Photovoltaic generating capacity prediction method based on illumination simulation

Abstract

The invention provides a photovoltaic generating capacity prediction method based on illumination simulation, and the method comprises the steps: sequentially completing the data acquisition and standardization, terrain modeling, sun position and direct radiation calculation, point-by-point illumination simulation to obtain shielding and scattering, total radiation aggregation and coupling plate inclination angle correction, and finally predicting the generating capacity in combination with a conversion efficiency parameter. And high-precision photovoltaic output prediction under a complex terrain is realized. According to the invention, the precision and reliability of photovoltaic power generation prediction can be improved, and powerful technical support is provided for planning, operation and scheduling of a photovoltaic power station.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photovoltaic power generation, and more particularly, to a photovoltaic power generation amount prediction method based on light simulation. BACKGROUND

[0002] With the increasing global energy demand and growing emphasis on environmental protection, photovoltaic power generation as a clean and renewable energy form has received widespread attention and application. Photovoltaic power generation systems convert solar energy into electricity, providing an effective way to address energy shortages and environmental pollution. However, accurate prediction of photovoltaic power generation is of great significance for power system planning, scheduling and operation. The prediction of photovoltaic power generation needs to consider multiple factors, including meteorological conditions, geographical environment, solar radiation intensity and photovoltaic panel characteristics. Existing photovoltaic power generation prediction methods mainly rely on historical data and simple meteorological models. Although these methods can provide prediction results to some extent, they often have problems such as insufficient accuracy, inability to adapt to complex terrain and environmental changes. For example, some methods only consider the influence of direct solar radiation, ignoring the influence of terrain shading, sky scattered radiation and other factors on photovoltaic power generation. In addition, existing technologies have certain limitations in data processing and model construction, such as data missing, improper handling of outliers, and poor adaptability of models to complex terrain.

[0003] In the implementation of the embodiments of the present application, there are at least the following problems or defects in the prior art: the prior art cannot accurately simulate the light conditions under complex terrain, resulting in low accuracy of photovoltaic power generation prediction; the prior art lacks effective data filling and outlier removal methods in the data preprocessing stage, affecting the reliability of the prediction model; the prior art fails to fully consider the influence of photovoltaic panel installation inclination and azimuth on power generation, resulting in deviation between the prediction results and the actual situation. SUMMARY

[0004] The present application provides a photovoltaic power generation amount prediction method based on light simulation, comprising: S1, acquiring and processing data step: acquiring meteorological data sources and geographical data sources of a target area, and preprocessing the meteorological data sources and the geographical data sources to generate standardized meteorological data and geographical data; S2, constructing a terrain model step: based on the preprocessed geographical data, constructing a digital elevation model of the target area; S3, calculating the solar position and direct radiation step: based on the preprocessed meteorological data and geographical data, calculating the solar position information of each grid point in the target area within a predetermined time period, and further calculating the direct radiation of each grid point; S4, light simulation analysis step: based on the digital elevation model and the solar position information of each grid point, performing time-point-by-time-point light simulation analysis to calculate the terrain shading coefficient and sky scattered radiation amount of each grid point; S5, total radiation amount aggregation step: aggregating the direct radiation amount, the terrain shading coefficient and the sky scattered radiation amount of each grid point in the preset time period to calculate the total radiation amount of the photovoltaic panel receiving surface of the target region; S6, predicted power generation amount step: based on the total radiation amount and a preset photovoltaic panel conversion efficiency parameter, predicting the photovoltaic power generation amount of the target region in the preset time period.

[0005] Further, the step S1 specifically comprises the following sub-steps: S11, obtaining the meteorological data source, the meteorological data source comprising historical solar direct radiation intensity, historical solar scattered radiation intensity, historical ambient temperature, historical humidity and historical air pressure data from satellite remote sensing; obtaining the geographic data source, the geographic data source comprising longitude and latitude coordinates, elevation data, slope data and slope direction data of the target region from a digital elevation model database; S12, preprocessing the meteorological data source: firstly, using time series linear interpolation method to fill in the missing meteorological data; secondly, using threshold method to eliminate the abnormal meteorological data obviously beyond the reasonable range, and again using time series linear interpolation method to fill in; finally, resampling the processed meteorological data to the same spatial resolution as the geographic data; S13, preprocessing the geographic data source: converting the geographic data from different coordinate systems to the same geographic coordinate system, and resampling all geographic data to the same spatial resolution to generate the standardized geographic data.

[0006] Further, in the step S3, the specific method for calculating the solar position information of each grid point is: according to the longitude and latitude coordinates in the geographic data and the world standard time, the solar elevation angle and the solar azimuth angle of each grid point at each time point in the preset time period are calculated by a solar position algorithm.

[0007] Further, in the step S3, the specific method for calculating the direct radiation amount of each grid point is: based on the solar direct radiation intensity in the preprocessed meteorological data, and combined with the calculated solar elevation angle, the direct radiation amount of each grid point at each time point is calculated by formula (I):

[0008] wherein, represents the direct radiation amount, represents the direct solar radiation intensity in the meteorological data, represents the solar elevation angle.

[0009] Further, in the step S4, the specific method for calculating the terrain shading coefficient is: based on the digital elevation model and the solar position information of each grid point at each time point, determining whether there is terrain shielding between the current grid point and the sun by a line-of-sight analysis algorithm; the line-of-sight analysis algorithm takes the current grid point as the starting point, calculates the visibility of the terrain obstacles along the way along the solar azimuth angle direction and at the solar elevation angle as the elevation angle; if there is shielding, the terrain shading coefficient is 0; if there is no shielding, the terrain shading coefficient is 1.

[0010] Further, in the step S4, the specific method for calculating the sky scattered radiation amount is: based on the preprocessed solar scattered radiation intensity in the meteorological data, the digital elevation model and the terrain shading coefficient, the sky scattered radiation amount is calculated by anisotropic Hay model; the Hay model first calculates the anisotropy index of sky scattered radiation, and then combines the sky view factor provided by the digital elevation model to finally calculate the sky scattered radiation amount of each grid point.

[0011] Further, in the step S5, the specific method for calculating the total radiation amount is: superimposing the direct radiation amount corrected by the terrain shading coefficient and the sky scattered radiation amount, and calculating the instantaneous total radiation amount of each grid point at each time point by formula (II), and then integrating and summing the instantaneous total radiation amount of all time points in the preset time period to obtain the cumulative total radiation amount in the preset time period:

[0012] wherein, represents the cumulative total radiation amount, t represents the time point, start represents the starting time of the preset time period, and end represents the end time of the preset time period, represents the direct radiation amount at the time point t, represents the terrain shading coefficient at the time point t, represents the sky scattered radiation amount at the time point t, represents the time step.

[0013] Further, in the step S5, before calculating the cumulative total radiation amount, it further includes a correction step of photovoltaic panel installation inclination and azimuth: according to the photovoltaic panel installation parameters of the target area, the instantaneous total radiation amount is converted from the horizontal plane coordinate system to the photovoltaic panel inclined plane coordinate system by using a rotation matrix to correct the cosine projection loss.

[0014] Further, in the step S6, the photovoltaic panel conversion efficiency parameters include photovoltaic panel nominal efficiency, temperature attenuation coefficient, dust loss coefficient and inverter efficiency; the specific method for predicting the photovoltaic power generation is: inputting the cumulative total radiation into a photovoltaic power generation calculation model, combining the photovoltaic panel conversion efficiency parameters, and calculating the photovoltaic power generation through formula (III):

[0015] Wherein, E represents the predicted photovoltaic power generation, represents the cumulative total radiation, and A represents the photovoltaic panel area, represents the photovoltaic panel nominal efficiency, represents the temperature attenuation coefficient, represents the average ambient temperature in the preset time period, represents the standard test temperature, represents the dust loss coefficient, represents the inverter efficiency.

[0016] Further, in the step S5, before integrating and summing the instantaneous total radiation, a data verification and compensation step is further included: S51, traversing all grid points, verifying the proportion of valid data points of each grid point in the preset time period; the valid data point refers to a data point whose calculated instantaneous total radiation is non-negative and less than the solar constant; S52, if the proportion of valid data points of a certain grid point is lower than a first preset threshold, it is determined that the grid point is an invalid grid point; S53, for the grid point determined to be invalid, the arithmetic mean of the instantaneous total radiation of its surrounding valid grid points at the same time is used for replacement; S54, if there is no valid grid point in the surrounding, the historical average radiation of the same period in the geographical area where the grid point is located is used for replacement.

[0017] The above embodiments of the present application have at least the following beneficial effects: 1. By constructing a digital elevation model and combining solar position information to perform light simulation analysis point by point in time, the terrain shading coefficient and sky scattered radiation of each grid point can be accurately calculated, thereby effectively solving the problem of insufficient photovoltaic power generation prediction accuracy caused by ignoring terrain shading and scattered radiation in the prior art, improving the accuracy of the prediction result, and being especially suitable for photovoltaic power generation prediction in complex terrain areas.

[0018] 2. In the data preprocessing stage, the time series linear interpolation method is used to fill in the missing data, and the threshold method is used to eliminate abnormal data, at the same time, different sources of geographic data are unified to the same coordinate system and resampled to the same spatial resolution, effectively solving the problems of data missing, improper handling of abnormal values and inconsistency of geographic data in the prior art, improving the reliability and usability of the data, and providing a solid data foundation for subsequent light simulation and power generation prediction.

[0019] 3. In the calculation of total radiation, the correction steps of photovoltaic panel installation inclination and azimuth are introduced, the instantaneous total radiation is converted from the horizontal plane coordinate system to the photovoltaic panel inclined plane coordinate system by using the rotation matrix to correct the cosine projection loss, and at the same time, the parameters such as nominal efficiency of photovoltaic panel, temperature attenuation coefficient, dust loss coefficient and inverter efficiency are comprehensively considered in the prediction of power generation, fully solving the problem that the installation parameters of photovoltaic panel and various loss factors cannot be fully considered in the prior art to affect the power generation, making the prediction result more close to the actual power generation, and improving the practicability and reliability of photovoltaic power generation prediction. BRIEF DESCRIPTION OF DRAWINGS

[0020] The above and other objects, features and advantages of the exemplary embodiments of the present application will be more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which: Figure 1 The flowchart of the photovoltaic power generation prediction method based on light simulation provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0021] The technical solutions in the present application will be described clearly and completely below in conjunction with the drawings in the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. The components of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0022] In the traditional existing photovoltaic power generation prediction method, there is systematic deviation in the simulation of light conditions in complex terrain areas, and the coupling effect of terrain shielding effect and sky scattered radiation is not accurately quantified. The data preprocessing link lacks standardized procedures for abnormal value processing and spatial resolution unification of multi-source heterogeneous meteorological data, which affects the reliability of subsequent calculation links. The spatial aggregation process of photovoltaic panel receiving surface radiation does not consider the cosine projection loss caused by inclined installation, resulting in distortion of energy conversion model input parameters.

[0023] For example, there is a continuous undulating terrain in the deployment area of a certain mountain photovoltaic power station, with an elevation change range of more than 800 meters, and the slope distribution presents a multi-peak characteristic. The solar direct radiation intensity data provided by meteorological satellites has time series missing, and the original geographic coordinate system contains two standards of WGS84 and CGCS2000, with grid resolution fluctuating between 10 meters and 30 meters. When using traditional methods for calculation, the line-of-sight analysis algorithm does not establish a dynamic correlation between three-dimensional terrain and solar trajectory, resulting in a shielding judgment error of more than 40% on the sunny slope and the shady valley. The isotropic model is used for scattered radiation calculation, without correction combined with the actual sky view factor, resulting in a 12%-18% deviation of the estimated radiation value from the measured value during dawn and dusk. The radiation aggregation under the horizontal plane coordinate system does not perform inclination correction, resulting in a 6.7 GWh prediction error of the annual power generation of a 200 MW photovoltaic array.

[0024] If the above problems are not solved, the terrain shielding error will cause the power generation potential of the local block of the photovoltaic array to be overestimated or underestimated, affecting the accuracy of the power grid dispatching strategy. Data preprocessing defects may cause non-physical mutations in the radiation calculation results, reducing the robustness of the prediction model under extreme weather conditions. Coordinate system conversion error and inclination correction loss will cause inaccurate system-level energy balance calculation, directly affecting the reliability of investment return rate calculation of the power station, and causing deviation assessment risk in the power market transaction.

[0025] In the face of the above problems, the application aims at the data preprocessing link, and finds that the standardization processing of multi-source heterogeneous data needs to establish a unified coordinate system conversion and spatial resolution matching process. For the calculation error of the radiation of the photovoltaic panel receiving surface, it is realized that the horizontal plane radiation must be converted to the actual installation inclined plane coordinate system to correct the projection loss. The traditional method does not dynamically associate the terrain shielding judgment with the solar position, and the application proposes to perform visibility analysis through digital elevation model combined with time point by time point solar position information. At the same time, in order to solve the problem of matching the scattering radiation model with the actual terrain, an anisotropic radiation model combined with the sky view factor is used for correction. In the data preprocessing stage, the standardization process of meteorological data filling and geographical data resampling is designed to ensure the spatio-temporal consistency of multi-source data. Finally, the technical route of constructing a terrain model, dynamically calculating the solar position, calculating the radiation components in layers and performing coordinate conversion is determined to form a complete radiation calculation link.

[0026] As shown in Figure 1 The application proposes a photovoltaic power generation prediction method based on light simulation, which includes the following steps: S1, data acquisition and processing step: acquiring meteorological data sources and geographical data sources of the target area, and preprocessing the meteorological data sources and geographical data sources to generate standardized meteorological data and geographical data; S2, terrain model construction step: based on the preprocessed geographical data, constructing a digital elevation model of the target area; S3, solar position and direct radiation calculation step: based on the preprocessed meteorological data and geographical data, calculating the solar position information of each grid point in the target area within a preset time period, and further calculating the direct radiation of each grid point; S4, light simulation analysis step: based on the digital elevation model and the solar position information of each grid point, performing time point by time point light simulation analysis, and calculating the terrain shielding coefficient and sky scattering radiation of each grid point; S5, total radiation aggregation step: aggregating the direct radiation, terrain shielding coefficient and sky scattering radiation of each grid point within a preset time period, and calculating the total radiation of the photovoltaic panel receiving surface of the target area; S6, power generation prediction step: based on the total radiation and the preset photovoltaic panel conversion efficiency parameter, predicting the photovoltaic power generation of the target area within a preset time period.

[0027] The preprocessing data step refers to the standardization processing of meteorological data sources and geographic data sources. Specifically, missing data can be filled by time series linear interpolation method, abnormal values can be removed by threshold method, coordinate system conversion and spatial resolution resampling technology can be used to realize, to solve the problem of inconsistent original data or noise. Among them, the digital elevation model refers to the construction of a three-dimensional terrain model of the target area through geographic data, which can be generated by rasterizing elevation data combined with geographic information system software, and is used to accurately simulate the influence of terrain on light. Among them, the solar position information refers to the calculation of the solar elevation angle and azimuth angle of each grid point in the target area within a predetermined time period, which can be calculated by using the solar position algorithm based on latitude and longitude coordinates and world standard time, and can provide spatial geometric parameters for radiation calculation. Among them, the terrain shading coefficient is a binary parameter that determines whether the grid point is shaded by the terrain. Specifically, the visibility analysis algorithm can be used to calculate the visibility along the solar azimuth direction to eliminate the radiation error caused by terrain shadow. Among them, the sky scattered radiation is the non-direct radiation energy generated by atmospheric scattering. Specifically, the anisotropic Hay model combined with the sky view factor can be used to calculate and quantify the scattered radiation distribution under complex terrain. The total radiation aggregation step refers to the time and space accumulation of direct radiation, terrain shading coefficient and scattered radiation. Specifically, time integration algorithm can be used to superimpose instantaneous radiation data to accurately reflect the energy input of the photovoltaic panel receiving surface. The photovoltaic panel conversion efficiency parameter is a key physical coefficient for converting radiation to power generation. Specifically, a multi-dimensional calculation model can be constructed by using temperature attenuation coefficient, dust loss coefficient and inverter efficiency to improve the engineering practicability of power generation prediction.

[0028] The present application includes the terrain shading effect and the sky scattered radiation into the photovoltaic power generation prediction model. Through the cooperative application of digital elevation model construction, line of sight analysis algorithm and anisotropic radiation model, the spatial resolution difference between geographic and meteorological data is eliminated in the data preprocessing stage, and the direct and scattered radiation under complex terrain is accurately simulated in the radiation calculation stage. Finally, high-precision power generation prediction results are output through space-time aggregation and efficiency parameter correction.

[0029] Further, in the data acquisition and processing step, the meteorological data sources include historical solar direct radiation intensity, historical solar scattered radiation intensity, historical environmental temperature, historical humidity and historical air pressure data obtained by satellite remote sensing. The geographic data sources include the latitude and longitude coordinates, elevation data, slope data and aspect data of the target area in the digital elevation model database. The time series linear interpolation is used to fill the missing values of the meteorological data, and the threshold method is used to remove the abnormal values. The geographic data is converted to the same geographic coordinate system and resampled to the same spatial resolution.

[0030] When constructing the terrain model, the preprocessed geographic data is used to generate a three-dimensional digital elevation model with a resolution of 10 meters.

[0031] When calculating the position of the sun, a solar position algorithm is used to calculate the solar elevation angle and azimuth angle of each grid point in a preset time period with a time step of 5 minutes. The direct radiation amount calculation considers the influence of the solar elevation angle.

[0032] The light simulation analysis adopts a line-of-sight analysis algorithm, takes the current grid point as the starting point, and judges whether there is terrain shielding along the solar azimuth angle direction and at the solar elevation angle. The sky scattered radiation amount calculation adopts the anisotropic Hay model and is modified in combination with the sky view factor.

[0033] When aggregating the total radiation amount, the direct radiation amount and the sky scattered radiation amount are superimposed, and the influence of the terrain shielding coefficient is considered. The instantaneous total radiation amount of all time points in the preset time period is integrated and summed to obtain the cumulative total radiation amount.

[0034] When predicting the power generation amount, the cumulative total radiation amount is input into a photovoltaic power generation amount calculation model considering parameters such as the nominal efficiency of the photovoltaic panel, the temperature attenuation coefficient, the dust loss coefficient, and the inverter efficiency, to obtain the final prediction result.

[0035] The application further proposes a specific method for obtaining and processing data steps: obtaining a meteorological data source, the meteorological data source including historical solar direct radiation intensity, historical solar scattered radiation intensity, historical environmental temperature, historical humidity, and historical air pressure data from satellite remote sensing; obtaining a geographic data source, the geographic data source including longitude and latitude coordinates, elevation data, slope data, and slope direction data of the target region from a digital elevation model database; preprocessing the meteorological data source: first, filling in missing meteorological data using a time series linear interpolation method; second, removing abnormal meteorological data that obviously exceeds a reasonable range using a threshold method, and filling in the data again using a time series linear interpolation method; and finally, resampling the processed meteorological data to the same spatial resolution as the geographic data; preprocessing the geographic data source: converting geographic data from different coordinate systems to the same geographic coordinate system, and resampling all geographic data to the same spatial resolution to generate standardized geographic data.

[0036] Among them, the preprocessing of the meteorological data source fills in missing data through time series linear interpolation, ensures time continuity, removes outliers again after removing outliers using a threshold method, and eliminates unreasonable data interference; the meteorological data is resampled to the resolution of the geographic data to ensure spatial matching. The preprocessing of the geographic data source eliminates spatial misplacement problems through coordinate system unification; resampling to the same resolution ensures consistency in subsequent grid calculations.

[0037] Specifically, in the meteorological data imputation process, time-series linear interpolation fills in missing values ​​based on the data trends of adjacent time points. For example, if temperature data is missing at a certain moment, linear interpolation is performed using data from the two hours before and after the missing value. Outlier removal uses a preset threshold range; for example, data with direct solar radiation intensity exceeding the extra-atmospheric solar constant is considered outlier, directly removed, and then re-interpolated. Meteorological data resampling uses spatial interpolation algorithms to adjust the data resolution to be consistent with the geographic data. For example, bilinear interpolation is used to adjust meteorological data from a 1-kilometer resolution to a 30-meter resolution. Geographic data coordinate transformation uses a seven-parameter Bursa model to achieve accurate conversion between different coordinate systems, such as converting the WGS84 coordinate system to the UTM projected coordinate system. During resampling, the nearest neighbor method is used to maintain the original attributes of the geographic data; for example, elevation data retains its integer properties after resolution adjustment. Thus, the preprocessed meteorological and geographic data meet the standardization requirements in time, space, and attribute dimensions, providing reliable input for subsequent terrain model construction and radiation calculation.

[0038] Furthermore, meteorological data sources are acquired, including historical direct solar radiation intensity, historical diffuse solar radiation intensity, historical ambient temperature, historical humidity, and historical air pressure data from satellite remote sensing; geographic data sources are also acquired, including latitude and longitude coordinates, elevation data, slope data, and aspect data of the target area from a digital elevation model database. Preprocessing of meteorological data sources: First, missing meteorological data are filled in using time series linear interpolation; second, abnormal meteorological data that clearly exceeds the reasonable range are removed using a threshold method and filled in again using time series linear interpolation; finally, the processed meteorological data are resampled to the same spatial resolution as the geographic data. Preprocessing of geographic data sources: Geographic data from different coordinate systems are uniformly converted to the same geographic coordinate system, and all geographic data are resampled to the same spatial resolution to generate standardized geographic data.

[0039] Specifically, when acquiring meteorological data sources, historical meteorological data for the target area over the past year can be downloaded by accessing the API interface of the satellite remote sensing data platform. Geographic data sources can be obtained from the National Geographic Information Resource Catalog Service System.

[0040] When preprocessing meteorological data, the `interpolate()` function from the pandas library is first used to perform linear interpolation on missing data. Then, reasonable thresholds are set based on the physical characteristics of each meteorological parameter; for example, the temperature range is limited to -50°C to 50°C. Data outside this range is considered outliers and removed. For the removed outliers, linear interpolation is used again to impute them. Finally, the `interpolate.griddata()` function from the scipy library is used to resample the meteorological data to the same 1km × 1km spatial resolution as the geographic data.

[0041] For the preprocessing of geographic data, the GDAL library is first used to convert data from different coordinate systems to the WGS84 coordinate system. Then, the interpolate.griddata() function of the scipy library is used to resample all geographic data to a spatial resolution of 1km×1km, generating a standardized geographic dataset.

[0042] This application further proposes a method based on latitude and longitude coordinates and world standard time in geographic data, using a solar position algorithm to cyclically calculate the solar altitude angle and solar azimuth angle of each grid point at each time point within a preset time period.

[0043] The solar position algorithm employs an astronomical calendar model, establishing a solar declination angle calculation model based on Earth's orbital parameters and axial tilt. Latitude and longitude coordinates are used as input parameters to determine the geographical location of the target area, and Coordinated Universal Time (UTC) is used for time synchronization to eliminate time zone errors. The iterative calculation process uses a preset time step as the iteration interval, continuously performing solar position calculations in the time dimension. The time step is set to 15 minutes to achieve a balance between time resolution and computational efficiency. The independent calculation of each grid point adopts a parallel processing architecture, achieving spatial data partitioning through distributed computing nodes.

[0044] Specifically, the solar position algorithm first calculates the Julian day parameters based on Coordinated Universal Time (UTC), and then calculates the solar mean anomaly angle using the Earth's orbital eccentricity. Further, it solves for the solar declination angle using the obliquity of the ecliptic model, and calculates the hour angle parameters based on latitude and longitude coordinates. The solar altitude angle is calculated by coupling the declination angle, hour angle, and geographic latitude using spherical trigonometry formulas, while the solar azimuth angle is corrected using an azimuth angle correction formula to eliminate projection errors. During the iterative calculation process, an independent calculation thread is generated for each time point to ensure the continuity of the time series data. The spatial independence between grid points isolates the influence of terrain differences on the solar position, preventing error propagation. Through a dual iterative mechanism of time and space, dynamic modeling of the solar position of all grid points within a preset time period is achieved, providing accurate geometric parameter input for subsequent radiance calculations.

[0045] Furthermore, when calculating the sun's position information, the specific location of the target area is first determined based on the latitude and longitude coordinates in the geographic data. For example, for a photovoltaic power station located at 30°N, 120°E, its latitude and longitude coordinates are (30°N, 120°E). Next, a time series is established based on Coordinated Universal Time (UTC), such as starting from 00:00:00 UTC on January 1, 2023, with 1-hour intervals, to calculate the sun's position over a 24-hour period.

[0046] Specifically, the NREL solar position algorithm is used for calculation. This algorithm takes into account factors such as the Earth's rotation, revolution, and axial tilt, and can accurately calculate the Sun's position on the celestial sphere. For each point in time, the algorithm inputs date, time, longitude, and latitude, and outputs the solar altitude angle and azimuth angle. For example, at 12:00:00 UTC on January 1, 2023, for the above position, the calculated solar altitude angle is approximately 36.5°, and the azimuth angle is approximately 180.3° (180° for due south).

[0047] Furthermore, the calculation process is embedded in a loop structure, repeating the process for each time point and each grid point within a preset time period. Assuming the target area is divided into a 100×100 grid, 240,000 (24 hours × 100 × 100) solar position calculations are required. The calculation results are stored in array form, with each element containing time, grid coordinates, solar altitude angle, and azimuth information.

[0048] Therefore, the systematic calculation of the sun's position provides basic data support for subsequent calculations of direct radiation, topographic shading, and diffuse radiation.

[0049] This application further proposes a method to calculate the direct radiation intensity of each grid point at each time point using formula (I), based on the direct solar radiation intensity in the preprocessed meteorological data and combined with the calculated solar altitude angle:

[0050] in, This indicates the amount of direct radiation. This represents the intensity of direct solar radiation in meteorological data. This indicates the solar altitude angle.

[0051] The solar direct radiation intensity is derived from meteorological data processed using time-series linear interpolation and thresholding to ensure data integrity and rationality. The solar altitude angle is calculated iteratively for each time point within a preset time period using latitude and longitude coordinates and Coordinated Universal Time (UTC) from geographic data, employing a solar position algorithm. Formula (I) simulates the vertical component of solar radiation on the horizontal plane by multiplying the solar direct radiation intensity by the sine of the solar altitude angle, thus reflecting the effective radiation intensity actually reaching the Earth's surface.

[0052] Specifically, the reliability of direct solar radiation intensity data has been improved during the preprocessing stage by filling in and removing outliers, while the solar altitude angle is calculated iteratively to ensure continuity over time. During the calculation, the sine of the solar altitude angle is used to correct for the incident angle of solar radiation, avoiding overestimation or underestimation of radiation due to neglecting changes in the sun's position. For example, when the solar altitude angle approaches zero, the sine value also approaches zero, and the calculated direct radiation automatically approaches zero, consistent with the actual physical phenomena during sunrise and sunset. This formula allows for the dynamic combination of meteorological and geographic data, accurately quantifying the effective radiation of each grid point at different times, providing accurate input for subsequent terrain shading correction and total radiation aggregation.

[0053] Furthermore, when calculating the direct radiation at each grid point, the solar direct radiation intensity is first obtained from the preprocessed meteorological data. The solar direct radiation intensity is typically expressed in W / m². 2 The units are [units]. Further, combining the previously calculated solar altitude angle, these two parameters are substituted into formula (I) for calculation. Specifically, formula (I) is: In this formula, This indicates the amount of direct radiation. This represents the intensity of direct solar radiation in meteorological data. This represents the solar altitude angle. From this, the direct radiation at each grid point at each time point can be obtained.

[0054] For example, suppose that at a certain time point, the intensity of direct solar radiation at a certain grid point is 800 W / m. 2 The solar altitude angle is 30°. Substituting these values ​​into formula (I), we get: =800·sin(30°)≈400W / m 2 This means that the direct radiation received by this grid point at this time is approximately 400 W / m. 2 .

[0055] This application further proposes a method based on a digital elevation model and the solar position information of each grid point at each time point. A line-of-sight analysis algorithm is used to determine whether there is terrain occlusion between the current grid point and the sun. The line-of-sight analysis algorithm takes the current grid point as the starting point, along the solar azimuth angle and the solar altitude angle as the elevation angle, and calculates the visibility with terrain obstacles along the way. If there is occlusion, the terrain occlusion coefficient is 0; if there is no occlusion, the terrain occlusion coefficient is 1.

[0056] The line-of-sight analysis algorithm obtains three-dimensional terrain data through a digital elevation model, determines the line-of-sight direction by combining the solar azimuth angle and the solar altitude angle to determine the line-of-sight elevation angle, thus forming a line-of-sight path in three-dimensional space. The visibility calculation uses the ray tracing method to detect elevation data point by point along the line-of-sight path. If the terrain elevation at any point on the path exceeds the line-of-sight elevation, it is determined that there is an obstruction. The terrain occlusion coefficient is processed using binary processing to eliminate the uncertainty of the occlusion state.

[0057] Specifically, within the 3D terrain mesh provided by the digital elevation model (DEM), the current grid point coordinates are used as the starting point. The horizontal angle is determined based on the solar azimuth angle, and the vertical elevation angle is determined based on the solar altitude angle, generating a spatial ray path. Sampling is performed along the ray path at a preset step size, converting the geographic coordinates of each sampling point into a grid index in the DEM, and extracting the elevation value of the corresponding location. Simultaneously, based on the geometric relationship of the ray path, the theoretical line-of-sight elevation of each sampling point is calculated. When the actual terrain elevation is detected to exceed the line-of-sight elevation, the calculation is immediately terminated and the system returns to an occlusion state. By using a binary terrain occlusion coefficient, the complex calculations for partial occlusion or penumbra areas are eliminated, ensuring the clarity of the direct radiation correction logic. For example, in a ridge area with a 30-degree slope, when the solar azimuth angle is 120 degrees, the line-of-sight path extends southeast. If the elevation of the mountain ahead exceeds the line-of-sight path elevation by 30%, the grid point is determined to be in the shadow area, and the direct radiation is set to zero.

[0058] Furthermore, when calculating the terrain shading coefficient, the algorithm first uses a line-of-sight analysis to determine whether there is terrain shading between the current grid point and the sun, based on the digital elevation model and the solar position information of each grid point at each time point. The line-of-sight analysis algorithm takes the current grid point as the starting point, along the solar azimuth direction and the solar altitude angle as the elevation angle, to calculate the visibility with terrain obstacles along the way.

[0059] Specifically, the line-of-sight analysis algorithm employs ray tracing. Starting from the current grid point, it samples the digital elevation model at fixed step sizes, following the directions of the solar azimuth and elevation angles. The elevation of each sampled point is compared with the theoretical ray height. If the elevation of the sampled point is greater than the theoretical ray height, occlusion is determined to exist.

[0060] Furthermore, to improve computational efficiency, an adaptive step size strategy can be adopted. A smaller sampling step size is used in areas closer to the current grid point, while a larger sampling step size is used in more distant areas. For example, a 10-meter step size can be used in the 0-1000 meter range, a 50-meter step size in the 1000-5000 meter range, and a 100-meter step size beyond 5000 meters.

[0061] Therefore, if line-of-sight analysis indicates the presence of obstruction, the terrain obstruction coefficient is set to 0; if no obstruction exists, the terrain obstruction coefficient is set to 1. This binarized terrain obstruction coefficient can be directly used for subsequent radiation calculations.

[0062] This application further proposes to calculate the sky scattering radiation based on the solar scattering radiation intensity, digital elevation model, and terrain shading coefficient in the preprocessed meteorological data using an anisotropic Hay model. The Hay model first calculates the anisotropy index of the sky scattering radiation, and then combines it with the sky view factor provided by the digital elevation model to finally calculate the sky scattering radiation of each grid point.

[0063] The anisotropy index is calculated by establishing a radiation distribution weighting function using solar position and atmospheric transmittance parameters. The sky view factor is obtained through three-dimensional spatial geometric analysis using a digital elevation model to determine the proportion of visible sky within a hemisphere. The terrain shading coefficient, as a binarization parameter, participates in the shading correction of sky scattered radiation, automatically blocking the scattered radiation component in that direction when a grid point is obscured by terrain. The Hay model achieves coupled calculation of terrain features and atmospheric conditions by establishing the product relationship between the anisotropy index and the sky view factor.

[0064] Specifically, the method first obtains a baseline value of solar diffuse radiation intensity from standardized meteorological data. Then, it calculates the anisotropy index based on solar altitude angle and atmospheric transmittance parameters. This index quantifies the directional distribution characteristics of diffuse radiation. Simultaneously, a three-dimensional line-of-sight analysis is performed based on a digital elevation model to calculate the sky view factor for each grid point. This factor represents the proportion of sky area not obstructed by terrain at that point. The anisotropy index is multiplied by the sky view factor, and then multiplied by the baseline diffuse radiation intensity to obtain the terrain-corrected sky diffuse radiation. This method effectively solves the problem of inaccurate calculation of diffuse radiation in complex terrain areas by quantifying the impact of terrain shading on diffuse radiation, making the photovoltaic power generation prediction results more consistent with the radiation distribution patterns under actual terrain conditions.

[0065] Furthermore, in calculating sky diffuse radiance, the calculation is first performed using an anisotropic Hay model based on the solar diffuse radiance intensity, digital elevation model, and terrain shading coefficient from the preprocessed meteorological data. Specifically, the Hay model first calculates the anisotropy index of sky diffuse radiance, and then combines it with the sky view factor provided by the digital elevation model to finally calculate the sky diffuse radiance for each grid point.

[0066] In practical applications, the anisotropy index can be calculated using the following formula: AI = 1 - (Idiff / I0) Where AI is the anisotropy index, Idiff is the intensity of scattered radiation on the horizontal plane, and I0 is the intensity of solar radiation outside the atmosphere.

[0067] Furthermore, the sky view factor can be calculated by analyzing the terrain height within a 360-degree range around each grid point in the digital elevation model. For example, the surrounding space can be divided into 72 sectors of 5 degrees, the maximum elevation angle within each sector can be calculated, and then the sky view factor can be calculated based on these elevation angles.

[0068] Finally, the amount of scattered radiation from the sky can be calculated using the following formula:

[0069] in, This represents the amount of scattered radiation on the inclined plane. The angle of inclination of the inclined plane. The solar altitude angle, 1 is the zenith angle, and SVF is the sky view factor.

[0070] This application further proposes to perform a data verification and compensation step before calculating the cumulative total radiation, including traversing all grid points to verify the proportion of valid data points, and after determining invalid grid points, using the surrounding valid grid points or the average value of historical radiation for the same period for data compensation.

[0071] Valid data points are defined as those with non-negative instantaneous total irradiance that is less than the solar constant. Invalid grid points with insufficient valid data points are filtered out by setting a first preset threshold. For invalid grid points, the arithmetic mean of the instantaneous total irradiance of surrounding valid grid points at the same time is used as a substitute. If there is no valid data in the surrounding area, the average historical irradiance of the same period in that area is used to fill the gap.

[0072] Specifically, the data verification and compensation steps ensure the validity of each data point by verifying it grid by grid, avoiding the impact of outliers or missing values ​​on the integration results. When the proportion of valid data points for a certain grid is lower than a preset threshold, the system automatically triggers a compensation mechanism, using spatially adjacent or temporally similar valid data for substitution. For example, if a grid's radiation exceeds the solar constant at consecutive time points due to sensor failure, the system will identify it as an invalid grid and use the average data from adjacent grids for interpolation. If there is no valid adjacent data for that area, historical data from the same period will be used to complete the compensation, thereby ensuring the completeness and accuracy of the cumulative total radiation calculation. Thus, the photovoltaic power generation prediction model can output more accurate power generation results based on more reliable total radiation data.

[0073] Furthermore, when calculating the total radiance, the direct radiance, corrected for terrain shading, is first superimposed with the sky-scattered radiance. Specifically, for each grid point within the target area, the instantaneous total radiance is calculated at each time point. The calculation formula is:

[0074] in, This represents the instantaneous total radiation. Indicates a point in time. Indicates a point in time The amount of direct radiation. Indicates a point in time Terrain shading coefficient, Indicates a point in time The amount of scattered radiation from the sky, Indicates the time step.

[0075] For example, suppose the preset time period is 24 hours a day, with a time step of 1 hour. For a grid point within the target area, first calculate the instantaneous total radiation for each hour, then add the instantaneous total radiation for 24 hours to obtain the cumulative total radiation for that grid point in one day. Repeating this process, the cumulative total radiation for all grid points within the target area can be obtained.

[0076] This application further proposes to add a correction step for the installation tilt angle and azimuth angle of the photovoltaic panel before calculating the cumulative total radiation: based on the installation parameters of the photovoltaic panel in the target area, the instantaneous total radiation is transformed from the horizontal coordinate system to the photovoltaic panel tilt plane coordinate system using a rotation matrix to correct the cosine projection loss.

[0077] The rotation matrix is ​​constructed based on the photovoltaic panel's installation tilt angle and azimuth angle parameters, achieving vector projection transformation of radiation through three-dimensional spatial coordinate transformation. The photovoltaic panel installation parameters include fixed installation angles recorded in the mechanical structure design documents or dynamic angle data fed back in real time by the tracking system. The horizontal coordinate system is established with the ground plane as the reference plane, and the tilted coordinate system is established with the normal direction of the photovoltaic panel surface as the reference axis. During the coordinate transformation process, the solar incident vector is decomposed to the photovoltaic panel normal direction using a direction cosine matrix, eliminating cosine projection errors caused by tilted installation.

[0078] Specifically, the tilt angle and azimuth angle parameters of the photovoltaic panel are quantified as input variables of a rotation matrix, constructing a transformation matrix that rotates around the geographic coordinate system axes. Instantaneous total radiation data is input as a three-dimensional vector into the rotation matrix, and after coordinate transformation, the radiation components in the tilted plane coordinate system are output. This process converts the radiation received on the horizontal plane into the effective radiation actually received on the tilted plane, correcting for the cosine loss caused by the deviation between the photovoltaic panel surface and the angle of incidence of sunlight. For example, when the photovoltaic panel tilt angle is 30 degrees and the azimuth angle is due south, the rotation matrix will automatically adjust the projected components of the radiation vector, making the calculation results closer to the radiation reception efficiency under actual installation conditions. Through this correction step, the differences in radiation reception of the photovoltaic panel at different installation angles are accurately quantified, significantly improving the spatial adaptability and computational accuracy of the power generation prediction model.

[0079] As a preferred embodiment, the specific implementation of this application is as follows: When calculating the cumulative total radiation, the tilt angle and azimuth angle of the photovoltaic panel installation need to be corrected. The photovoltaic panel installation parameters include a tilt angle of 35° and an azimuth angle of 10° west of south. The instantaneous total radiation in the horizontal coordinate system is transformed to the tilted coordinate system by constructing a rotation matrix. Specifically, a local coordinate system is established with the normal direction of the photovoltaic panel as the reference axis. The rotation angles around the X and Z axes of the geographic coordinate system are calculated based on the tilt angle and azimuth angle to generate a three-dimensional rotation matrix. The horizontal radiation vector is input into the rotation matrix for spatial transformation to eliminate the cosine projection loss caused by the tilt of the photovoltaic panel, and the corrected tilted radiation vector is output. Furthermore, the instantaneous total radiation of each grid point is transformed point by point to ensure that the radiation calculation is consistent with the actual spatial orientation of the photovoltaic panel's light-receiving surface.

[0080] This application further proposes a correction step to add the installation tilt angle and azimuth angle of the photovoltaic panel before calculating the cumulative total radiation. Specifically, it includes: based on the installation parameters of the photovoltaic panel in the target area, using a rotation matrix to transform the instantaneous total radiation from the horizontal coordinate system to the photovoltaic panel tilt coordinate system to correct the cosine projection loss.

[0081] The photovoltaic (PV) panel installation parameters include the installation tilt angle and azimuth angle. The installation tilt angle refers to the angle between the PV panel and the horizontal plane, and the azimuth angle refers to the angle of deflection of the PV panel from due south. The rotation matrix uses a three-dimensional spatial coordinate system transformation method to rotate the radiation vector in the horizontal plane coordinate system around the geographic coordinate system axis, making the calculation result perpendicular to the actual tilt surface of the PV panel. During the correction process, the instantaneous total radiation data is converted into radiation intensity in the tilt surface coordinate system, eliminating the cosine projection loss caused by the PV panel installation angle. For example, when the PV panel installation tilt angle is 30 degrees and the azimuth angle is 15 degrees east of south, the rotation matrix rotates the radiation vector in the horizontal plane coordinate system by 30 degrees around the north-south axis and then by 15 degrees around the vertical axis to obtain the radiation in the tilt surface coordinate system.

[0082] Specifically, photovoltaic (PV) panel installation parameters are obtained through on-site measurements or design drawings and input into the calculation model. The instantaneous total radiance data in the horizontal coordinate system contains three-dimensional vector components, corresponding to the east, north, and vertical directions, respectively. The rotation matrix is ​​constructed based on the installation tilt angle and azimuth angle, and the original radiance vector is transformed to the tilted plane coordinate system through matrix multiplication. The component of the transformed radiance vector in the tilted plane normal direction is the corrected effective radiance. For example, when there is an angle between the tilted plane of the PV panel and the incident direction of sunlight, the radiance in the horizontal coordinate system needs to be multiplied by the cosine of that angle. However, through coordinate system transformation, the radiance component in the tilted plane normal direction can be directly obtained, avoiding cosine projection loss. This correction step is completed before integration and summation to ensure that the cumulative total radiance accurately reflects the actual radiant energy received by the PV panel, thereby improving the accuracy of power generation prediction.

[0083] As a preferred embodiment, the specific implementation of this application is as follows: In the actual application of the photovoltaic power station in the target area, the photovoltaic panel conversion efficiency parameters are set as follows: photovoltaic panel nominal efficiency 18%, temperature decay coefficient 0.0045 / ℃, dust loss coefficient 0.97, and inverter efficiency 98%. When the cumulative total radiation is calculated to be 1250 kWh / m²... 2 At that time, the average ambient temperature was monitored to be 28.6℃, and the standard test temperature was taken as 25℃. In specific implementation, the above parameters were substituted into the calculation model, using the formula E=1250×5000×0.18×(1-0.0045×(28.6-25))×0.97×0.98, where the total area of ​​the photovoltaic panels is 5000m². 2 The final predicted power generation value was 1,045,300 kWh. During the calculation, the temperature degradation coefficient was calibrated based on the temperature characteristic curve of the monocrystalline silicon module, and the dust loss coefficient was determined based on measured data from the quarterly cleaning cycle.

[0084] This application further proposes adding a data verification and compensation step before integral summation, including traversing all grid points and verifying the proportion of valid data points; identifying invalid grid points; and replacing them with data from surrounding valid grid points or historical data from the same period.

[0085] Valid data points are defined as those with a non-negative instantaneous total irradiance that is less than the solar constant. Invalid grid points are filtered out by setting a first preset threshold. For invalid grid points, the arithmetic mean of surrounding valid grid points is used as a replacement. If no valid data is available in the vicinity, the average historical irradiance of the same period in the region is used to fill the gap. The selection range of surrounding valid grid points can be dynamically adjusted according to the spatial resolution of the geographic region. For example, valid data can be searched within a 3×3 or 5×5 neighborhood centered on the target grid point. The time window for historical data can be set to the historical average of the same season or month to maintain temporal consistency.

[0086] Specifically, before calculating the cumulative total radiance, the system automatically iterates through the time-series data of each grid point and calculates the percentage of valid data points. When the percentage of valid data for a grid point falls below a threshold, a data compensation mechanism is triggered. For example, if a grid point experiences abnormal instantaneous total radiance at 50% of its time points due to cloud cover, the system will mark it as invalid and interpolate by using the average of valid data from the eight adjacent grid points. If all adjacent grid points are invalid, the average radiance for that location in the same month over the past five years will be used instead. This dual data compensation strategy ensures the completeness of radiance data for each grid point, preventing underestimation of the cumulative value due to missing local data. This process forms a closed-loop verification during the data aggregation phase, improving the robustness of radiance calculation under complex terrain.

[0087] As a preferred embodiment, the solution of this application is implemented as follows: In the data processing module of the photovoltaic power generation prediction system, the data verification and compensation steps are executed by an automated algorithm. When traversing all grid points, the system automatically identifies the numerical range of the instantaneous total radiation of each grid point within the prediction period, setting the value to be between 0 and 1361 W / m². 2 Data points between points are marked as valid data points. When the percentage of valid data points for a given grid point falls below 85%, the system triggers a data compensation mechanism: First, a 3×3 neighborhood window is generated centered on the grid point. Instantaneous radiance data of valid grid points within the neighborhood are extracted, and the arithmetic mean is calculated to overwrite the original invalid data. If there is no valid data within the neighborhood window, the system calls the geographic information database to match the historical average radiance of the latitude and longitude grid where the grid point is located for replacement. Historical data is derived from satellite observation records of the same month, day, and time within the last five years.

[0088] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A photovoltaic power generation amount prediction method based on illumination simulation, characterized by, The method comprises the following steps: S1, obtaining and processing data step: obtaining meteorological data sources and geographic data sources of a target area, and preprocessing the meteorological data sources and the geographic data sources to generate standardized meteorological data and geographic data; S2, constructing a terrain model step: based on the preprocessed geographic data, a digital elevation model of the target area is constructed; S3, calculating the position of the sun and direct radiation step: based on the preprocessed meteorological data and geographic data, the solar position information of each grid point in the target area within a predetermined time period is calculated, and the direct radiation of each grid point is further calculated; S4, light simulation analysis step: based on the digital elevation model and the solar position information of each grid point, a time point by time point light simulation analysis is performed to calculate the terrain shading coefficient and sky scattered radiation of each grid point; S5, aggregating total radiation step: aggregating the direct radiation, terrain shading coefficient and sky scattered radiation of each grid point within the predetermined time period to calculate the total radiation of the photovoltaic panel receiving surface of the target area; S6, predicting power generation step: based on the total radiation and a predetermined photovoltaic panel conversion efficiency parameter, the photovoltaic power generation of the target area within the predetermined time period is predicted.

2. The photovoltaic power generation amount prediction method based on illumination simulation according to claim 1, characterized in that, The step S1 specifically comprises the following sub-steps: S11, obtaining the meteorological data sources, which include historical solar direct radiation intensity, historical solar scattered radiation intensity, historical environmental temperature, historical humidity and historical air pressure data from satellite remote sensing; obtaining the geographic data sources, which include longitude and latitude coordinates, elevation data, slope data and slope direction data of the target area from a digital elevation model database; S12, preprocessing the meteorological data sources: first, using time series linear interpolation method to fill in the missing meteorological data; second, using threshold method to eliminate abnormal meteorological data that obviously exceeds the reasonable range, and again using time series linear interpolation method to fill in; finally, resampling the processed meteorological data to the same spatial resolution as the geographic data; S13, preprocessing the geographic data sources: converting the geographic data from different coordinate systems to the same geographic coordinate system, and resampling all geographic data to the same spatial resolution to generate the standardized geographic data.

3. The photovoltaic power generation amount prediction method based on illumination simulation according to claim 1, characterized in that, In the step S3, the specific method for calculating the solar position information of each grid point is: according to the longitude and latitude coordinates in the geographic data and the world standard time, the solar elevation angle and solar azimuth angle of each grid point at each time point within the predetermined time period are calculated by a solar position algorithm.

4. The photovoltaic power generation amount prediction method based on illumination simulation according to claim 3, characterized in that, In the step S3, the specific method for calculating the direct radiation of each grid point is: based on the solar direct radiation intensity in the preprocessed meteorological data, and combined with the calculated solar elevation angle, the direct radiation of each grid point at each time point is calculated by formula (I): wherein represents the direct radiation amount, represents the solar direct radiation intensity in the meteorological data, represents the solar elevation angle.

5. The photovoltaic power generation amount prediction method based on illumination simulation according to claim 1, characterized in that, In the step S4, the specific method for calculating the terrain shading coefficient is: based on the digital elevation model and the solar position information of each grid point at each time point, determining whether there is terrain shielding between the current grid point and the sun by a line-of-sight analysis algorithm; The line-of-sight analysis algorithm takes the current grid point as the starting point, calculates the visibility of the terrain obstacles along the way along the solar azimuth direction and at the solar elevation angle as the elevation angle; If there is shielding, the terrain shading coefficient is 0; If there is no shielding, the terrain shading coefficient is 1.

6. The photovoltaic power generation amount prediction method based on illumination simulation according to claim 5, characterized in that, In the step S4, the specific method for calculating the sky scattered radiation is: based on the solar scattered radiation intensity in the preprocessed meteorological data, the digital elevation model and the terrain shading coefficient, the sky scattered radiation is calculated by an anisotropic Hay model; The Hay model first calculates the anisotropic index of the sky scattered radiation, and then combines the sky view factor provided by the digital elevation model to finally calculate the sky scattered radiation of each grid point.

7. The photovoltaic power generation amount prediction method based on illumination simulation according to claim 1, characterized in that, In the step S5, the specific method for calculating the total radiation is: superimposing the direct radiation modified by the terrain shading coefficient and the sky scattered radiation, and calculating the instantaneous total radiation of each grid point at each time point by formula (II), and then integrating and summing the instantaneous total radiation of all time points in the preset time period to obtain the cumulative total radiation in the preset time period: wherein, denotes the cumulative total radiation, t denotes a time point, start denotes a start time of the preset time period, end denotes an end time of the preset time period, denotes the direct radiation at time point t, denotes the terrain shading factor at time point t, denotes the sky diffuse radiation at time point t, denotes a time step.

8. The photovoltaic power generation amount prediction method based on illumination simulation according to claim 7, characterized in that, In the step S5, before calculating the cumulative total radiation, it further includes a correction step of the installation inclination and azimuth of the photovoltaic panel: according to the installation parameters of the photovoltaic panel of the target area, the instantaneous total radiation is converted from the horizontal plane coordinate system to the photovoltaic panel inclined plane coordinate system by using a rotation matrix to correct the cosine projection loss. 9.The photovoltaic power generation amount prediction method based on illumination simulation of claim 1, wherein, In the step S6, the photovoltaic panel conversion efficiency parameters include the nominal efficiency of the photovoltaic panel, the temperature attenuation coefficient, the dust loss coefficient and the inverter efficiency; the specific method for predicting the photovoltaic power generation is: inputting the cumulative total radiation into a photovoltaic power generation calculation model, and combining the photovoltaic panel conversion efficiency parameters, and calculating the photovoltaic power generation by formula (III): wherein E represents the predicted photovoltaic power generation amount, represents the cumulative total radiation amount, A represents the photovoltaic panel area, represents the nominal efficiency of the photovoltaic panel, represents the temperature attenuation coefficient, represents the average ambient temperature in the preset time period, represents the standard test temperature, represents the dust loss coefficient, represents the inverter efficiency. 10.The photovoltaic power generation amount prediction method based on illumination simulation of claim 7, wherein, In the step S5, before integrating and summing the instantaneous total radiation, it further includes a data verification and compensation step: S51, traverse all grid points, and verify the proportion of valid data points of each grid point in the preset time period; the valid data point refers to the data point whose calculated instantaneous total radiation is non-negative and less than the solar constant; S52, if the proportion of valid data points of a certain grid point is lower than a first preset threshold, it is determined that the grid point is invalid; S53, for the grid point determined to be invalid, the arithmetic mean of the instantaneous total radiation of its surrounding valid grid points at the same time is used for replacement; S54, if there is no valid grid point around, the historical average radiation of the same period in the geographical area where the grid point is located is used for replacement.

Images