Optical power prediction method based on WRF-Solar assimilation multi-source data and related device

Abstract

The invention provides an optical power prediction method based on WRF-Solar assimilation multi-source data and a related device, and belongs to the technical field of optical power prediction. According to the method, a WRF-Solar model and a WRF-DA 4D-Var assimilation framework are constructed, preprocessed historical meteorological data and historical radiation observation data are input into the WRF-Solar model and the WRF-DA 4D-Var assimilation framework, and a prediction result using the historical meteorological data and the historical radiation observation data is obtained; performing error analysis on prediction results of the historical meteorological data and the historical radiation observation data to obtain an optimal WRF-Solar model parameter and an optimal WRF-DA 4D-Var assimilation parameter; the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters are utilized to construct a final prediction model; predicting the solar radiation and meteorological elements to be predicted by using the final prediction model to obtain prediction results of the solar radiation and meteorological elements; and obtaining an optical power prediction result based on the prediction results of the solar radiation and the meteorological elements. According to the invention, the problem of low accuracy of optical power prediction is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of optical power prediction, and in particular relates to an optical power prediction method based on WRF-Solar assimilation of multi-source data and a related device. Background Art

[0002] As a clean renewable energy generation method, the output power of photovoltaic power generation is directly affected by the intensity of solar radiation. Solar radiation has significant temporal and spatial variation characteristics, especially in the short and ultra-short term, and is affected by multiple factors such as clouds, aerosols and meteorological conditions. The actual output of photovoltaic power generation has a strong fluctuation characteristic during this period. Therefore, accurate prediction of the intensity and temporal and spatial variation of solar radiation is the key to improving the power generation efficiency and economic benefits of photovoltaic power stations.

[0003] The methods currently used in the prior art for predicting optical power are: 1) predicting optical power based on statistical models; 2) predicting optical power based on physical models; 3) predicting optical power based on machine learning algorithms.

[0004] When predicting light power based on statistical models (such as ARMA model and ARIMA model), this method usually relies on historical meteorological data and historical photovoltaic power data. By mining the potential laws between historical data, it can provide ultra-short-term and short-term prediction results. However, this method has limitations in complex meteorological conditions, especially poor response to sudden weather changes. This method often cannot accurately capture instantaneous changes through the average statistical pattern of historical data, resulting in low accuracy in solar light resource prediction. ARMA stands for Auto Regressive Moving Average Model, which means autoregressive moving average model in Chinese. ARIMA stands for Auto Regressive Integrated Moving Average Model, which means autoregressive integrated moving average model in Chinese.

[0005] When the method based on physical model predicts optical power, this method is based on atmospheric physical processes and solar radiation transfer models, and uses numerical weather forecast models (such as WRF, Weather Research and Forecasting) to simulate meteorological elements and solar radiation. This method relies on the simulation of atmospheric physical processes. By applying the basic principles of atmospheric physics and radiation transfer, combined with meteorological observation data, it generates weather forecast results for a period of time in the future, including temperature, humidity, wind direction, wind speed, cloud cover and radiation. However, this method faces difficulties in dealing with cloud characteristics (such as rapid changes in clouds, complex spatial and temporal distribution characteristics, diversity of cloud thickness and height, and complex microphysical processes in clouds). Moreover, this method has certain uncertainties in the prediction and simulation of meteorological elements and light resources. These uncertainties mainly come from errors in the parameterization scheme of physical processes, uncertainties in the initial and boundary conditions of the model, model resolution limitations, errors in meteorological inputs, and errors caused by the limited ability of the model to handle local micrometeorological phenomena.

[0006] When predicting optical power based on machine learning algorithms, this method requires a large amount of training data to capture the complex relationship between meteorological factors and optical power. When the amount of data is insufficient or the data quality is poor, the prediction effect of this method is often not ideal. Moreover, the machine learning algorithm model is a purely data-driven model that lacks physical constraints and does not consider the physical relationship between meteorological factors or the principles of atmospheric dynamics. Under extreme weather conditions, this method is difficult to give a more reliable prediction result. Summary of the invention

[0007] The object of the present invention is to provide a method and related device for optical power prediction based on WRF-Solar assimilation of multi-source data, so as to solve the problem of low accuracy of optical power prediction in the prior art.

[0008] In order to achieve the above object, the present invention adopts the following technical solution: In a first aspect, the present invention provides an optical power prediction method based on WRF-Solar assimilation of multi-source data, comprising the following steps: Obtain historical meteorological data and historical radiation observation data, pre-process the obtained historical meteorological data and historical radiation observation data, remove outliers and correct deviations; Construct the WRF-Solar model and WRF-DA 4D-Var assimilation framework, input the preprocessed historical meteorological data and historical radiation observation data into the WRF-Solar model and WRF-DA 4D-Var assimilation framework, and obtain the prediction results using the historical meteorological data and historical radiation observation data; The error analysis of the prediction results using historical meteorological data and historical radiation observation data was performed to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters; The final prediction model was constructed using the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters; The final prediction model is used to predict the solar radiation and meteorological elements to be predicted, and the prediction results of the solar radiation and meteorological elements are obtained; Based on the prediction results of solar radiation and meteorological elements, the optical power prediction results are obtained.

[0009] A further improvement of the present invention is that in the steps of obtaining historical meteorological data and historical radiation observation data, preprocessing the obtained historical meteorological data and historical radiation observation data, removing outliers and correcting deviations, spatial and temporal interpolation processing is also performed on the obtained historical radiation observation data to keep consistent with the temporal and spatial resolution of the WRF-Solar model.

[0010] A further improvement of the present invention is that, in the step of performing error analysis on the prediction results of historical meteorological data and historical radiation observation data to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters, specific error analysis indicators include mean absolute error, mean deviation, root mean square error and correlation coefficient.

[0011] A further improvement of the present invention is that, in the step of performing error analysis on the prediction results of historical meteorological data and historical radiation observation data to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters, the specific error analysis process is: when the indicators of the error analysis are within a preset threshold range, the WRF-Solar model parameters and WRF-DA 4D-Var assimilation parameters with the smallest error values ​​are used as the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters.

[0012] A further improvement of the present invention is that, when the index of the error analysis is not within a preset threshold range, the WRF-Solar model parameters and the WRF-DA 4D-Var assimilation parameters are adjusted until the index of the error analysis is within a preset threshold range.

[0013] A further improvement of the present invention is that, in the step of obtaining the optical power prediction result based on the prediction results of solar radiation and meteorological elements, the optical power prediction result is obtained based on the prediction results of solar radiation and meteorological elements using the pvlib-python tool.

[0014] A further improvement of the present invention is that the cost function expression of the WRF-DA 4D-Var assimilation framework is:

[0015] in, is the cost function, For the initial moment The mode status, is the background field, i.e. the initial moment The pattern prior, is the background error covariance matrix, The number of time steps in the assimilation time window, For the model from the initial moment arrive The forward model operator at time , integrated to get Mode status at the moment , for The observation vector at time, is the observation operator, for The observation error covariance matrix at time , To set the initial mode state By integrating the model Mode status at the moment , and then through the observation operator Project it into the observation space and get the value of the model simulation in the observation space.

[0016] In a second aspect, the present invention provides an optical power prediction system based on WRF-Solar assimilation of multi-source data, including a data acquisition module, a data preprocessing module, a historical data prediction result acquisition module, an optimal parameter acquisition module, a final prediction model construction module, a solar radiation and meteorological element prediction module and an optical power prediction module; The data acquisition module is used to acquire historical meteorological data and historical radiation observation data; The data preprocessing module is used to preprocess the obtained historical meteorological data and historical radiation observation data, remove abnormal values ​​and correct deviations; The historical data prediction result acquisition module is used to construct the WRF-Solar model and the WRF-DA 4D-Var assimilation framework, input the pre-processed historical meteorological data and the historical radiation observation data into the WRF-Solar model and the WRF-DA4D-Var assimilation framework, and obtain the prediction results using the historical meteorological data and the historical radiation observation data; The optimal parameter acquisition module is used to perform error analysis on the prediction results using historical meteorological data and historical radiation observation data to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters; The final prediction model construction module is used to construct the final prediction model using the optimal WRF-Solar model parameters and the optimal WRF-DA4D-Var assimilation parameters; The solar radiation and meteorological element prediction module is used to predict the solar radiation and meteorological elements to be predicted using the final prediction model to obtain the prediction results of the solar radiation and meteorological elements; The optical power prediction module is used to obtain an optical power prediction result based on the prediction results of solar radiation and meteorological elements.

[0017] In a third aspect, the present invention provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the optical power prediction method based on WRF-Solar assimilation of multi-source data introduced above are implemented.

[0018] In a fourth aspect, the present invention provides a storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the optical power prediction method based on WRF-Solar assimilation of multi-source data introduced above are implemented.

[0019] Compared with the prior art, the present invention has the following beneficial effects: The present invention belongs to an improved invention. Compared with the existing optical power prediction method based on WRF-Solar assimilation of multi-source data, the optical power prediction method based on WRF-Solar assimilation of multi-source data proposed by the present invention uses the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters to construct a final prediction model, and uses the final prediction model to predict the solar radiation and meteorological elements to be predicted, and obtains the prediction results of solar radiation and meteorological elements, and obtains the optical power prediction results based on the prediction results of solar radiation and meteorological elements. It can be seen that the present invention uses the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters to construct the final prediction model to obtain the prediction results of solar radiation and meteorological elements, and then obtains the optical power prediction results based on the prediction results of solar radiation and meteorological elements. On the one hand, the present invention adopts the WRF-Solar model specially designed for solar energy resource prediction, and captures local weather phenomena and solar radiation changes through high resolution, which has stronger prediction ability than the traditional statistical model WRF-Solar model. On the other hand, the present invention adopts a combination of the WRF-Solar model and the WRF-DA 4D-Var data assimilation technology to improve the capture and simulation capabilities of different weather systems (such as storms and thick cloud cover), improve the prediction accuracy of solar radiation and meteorological elements, and effectively solve the problem of low accuracy of optical power prediction in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a flow chart of the optical power prediction method based on WRF-Solar assimilation of multi-source data of the present invention; Figure 2 It is a schematic diagram of the optical power prediction system based on WRF-Solar assimilation of multi-source data of the present invention; Figure 3 It is a schematic diagram of the structure of the electronic device of the present invention. DETAILED DESCRIPTION

[0021] In order to further understand the content of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the embodiments are only for explaining the present invention and are not intended to limit it.

[0022] The optical power prediction method based on WRF-Solar assimilation of multi-source data proposed in the present invention uses the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters to build a final prediction model, uses the final prediction model to predict solar radiation and meteorological elements to obtain prediction results of solar radiation and meteorological elements, and obtains optical power prediction results based on the prediction results of solar radiation and meteorological elements. Compared with the prior art, the present invention effectively solves the problem of low accuracy of optical power prediction.

[0023] Embodiment 1: The flowchart of the optical power prediction method based on WRF-Solar assimilation of multi-source data in the present invention is as follows: Figure 1 As shown, the optical power prediction method based on WRF-Solar assimilation of multi-source data of the present invention comprises the following steps: S1. Obtain historical meteorological data and historical radiation observation data, pre-process the obtained historical meteorological data and historical radiation observation data, eliminate outliers and correct deviations.

[0024] S2. Construct the WRF-Solar model and WRF-DA 4D-Var assimilation framework, input the preprocessed historical meteorological data and historical radiation observation data into the WRF-Solar model and WRF-DA 4D-Var assimilation framework, and obtain the prediction results using historical meteorological data and historical radiation observation data.

[0025] S3. Error analysis is performed on the prediction results using historical meteorological data and historical radiation observation data to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters.

[0026] S4. Construct the final prediction model using the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters.

[0027] S5. The final prediction model is used to predict the solar radiation and meteorological elements to be predicted, and the prediction results of the solar radiation and meteorological elements are obtained.

[0028] S6. Based on the prediction results of solar radiation and meteorological elements, obtain the optical power prediction results.

[0029] Embodiment 2: The schematic diagram of the optical power prediction system based on WRF-Solar assimilation of multi-source data in the present invention is as follows Figure 2As shown, the optical power prediction system based on WRF-Solar assimilation of multi-source data of the present invention includes a data acquisition module, a data preprocessing module, a historical data prediction result acquisition module, an optimal parameter acquisition module, a final prediction model construction module, a solar radiation and meteorological element prediction module and an optical power prediction module.

[0030] The data acquisition module is used to obtain historical meteorological data and historical radiation observation data.

[0031] The data preprocessing module is used to preprocess the historical meteorological data and historical radiation observation data, eliminate outliers and correct deviations.

[0032] The historical data prediction result acquisition module is used to build the WRF-Solar model and the WRF-DA 4D-Var assimilation framework, input the preprocessed historical meteorological data and historical radiation observation data into the WRF-Solar model and the WRF-DA 4D-Var assimilation framework, and obtain the prediction results using the historical meteorological data and historical radiation observation data.

[0033] The optimal parameter acquisition module is used to perform error analysis on the prediction results using historical meteorological data and historical radiation observation data to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters.

[0034] The final prediction model building module is used to build the final prediction model using the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters.

[0035] The solar radiation and meteorological element prediction module is used to predict the solar radiation and meteorological elements to be predicted using the final prediction model to obtain the prediction results of solar radiation and meteorological elements.

[0036] The optical power prediction module is used to obtain optical power prediction results based on the prediction results of solar radiation and meteorological elements.

[0037] Embodiment 3: S1. Obtain historical meteorological data and historical radiation observation data, pre-process the obtained historical meteorological data and historical radiation observation data, eliminate outliers and correct deviations.

[0038] First, historical meteorological data (such as the US National Centers for Environmental Prediction NECP Global Forecast System (GFS) dataset including the locations of photovoltaic power plants) and historical radiation observation data (such as the radiation observation dataset of the Japanese Himawari-8 satellite, ground observation station observation data, and laser cloud radar observation data) are obtained.

[0039] Then the historical meteorological data and historical radiation observation data are preprocessed to remove outliers and correct deviations.

[0040] In this step, the historical radiation observation data are also interpolated in space and time to make them consistent with the time and space resolution of the WRF-Solar model, which is convenient for the WRF-Solar model to assimilate the observation data.

[0041] S2. Construct the WRF-Solar model and WRF-DA 4D-Var assimilation framework, input the preprocessed historical meteorological data and historical radiation observation data into the WRF-Solar model and WRF-DA 4D-Var assimilation framework, and obtain the prediction results using historical meteorological data and historical radiation observation data.

[0042] The cost function expression of the WRF-DA 4D-Var assimilation framework is:

[0043] in, is the cost function, For the initial moment The mode status, is the background field, i.e. the initial moment The pattern prior, is the background error covariance matrix, The number of time steps in the assimilation time window, For the model from the initial moment arrive The forward model operator at time , integrated to get Mode status at the moment , for The observation vector at time, is the observation operator, for The observation error covariance matrix at time , To set the initial mode state By integrating the model Mode status at the moment , and then through the observation operator Project it into the observation space and get the value of the model simulation in the observation space.

[0044] The first term in the cost function expression of the WRF-DA 4D-Var assimilation framework Represents the initial state of the weighted background error covariance matrix With background field , which can make the optimized initial state not deviate too much from the background field, unless the observational data provides stronger evidence.

[0045] The second term in the cost function expression of the WRF-DA 4D-Var assimilation framework It represents the sum of the deviations between the results mapped to the observation space by the WRF-Solar model simulation and the actual observations in the entire assimilation time window. By minimizing this term to optimize the initial state, the simulation results of the WRF-Solar model can be optimized at time and better fit to the observed data.

[0046] Conventional meteorological elements that can be predicted by the WRF-Solar model include temperature, humidity, and air pressure. Using linear observation operators, algebraic observation operators, vertical interpolation observation operators, localized observation operators, and remote sensing observation operators, the results of the WRF-Solar model simulation are projected into the observation space to compare with the observation results of the observation site. Since both the satellite observation channel and the ground-based lidar observe radiation data, the meteorological elements predicted by the WRF-Solar model need to be converted into radiation through the radiation transfer model RTM, and then projected into the observation space of the satellite and lidar for comparison and assimilation. Common radiation transfer models include CRTM (Community Radiative Transfer Model) and RTTOV (Radiative Transfer for TOVS). The function of the radiation transfer model RTM is to convert physical quantities such as temperature and humidity predicted by the WRF-Solar model into brightness temperature observed by satellites. The specific configuration steps of the radiation transfer model RTM are as follows: The WRF-Solar model can use the RTM model to simulate the radiation information on the Himawari-8 observation channel, and calculate the radiation brightness temperature of the corresponding satellite observation channel through the cloud micro- and macro-physical properties predicted by the WRF-Solar model and the environmental background elements (such as cloud particle size, shape, direction, cloud amount, cloud top temperature, morphology, optical thickness, aerosol, temperature, humidity, water vapor, surface albedo and vertical temperature profile, etc.). The WRF-DA 4D-Var assimilation system allows the required RTM model to be adopted through the namelist parameter configuration file. The radiation transfer model RTM used in this embodiment is the CRTM radiation transfer model.

[0047] The reasons why the CRTM radiation transmission model is adopted in this embodiment are described as follows: 1) CRTM has been integrated into the WRF-DA 4D-Var assimilation system, has no licensing restrictions, and can be automatically compiled; 2) CRTM largely solves the atmospheric scattering and surface emissivity simulation, allowing more satellite data to be accurately assimilated; 3) CRTM supports radiation transfer simulation of multiple satellite sensors, including meteorological satellites such as Himawari-8, GOES, MODIS and METEOSAT. The spectral channels of these satellites usually cover visible light, infrared and microwave bands (for example, the 16 channels of Himawari-8 satellite cover a wide spectral range from visible light to infrared), and CRTM can process these different bands and perform corresponding radiation brightness temperature simulation.

[0048] 4) CRTM is able to take into account the absorption of atmospheric gases and the multiple scattering effects of water clouds, ice clouds and various aerosols, providing fast and accurate satellite radiance simulations under various weather and surface conditions.

[0049] The following are some of the parameters that need to be configured in this step: A. Basic WRF configuration parameters The WRF mode (also called the WRF model) uses a three-layer bidirectional nested grid setting. The resolution of the horizontal grid from the outer layer to the inner layer is 9km, 3km and 1km. The innermost grid covers the area where the photovoltaic power station site is located. The number of vertical grid layers is set to 40, 50 and 50 from the outer to the inner. The time step of the WRF model integration can be selected according to the grid resolution and physical scheme. The higher the resolution, the shorter the time step is required to ensure the stability of the model. The initial conditions and boundary conditions that drive the operation of the WRF model are derived from the NCEP Global Forecast System GFS forecast data. The WRF forecast results are output once an hour.

[0050] B. Planetary Boundary Layer Scheme Configuration The planetary boundary layer scheme affects the turbulent mixing process in the lower atmosphere, which in turn affects the temperature, humidity and wind speed near the ground. These parameters will directly affect the optical power prediction. The planetary boundary layer scheme can adopt the YSU boundary layer scheme or the MYNN boundary layer scheme. The YSU boundary layer scheme is a non-local mixing scheme, which is suitable for a variety of weather and complex terrain conditions. The MYNN boundary layer scheme is a local and non-local mixing scheme, which is suitable for complex airflow and terrain conditions.

[0051] C. Near-ground layer solution configuration The near-surface layer scheme adopts the Monin-Obukhov near-surface layer scheme, which is based on the Monin-Obukhov theory and is mainly used to describe boundary layer atmospheric processes under stable and unstable conditions, especially the exchange of temperature, humidity and momentum in the near-surface layer (usually within tens of meters).

[0052] D. Land surface process scheme configuration The land surface process simulates the energy exchange process between the atmosphere and the surface, which will affect the simulation of parameters such as surface temperature, humidity and surface albedo. The Noah LSM land surface hydrological model scheme can be used here, which is suitable for various climatic conditions and can well simulate the energy exchange process between the surface and the atmosphere.

[0053] E. Microphysical process scheme configuration Microphysical schemes are used to describe the generation, development and dissipation of water-forming substances (such as clouds, rain, snow and ice) in the atmosphere. The use of appropriate microphysical schemes is crucial for radiation simulation. The microphysical scheme can adopt the Thompson microphysical scheme or the WSM6 microphysical scheme. The Thompson microphysical scheme or the WSM6 microphysical scheme can accurately simulate multiple types of water-forming substances, thereby accurately predicting the impact of clouds on solar radiation.

[0054] F. Convection parameterization scheme configuration The Kain-Fritsch convection parameterization scheme is used for the outer grid (usually because the outer grid resolution is too low to resolve convection), while the convection parameterization scheme is not required for the inner grid (this is because when the inner grid resolution reaches 3 km or higher, the WRF model can directly resolve mesoscale convection activity without relying on the convection parameterization scheme).

[0055] G. Radiation solution configuration The long-wave radiation scheme uses the RRTMG long-wave radiation scheme, and the short-wave radiation scheme uses the Dudhia short-wave radiation scheme, which can accurately simulate the long-wave and short-wave radiation in the atmosphere and are suitable for the application scenario of optical power prediction. The solar zenith angle tracking and correction options are enabled in the WRF-Solar model to ensure that the changes in the solar altitude angle at different times are accurately simulated.

[0056] H. Construction of background error covariance and observation error covariance in assimilation process a. The construction of the background error covariance matrix is ​​the core of data assimilation. The matrix reflects the errors and correlations of model state variables (such as temperature, humidity and wind speed) and can usually be constructed through the NMC method or the ensemble Kalman filter (EnKF).

[0057] b. The observation error covariance matrix is ​​used to represent the errors and correlations of observation data. The errors of satellite, ground and lidar observations are estimated, and the error levels of various types of observation data are reasonably set to construct the observation error covariance matrix.

[0058] The contents of some parameters configured in this step are described as follows: The WRF mode of this embodiment adopts a three-layer bidirectional nested grid setting. The resolution of the horizontal grid from the outer layer to the inner layer is 9km, 3km and 1km. The innermost grid covers the area where the photovoltaic power station site is located, and the number of vertical grid layers is set to 40, 50 and 50 from the outside to the inside. The nested grids in the three-layer nested area are marked as d0, d1 and d2 from the outside to the inside. The number of grid points of the d0, d1 and d2 grids is set to 60×60, 80×80 and 100×100 respectively. The latitude range of the d0 grid is 26.70°N~31.68°N, and the longitude range is 117.71°E~123.17°E. The latitude range of the d1 grid is 28.13°N~30.25°N, and the longitude range is 119.50°E~122.00°E. The latitude range of the d2 grid is 28.73°N~29.73°N, and the longitude range is 120.43°E~121.45°E. The time step of the WRF model integration is 45s, 15s, and 5s from the outer grid to the inner grid. The initial conditions and boundary conditions driving the WRF model operation are all derived from the NCEP Global Forecast System GFS forecast data, and the WRF forecast results are output once an hour.

[0059] The planetary boundary layer scheme of this embodiment adopts the YSU boundary layer scheme. The near-surface layer scheme adopts the Monin-Obukhov near-surface layer scheme. The land surface process scheme adopts the Noah LSM land surface hydrological model scheme. The microphysical process scheme adopts the Thompson microphysical scheme and the WSM6 microphysical scheme, and the optimal schemes are subsequently determined based on sensitivity tests. The long-wave radiation scheme adopts the RRTMG long-wave radiation scheme, and the short-wave radiation scheme adopts the Dudhia scheme. The d0 grid adopts the Kain-Fritsch convection parameterization scheme, and the d1 and d2 grids do not use the convection parameterization scheme. The NMC method is used to construct the background error covariance matrix. The errors of satellite, ground and laser cloud radar observations are estimated, the error levels of various types of observation data are reasonably set, and the observation error covariance matrix is ​​constructed. The assimilation time window is set to 6 hours.

[0060] S3. Error analysis is performed on the prediction results using historical meteorological data and historical radiation observation data to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters.

[0061] The error analysis of the prediction results (such as solar radiation and meteorological element prediction results) using historical meteorological data and historical radiation observation data was performed to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters. Solar radiation specifically includes GHI (Global Horizontal Irradiance), DNI (Direct Normal Irradiance) and DHI (Diffuse Horizontal Irradiance).

[0062] In this step, specific indicators for error analysis of meteorological element prediction results include mean absolute error, mean deviation, root mean square error and correlation coefficient.

[0063] In this step, the specific error analysis process is as follows: when the error analysis indicators (mean absolute error, mean deviation, root mean square error and correlation coefficient) are within the preset threshold (the preset threshold can be set according to the actual simulation results), the WRF-Solar model parameters and WRF-DA 4D-Var assimilation parameters with the smallest error values ​​are used as the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters.

[0064] The formula for calculating the mean absolute error is:

[0065] in, is the mean absolute error, is the first i The predicted value, The first i Observations, N is the number of samples in the model assimilated forecast ensemble or surface meteorological observation ensemble.

[0066] The formula for calculating the mean deviation is:

[0067] in, is the mean deviation, is the first i The predicted value, The first i Observations, N is the number of samples in the model assimilated forecast ensemble or surface meteorological observation ensemble.

[0068] The formula for calculating the root mean square error is:

[0069] in, is the root mean square error, The first i Observations, is the first i The predicted value, N is the number of samples in the model assimilated forecast ensemble or surface meteorological observation ensemble.

[0070] The formula for calculating the correlation coefficient is:

[0071] in, is the correlation coefficient, is the covariance between the model assimilated forecast and the surface meteorological observation, and are the standard deviations of the model assimilated forecasts and surface meteorological observations, respectively.

[0072] For the error analysis of GHI (Global Horizontal Irradiance), in order to reduce the dependence of statistical indicators on GHI values, the mean absolute error, mean deviation and root mean square error are standardized. The specific calculation formula is as follows: The calculation formula for the standardized mean absolute error is:

[0073] in, is the standardized mean absolute error, is the mean absolute error before standardization, It is the arithmetic mean of the GHI values ​​observed on the ground.

[0074] The formula for calculating the standardized mean deviation is:

[0075] in, is the standardized mean deviation, is the mean deviation before standardization, It is the arithmetic mean of the GHI values ​​observed on the ground.

[0076] The calculation formula of the normalized root mean square error is:

[0077] in, is the normalized root mean square error, is the root mean square error before normalization, It is the arithmetic mean of the GHI values ​​observed on the ground.

[0078] When the indicators of error analysis (mean absolute error, mean deviation, root mean square error, and correlation coefficient) are not within the preset threshold range, the WRF-Solar model parameters and WRF-DA 4D-Var assimilation parameters are adjusted until the indicators of error analysis are within the preset threshold range.

[0079] S4. Construct the final prediction model using the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters.

[0080] S5. The final prediction model is used to predict the solar radiation and meteorological elements to be predicted, and the prediction results of the solar radiation and meteorological elements are obtained.

[0081] The final prediction model is used to predict the solar radiation and meteorological elements to be predicted (such as solar radiation and meteorological elements on different time scales in the future) to obtain the prediction results of solar radiation and meteorological elements (such as solar radiation and meteorological elements on different time scales in the future).

[0082] S6. Based on the prediction results of solar radiation and meteorological elements, obtain the optical power prediction results.

[0083] In this step, the pvlib-python tool is used to obtain the prediction results of the optical power (in this embodiment, the total AC output power of the photovoltaic system) based on the prediction results of solar radiation and meteorological elements. The specific process of obtaining the total AC output power of the photovoltaic system is described as follows: (1) Configure the PV system parameters (PV module parameters and inverter parameters) according to the actual situation of the PV system. You can extract the PV module parameters and inverter parameters from the database provided by pvlib-python, or you can manually set the PV module parameters and inverter parameters.

[0084] (2) Input the solar radiation results (such as GHI, DNI and DHI) predicted by the final prediction model into pvlib-python, and then use the pvlib.solarposition.get_solarposition() function of pvlib-python to calculate the azimuth and altitude of the sun based on the geographic location, time and date. Then use the pvlib.irradiance.get_total_irradiance() function to calculate the total radiation incident on the photovoltaic panel by combining GHI, DNI, DHI, the position of the sun and the tilt angle and azimuth of the photovoltaic panel. Among them, GHI is the global horizontal radiation, DNI is the direct normal radiation, and DHI is the diffuse horizontal radiation.

[0085] (3) The temperature model SAPM is used to calculate the operating temperature of the PV module. The surface ambient temperature and surface wind speed predicted by the WRF-Solar model are input into pvlib-python. The pvlib.temperature.TEMPERATURE_MODEL_PARAMETERS() function is used to call the temperature model SAPM and configure the relevant parameters of the temperature model SAPM. Finally, the pvlib.temperature.sapm_cell() function is used to calculate the temperature of the PV module.

[0086] (4) Use the pv_system.pvwatts_dc() function to calculate the DC output power of a single PV module in the PV system. Then, based on the number and arrangement (series or parallel) of PV modules in the PV system, calculate the total DC power of the PV system. Finally, use the pvlib.inverter.pvwatts_ac() function to calculate the total AC output power of the PV system.

[0087] Embodiment 4: See also Figure 3 As shown, the present invention also provides an electronic device 100 for an optical power prediction method based on WRF-Solar assimilation of multi-source data; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0088] The memory 101 can be used to store the computer program 103, and the processor 102 implements the steps of the optical power prediction method based on WRF-Solar assimilation multi-source data described in Example 1 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data (such as audio data) created according to the use of the electronic device 100, etc. In addition, the memory 101 can include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other non-volatile solid-state storage devices.

[0089] The at least one processor 102 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor, etc. The processor 102 is the control center of the electronic device 100, and uses various interfaces and lines to connect various parts of the entire electronic device 100.

[0090] The memory 101 in the electronic device 100 stores a plurality of instructions to implement an optical power prediction method based on WRF-Solar assimilation of multi-source data, and the processor 102 can execute the plurality of instructions to implement: Obtain historical meteorological data and historical radiation observation data, pre-process the obtained historical meteorological data and historical radiation observation data, remove outliers and correct deviations; Construct the WRF-Solar model and WRF-DA 4D-Var assimilation framework, input the preprocessed historical meteorological data and historical radiation observation data into the WRF-Solar model and WRF-DA 4D-Var assimilation framework, and obtain the prediction results using the historical meteorological data and historical radiation observation data; The error analysis of the prediction results using historical meteorological data and historical radiation observation data was performed to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters; The final prediction model was constructed using the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters; The final prediction model is used to predict the solar radiation and meteorological elements to be predicted, and the prediction results of the solar radiation and meteorological elements are obtained; Based on the prediction results of solar radiation and meteorological elements, the optical power prediction results are obtained.

[0091] Embodiment 5: If the module / unit integrated in the electronic device 100 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory and read-only memory (ROM, Read-Only Memory).

[0092] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0093] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0094] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0095] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for predicting optical power based on WRF-Solar assimilation of multi-source data, characterized in that: The following steps are involved: Obtain historical meteorological data and historical radiation observation data, pre-process the obtained historical meteorological data and historical radiation observation data, remove outliers and correct deviations; Construct the WRF-Solar model and WRF-DA 4D-Var assimilation framework, input the preprocessed historical meteorological data and historical radiation observation data into the WRF-Solar model and WRF-DA 4D-Var assimilation framework, and obtain the prediction results using the historical meteorological data and historical radiation observation data; The error analysis of the prediction results using historical meteorological data and historical radiation observation data was performed to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters; The final prediction model was constructed using the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters; The final prediction model is used to predict the solar radiation and meteorological elements to be predicted, and the prediction results of the solar radiation and meteorological elements are obtained; Based on the prediction results of solar radiation and meteorological elements, the optical power prediction results are obtained.

2. The optical power prediction method based on WRF-Solar assimilation of multi-source data according to claim 1 is characterized in that: In the steps of obtaining historical meteorological data and historical radiation observation data, preprocessing the obtained historical meteorological data and historical radiation observation data, removing outliers and correcting deviations, the obtained historical radiation observation data is also subjected to spatial and temporal interpolation processing to keep the temporal and spatial resolutions consistent with the WRF-Solar model.

3. The optical power prediction method based on WRF-Solar assimilation of multi-source data according to claim 1, characterized in that: In the step of performing error analysis on the prediction results of historical meteorological data and historical radiation observation data to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters, specific error analysis indicators include mean absolute error, mean deviation, root mean square error and correlation coefficient.

4. The optical power prediction method based on WRF-Solar assimilation of multi-source data according to claim 3 is characterized in that: In the step of performing error analysis on the prediction results of historical meteorological data and historical radiation observation data to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters, the specific error analysis process is: when the indicators of the error analysis are within a preset threshold range, the WRF-Solar model parameters and WRF-DA 4D-Var assimilation parameters with the smallest error values ​​are used as the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters.

5. The optical power prediction method based on WRF-Solar assimilation of multi-source data according to claim 4 is characterized in that: When the index of the error analysis is not within the preset threshold range, the WRF-Solar model parameters and the WRF-DA 4D-Var assimilation parameters are adjusted until the index of the error analysis is within the preset threshold range.

6. The optical power prediction method based on WRF-Solar assimilation of multi-source data according to claim 1, characterized in that: In the step of obtaining the optical power prediction result based on the prediction results of solar radiation and meteorological elements, the optical power prediction result is obtained based on the prediction results of solar radiation and meteorological elements using the pvlib-python tool.

7. The optical power prediction method based on WRF-Solar assimilation of multi-source data according to claim 1, characterized in that: The cost function expression of the WRF-DA 4D-Var assimilation framework is: in, is the cost function, For the initial moment The mode status, is the background field, i.e. the initial moment The pattern prior, is the background error covariance matrix, The number of time steps in the assimilation time window, For the model from the initial moment arrive The forward model operator at time , integrated to get Mode status at the moment , for The observation vector at time, is the observation operator, for The observation error covariance matrix at time , To set the initial mode state By integrating the model Mode status at the moment , and then through the observation operator Project it into the observation space and get the value of the model simulation in the observation space.

8. An optical power prediction system based on WRF-Solar assimilation of multi-source data, characterized in that: It includes data acquisition module, data preprocessing module, historical data prediction result acquisition module, optimal parameter acquisition module, final prediction model construction module, solar radiation and meteorological element prediction module and optical power prediction module; The data acquisition module is used to acquire historical meteorological data and historical radiation observation data; The data preprocessing module is used to preprocess the obtained historical meteorological data and historical radiation observation data, remove abnormal values ​​and correct deviations; The historical data prediction result acquisition module is used to construct the WRF-Solar model and the WRF-DA 4D-Var assimilation framework, input the pre-processed historical meteorological data and the historical radiation observation data into the WRF-Solar model and the WRF-DA 4D-Var assimilation framework, and obtain the prediction results using the historical meteorological data and the historical radiation observation data; The optimal parameter acquisition module is used to perform error analysis on the prediction results using historical meteorological data and historical radiation observation data to obtain the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters; The final prediction model construction module is used to construct the final prediction model using the optimal WRF-Solar model parameters and the optimal WRF-DA 4D-Var assimilation parameters; The solar radiation and meteorological element prediction module is used to predict the solar radiation and meteorological elements to be predicted using the final prediction model to obtain the prediction results of the solar radiation and meteorological elements; The optical power prediction module is used to obtain an optical power prediction result based on the prediction results of solar radiation and meteorological elements.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the optical power prediction method based on WRF-Solar assimilation of multi-source data are implemented as described in any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the optical power prediction method based on WRF-Solar assimilation of multi-source data are implemented as described in any one of claims 1 to 7.

Images