model based on the historical multi-mode meteorological forecast data and the historical surface solar radiation data; Inputting the historical multi-mode meteorological forecast data into the surface solar radiation correction model to obtain corrected surface solar radiation data; Constructing a photovoltaic power generation prediction model based on the historical multi-mode meteorological forecast data, the historical photovoltaic power generation data and the revised surface solar radiation data; Inputting the historical multi-mode meteorological forecast data into the photovoltaic power generation prediction model to obtain predicted photovoltaic power generation data; A photovoltaic power generation power correction model is constructed based on the historical multi-mode meteorological forecast data and the predicted photovoltaic power generation power data.

2. The photovoltaic power generation prediction method according to claim 1, characterized in that: The historical photovoltaic power generation data includes: The latitude and longitude information of the photovoltaic power station at preset time intervals, the surface solar radiation data and the corresponding photovoltaic power generation data; The preset time period is 15 minutes.

3. The photovoltaic power generation prediction method according to claim 1, characterized in that: The acquisition of historical multi-mode meteorological forecast data and historical photovoltaic station observation data of the photovoltaic power station to be predicted includes: Observation data cleaning, the surface solar radiation data less than 0W / m 2 and greater than 2000W / m 2 The data of photovoltaic power generation less than 0MW and greater than 50MW were eliminated; When a value remains unchanged for more than 3 hours, data quality control is performed based on the actual situation.

4. The photovoltaic power generation prediction method according to claim 1, characterized in that: The pre-trained model also includes: Using the XGBoost algorithm to construct the surface solar radiation correction model; Using the XGBoost algorithm to construct the photovoltaic power generation prediction model; The photovoltaic power generation power correction model is constructed using the XGBoost algorithm.

5. The photovoltaic power generation prediction method according to claim 2, characterized in that: The acquisition of historical multi-mode meteorological forecast data and historical photovoltaic station observation data of the photovoltaic power station to be predicted includes: Access historical data from the European Centre for Medium-Range Weather Numerical Prediction Model; Access to historical data from the National Centers for Environmental Prediction Global Forecast System model; Obtain data from the China Meteorological Administration's wind and solar energy weather forecast system; The spatial interpolation method is used to determine the location of the photovoltaic power station to be predicted in the historical data of the European Center for Medium-Term Weather Numerical Forecast Model, the historical data of the Global Forecast System Model of the National Center for Environmental Prediction of the United States, and the data of the Wind and Solar Meteorological Forecast System of the China Meteorological Administration. The time interpolation method is used to obtain data for the preset time period from the historical data of the European Center for Medium-Term Weather Numerical Forecast Model, the historical data of the Global Forecast System Model of the National Centers for Environmental Prediction of the United States, and the data of the Wind and Solar Energy Meteorological Forecast System of the China Meteorological Administration.

6. The photovoltaic power generation prediction method according to claim 5, characterized in that: The spatial interpolation method is a four-point interpolation method, which is a method for extracting meteorological data of the four closest grid points in the mode corresponding to the longitude and latitude of the photovoltaic power generation photovoltaic power station; The time interpolation method uses cubic spline interpolation for variables with obvious time variation characteristics, such as surface solar radiation and temperature, and linear interpolation for variables such as wind speed and wind direction.

7. The photovoltaic power generation prediction method according to claim 1, characterized in that: The pre-trained model also includes: The historical multi-mode meteorological forecast data and the historical photovoltaic power forecast data are divided into a training set and a validation set in a ratio of 8:

2. A pre-training model is established based on the XGBoost algorithm. The hyperparameters of the constructed pre-training model are optimized using the grid search method to optimize the photovoltaic power correction model.

8. A photovoltaic power generation prediction system, characterized in that: The method is configured to predict photovoltaic power generation by executing the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: include: one or more processors; A storage device, used to store one or more programs, when the one or more programs are executed by the one or more processors, enables the one or more processors to implement the method according to any one of claims 1 to 7.

Images