A global horizontal irradiance forecast method, system, device and storage medium

Abstract

This application discloses a global horizontal irradiance forecasting method, system, device, and storage medium. This application proposes a new forecasting method that fully considers key meteorological element data and model characteristic data to forecast global horizontal irradiance, thereby improving forecast accuracy. This application can be widely applied in the field of artificial intelligence technology.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a global horizontal irradiance forecasting method, system, device and storage medium. Background Art

[0002] As global demand for renewable energy continues to grow, the proportion of solar energy in power systems continues to increase, especially in high-altitude and arid regions. High-precision insolation forecasts are crucial for the efficient operation of photovoltaic power plants and the stable scheduling of power grids. In recent years, numerical weather prediction (NWP) models, especially the widely used Weather Research and Forecasting (WRF) model, have played an important role in simulating atmospheric processes and predicting solar radiation. However, under certain climatic conditions, especially cloudy weather in winter, the WRF model often exhibits systematic biases in cloud cover, leading to a significant underestimation of the global horizontal irradiance (GHI).

[0003] Despite recent progress in NWP model optimization and AI technology applications, global horizontal irradiance forecasts remain difficult to accurately predict under certain terrain and climate conditions. Therefore, technical challenges remain that need to be addressed. Summary of the Invention

[0004] The purpose of this application is to solve one of the technical problems existing in the prior art to at least a certain extent.

[0005] To this end, an object of an embodiment of the present application is to provide a global horizontal irradiance forecasting method, system, device and storage medium, which can improve the accuracy of global horizontal irradiance forecasting.

[0006] In order to achieve the above-mentioned technical objectives, the technical solutions adopted in the embodiments of the present application include: a global horizontal irradiance forecasting method, comprising the following steps: obtaining key meteorological element data and model characteristic data; determining a basic initial water vapor field based on the key meteorological element data, and running the basic initial water vapor field to determine an initial water vapor mixing ratio; inputting the model characteristic data into a first neural network model to obtain a spatial distribution correction factor; based on the spatial distribution correction factor and the initial water vapor mixing ratio, determining a target water vapor mixing ratio and writing the target water vapor mixing ratio into a WRF intermediate file; running the WRF intermediate file and generating a simulated global horizontal irradiance, and determining a deviation index of a second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance; inputting the deviation index into the second neural network model and adjusting the network weights and network parameters of the second neural network model to obtain a target neural network; running the target neural network to obtain global horizontal irradiance.

[0007] The present application can obtain key meteorological element data and model characteristic data; determine the basic initial water vapor field according to the key meteorological element data, and run the basic initial water vapor field to determine the initial water vapor mixing ratio; input the model characteristic data into the first neural network model to obtain the spatial distribution correction factor; based on the spatial distribution correction factor and the initial water vapor mixing ratio, determine the target water vapor mixing ratio and write the target water vapor mixing ratio into the WRF intermediate file; run the WRF intermediate file and generate simulated global horizontal irradiance, and determine the deviation index of the second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance; input the deviation index into the second neural network model and adjust the network weights and network parameters of the second neural network model to obtain the target neural network; run the target neural network to obtain the global horizontal irradiance. The present application proposes a new prediction method, which fully considers the key meteorological element data and model characteristic data to predict the global horizontal irradiance, and the present application can improve the prediction accuracy.

[0008] In addition, the global horizontal irradiance forecasting method according to the above embodiment of the present invention may also have the following additional technical features:

[0009] Furthermore, in the embodiment of the present application, the key meteorological element data is obtained by the following steps:

[0010] Obtain original meteorological element data;

[0011] The WRF preprocessing system is run to perform data interpolation processing on the original meteorological element data to obtain key meteorological element data.

[0012] Furthermore, in an embodiment of the present application, determining the target water vapor mixing ratio based on the spatial distribution correction factor and the initial water vapor mixing ratio specifically includes:

[0013] Determining an adjusted water vapor mixing ratio according to the initial water vapor mixing ratio and a preset basic dehumidification ratio;

[0014] determining a drying amplitude according to the spatial distribution correction factor;

[0015] A target water vapor mixing ratio is determined according to the adjusted water vapor mixing ratio and the drying amplitude.

[0016] Furthermore, in an embodiment of the present application, determining the deviation index of the second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance specifically includes:

[0017] Calculating the difference between the measured global horizontal irradiance and the simulated global horizontal irradiance;

[0018] The difference is used as the deviation indicator.

[0019] Furthermore, in the embodiment of the present application, the inputting of the model feature data into the first neural network model to obtain the spatial distribution correction factor specifically includes:

[0020] Standardizing the model feature data and generating time series data;

[0021] Dividing the time series data into a training set, a validation set, and a test set, and inputting the training set into a Transformer model to train the model, thereby obtaining a trained first neural network model;

[0022] The test set is input into the trained first neural network model to obtain a spatial distribution correction factor.

[0023] Furthermore, in the embodiment of the present application, determining and adjusting the water vapor mixing ratio according to the initial water vapor mixing ratio and the preset basic dehumidification ratio specifically includes:

[0024] Inputting the initial water vapor mixing ratio and the preset basic dehumidification ratio into a first formula to obtain an adjusted water vapor mixing ratio;

[0025] The first formula is:

[0026] Qvap_base=Qvap_original×(1-ΔRH_base)

[0027] Qvap_base is the adjusted water vapor mixing ratio, Qvap_original is the initial water vapor mixing ratio, and ΔRH_base is the preset basic dehumidification ratio.

[0028] Furthermore, in the embodiment of the present application, determining the target water vapor mixing ratio according to the adjusted water vapor mixing ratio and the drying amplitude specifically includes:

[0029] Inputting the adjusted water vapor mixing ratio and the drying amplitude into a second formula to obtain a target water vapor mixing ratio;

[0030] Wherein, the second formula is:

[0031] Qvap_adj=Qvap_base×(1-ΔRH_final)

[0032] Qvap_adj is the target water vapor mixing ratio, Qvap_base is the adjusted water vapor mixing ratio, and ΔRH_final is the drying amplitude.

[0033] In addition, this application also provides a global horizontal irradiance forecast system, including:

[0034] A first processing unit is used to obtain key meteorological element data and model characteristic data;

[0035] A second processing unit is configured to determine a basic initial water vapor field according to the key meteorological element data, and to run the basic initial water vapor field to determine an initial water vapor mixing ratio;

[0036] a third processing unit, configured to input the model feature data into the first neural network model to obtain a spatial distribution correction factor;

[0037] a fourth processing unit, configured to determine a target water vapor mixing ratio based on the spatial distribution correction factor and the initial water vapor mixing ratio, and write the target water vapor mixing ratio into a WRF intermediate file;

[0038] a fifth processing unit, configured to run the WRF intermediate file and generate a simulated global horizontal irradiance, and determine a deviation index of the second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance;

[0039] a sixth processing unit, configured to input the deviation indicator into the second neural network model and adjust the network weights and network parameters of the second neural network model to obtain a target neural network;

[0040] The seventh processing unit is used to run the target neural network to obtain the global horizontal irradiance.

[0041] In addition, the present application also provides a global horizontal irradiance forecasting device, comprising:

[0042] at least one processor;

[0043] at least one memory for storing at least one program;

[0044] When the at least one program is executed by the at least one processor, the at least one processor implements a global horizontal irradiance forecasting method as described in any one of the above items.

[0045] In addition, the present application also provides a computer-readable storage medium, which stores processor-executable instructions. When the processor executes the processor, the processor-executable instructions are used to execute a global horizontal irradiance forecasting method as described in any of the above items.

[0046] The advantages and benefits of this application will be partially given in the following description, and partially become apparent from the following description, or learned through practice of this application:

[0047] The present application can obtain key meteorological element data and model characteristic data; determine the basic initial water vapor field according to the key meteorological element data, and run the basic initial water vapor field to determine the initial water vapor mixing ratio; input the model characteristic data into the first neural network model to obtain the spatial distribution correction factor; based on the spatial distribution correction factor and the initial water vapor mixing ratio, determine the target water vapor mixing ratio and write the target water vapor mixing ratio into the WRF intermediate file; run the WRF intermediate file and generate simulated global horizontal irradiance, and determine the deviation index of the second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance; input the deviation index into the second neural network model and adjust the network weights and network parameters of the second neural network model to obtain the target neural network; run the target neural network to obtain the global horizontal irradiance. The present application proposes a new prediction method, which fully considers the key meteorological element data and model characteristic data to predict the global horizontal irradiance, and the present application can improve the prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A schematic flow chart of a global horizontal irradiance forecasting method according to a specific embodiment of the present invention;

[0049] Figure 2 The figure is a flow chart of a global horizontal irradiance forecasting method in another specific embodiment of the present invention. DETAILED DESCRIPTION

[0050] The embodiments of the present invention are described in detail below with reference to the accompanying drawings to illustrate the principles and processes of the global horizontal irradiance forecasting method, system, device and storage medium in the embodiments of the present invention.

[0051] This application provides a global horizontal irradiance forecasting method, comprising the following steps:

[0052] Obtain key meteorological element data and model characteristic data;

[0053] Determine the basic initial water vapor field based on key meteorological element data, and run the basic initial water vapor field to determine the initial water vapor mixing ratio;

[0054] Inputting the model feature data into the first neural network model to obtain a spatial distribution correction factor;

[0055] Based on the spatial distribution correction factor and the initial water vapor mixing ratio, the target water vapor mixing ratio is determined and written into the WRF intermediate file;

[0056] Run the WRF intermediate file and generate simulated global horizontal irradiance, and determine the deviation index of the second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance;

[0057] Inputting the deviation indicator into the second neural network model and adjusting the network weights and network parameters of the second neural network model to obtain a target neural network;

[0058] Run the target neural network to obtain the global horizontal irradiance.

[0059] The present application can obtain key meteorological element data and model characteristic data; determine the basic initial water vapor field according to the key meteorological element data, and run the basic initial water vapor field to determine the initial water vapor mixing ratio; input the model characteristic data into the first neural network model to obtain the spatial distribution correction factor; based on the spatial distribution correction factor and the initial water vapor mixing ratio, determine the target water vapor mixing ratio and write the target water vapor mixing ratio into the WRF intermediate file; run the WRF intermediate file and generate simulated global horizontal irradiance, and determine the deviation index of the second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance; input the deviation index into the second neural network model and adjust the network weights and network parameters of the second neural network model to obtain the target neural network; run the target neural network to obtain the global horizontal irradiance. The present application proposes a new prediction method, which fully considers the key meteorological element data and model characteristic data to predict the global horizontal irradiance, and the present application can improve the prediction accuracy.

[0060] In addition, the global horizontal irradiance forecasting method according to the above embodiment of the present invention may also have the following additional technical features:

[0061] Furthermore, in the embodiment of the present application, the key meteorological element data is obtained by the following steps:

[0062] Obtain original meteorological element data;

[0063] Run the WRF preprocessing system to perform data interpolation processing on the original meteorological element data to obtain key meteorological element data.

[0064] Furthermore, in the embodiment of the present application, the target water vapor mixing ratio is determined based on the spatial distribution correction factor and the initial water vapor mixing ratio, specifically including:

[0065] Determine and adjust the water vapor mixing ratio according to the initial water vapor mixing ratio and the preset basic dehumidification ratio;

[0066] Determine the drying amplitude based on the spatial distribution correction factor;

[0067] The target water vapor mixing ratio is determined based on the adjusted water vapor mixing ratio and the drying range.

[0068] Furthermore, in the embodiment of the present application, based on the measured global horizontal irradiance and the simulated global horizontal irradiance, the deviation index of the second neural network model is determined, specifically including:

[0069] Calculate the difference between the measured global horizontal irradiance and the simulated global horizontal irradiance;

[0070] The difference is used as the deviation indicator.

[0071] Furthermore, in the embodiment of the present application, the model feature data is input into the first neural network model to obtain the spatial distribution correction factor, which specifically includes:

[0072] Standardize the model feature data and generate time series data;

[0073] The time series data is divided into a training set, a validation set, and a test set. The training set is input into the Transformer model and the model is trained to obtain the trained first neural network model.

[0074] The test set is input into the trained first neural network model to obtain the spatial distribution correction factor.

[0075] Furthermore, in the embodiment of the present application, the water vapor mixing ratio is determined and adjusted according to the initial water vapor mixing ratio and the preset basic dehumidification ratio, specifically including:

[0076] Inputting the initial water vapor mixing ratio and the preset basic dehumidification ratio into the first formula to obtain the adjusted water vapor mixing ratio;

[0077] The first formula is:

[0078] Qvap_base=Qvap_original×(1-ΔRH_base)

[0079] Qvap_base is the adjusted water vapor mixing ratio, Qvap_original is the initial water vapor mixing ratio, and ΔRH_base is the preset basic dehumidification ratio.

[0080] Furthermore, in the embodiment of the present application, the target water vapor mixing ratio is determined based on the adjustment of the water vapor mixing ratio and the drying range, specifically including:

[0081] Input the adjusted water vapor mixing ratio and drying amplitude into the second formula to obtain the target water vapor mixing ratio;

[0082] Among them, the second formula is:

[0083] Qvap_adj=Qvap_base×(1-ΔRH_final)

[0084] Qvap_adj is the target water vapor mixing ratio, Qvap_base is the adjusted water vapor mixing ratio, and ΔRH_final is the drying amplitude.

[0085] The following combination Figure 1 as well as Figure 2 The principle of this application is explained.

[0086] Reference Figure 1 as well as Figure 2 This paper proposes a comprehensive technical solution combining "adaptive drying correction of water vapor fields, grid assimilation, and artificial intelligence-assisted optimization" to optimize the accuracy of WRF's GHI forecasts under cloudy winter conditions at high altitudes. Specifically, this technical solution achieves a deep integration of physical models and AI technology through the following key steps:

[0087] First, the system acquires key meteorological data from global or regional numerical forecast centers (such as the European Centre for Medium-Range Weather Forecasts (ECMWF), including wind (U, V), temperature (T), water vapor mixing ratio (Qvap), and high-resolution topographic information. This data is interpolated using the WRF Preprocessing System (WPS) and converted to a mesoscale grid with a horizontal resolution of 3 kilometers and 60 to 70 vertical layers. This generates the basic initial field, ensuring the accuracy and adaptability of the model's initial conditions.

[0088] Next, the system performs relative humidity (RH) on the generated basic initial field. original ) calculation, identify the water vapor distribution of each grid and vertical layer. In view of the problem of excessive cloud cover in winter in some high-altitude areas, a basic dehumidification ratio (ΔRH base ), for example, 10% to 15%, and make a preliminary dryness correction on the water vapor mixing ratio (Qvap) to obtain a preliminary adjusted water vapor mixing ratio (Qvap base This step is intended to reduce the excessive condensation of water vapor in the WRF model under certain conditions, thereby suppressing the formation of excessive cloud cover.

[0089] Based on the preliminary drying correction, a deep learning model based on the Transformer architecture is introduced as an AI submodule. This model uses multi-source input features, including historical cloud cover deviation, precipitation probability, terrain characteristics, satellite cloud top brightness temperature, radar echo data, current wind field and temperature distribution, etc., and extracts and learns features through a multi-layer attention mechanism to generate a spatially distributed correction factor f. AI(t,x,y) This correction factor is used to dynamically adjust the basic drying ratio to obtain the final drying amplitude ΔRH final , ensuring that the drying process can adaptively reflect real-time and historical meteorological conditions and terrain characteristics.

[0090] Then, the system adjusts the water vapor mixing ratio (Qvap adj ) is written to the WRF intermediate file and grid assimilation functionality (such as FDDA or Grid Nudging) is enabled in the WRF configuration file. During the assimilation process, the appropriate assimilation coefficients and application altitude layers (e.g., above 700 hPa) are set so that the model periodically returns meteorological elements such as Qvap, U, V, and T to the adjusted target fields during integration, thereby continuously suppressing excessive cloud condensation under overcast conditions. This process ensures the integration of physical processes with data-driven corrections, maintaining the physical consistency and continuity of model calculations.

[0091] After the simulation is complete, the system calculates the deviations between several measured GHIs and the simulated GHIs, ultimately calculating metrics such as the bias and root mean square error (RMSE) between these measured and simulated GHIs. This deviation data is fed back to the AI ​​model as new training samples, used to iteratively optimize the AI ​​model's weights, implementing a closed-loop optimization mechanism of "model calculation - result evaluation - AI correction - model re-simulation." Through multiple rounds of iteration, the AI ​​model continuously learns and optimizes itself, improving its prediction and correction capabilities under different meteorological conditions, ensuring that the system can adapt to complex and changing weather environments.

[0092] To automate the entire process, the system integrates the aforementioned steps into a highly efficient forecasting system, encompassing modules such as data acquisition, preprocessing, AI correction, water vapor adjustment, WRF simulation, and closed-loop optimization. Furthermore, a high-performance computing cluster is used as the system's operating platform to ensure efficient large-scale data processing and deep learning model training. Furthermore, a modular design facilitates system expansion and maintenance, supporting the flexible replacement and upgrade of different AI model architectures.

[0093] Ultimately, by combining multi-source observational data, AI dynamic adjustment, and physical model assimilation, this invention effectively reduced the systematic bias of the WRF model in winter cloudy weather conditions in certain high-altitude regions, significantly improving the accuracy and reliability of the GHI forecast. This technical solution is not only suitable for weather forecasting and photovoltaic power plant site selection in high-altitude, dry and cold regions, but also has strong scalability and adaptability, allowing for widespread application in rainy, cloudy, and foggy scenarios in other regions or seasons, providing high-precision solar radiation forecast support.

[0094] In other examples, to verify the effectiveness of the proposed "adaptive drying correction of water vapor field, grid assimilation, and artificial intelligence-assisted optimization" technical solution, five typical winter cloudy days in January 2024 at a photovoltaic power station in a high-altitude area were selected as examples. The specific implementation steps are as follows:

[0095] First, the system obtained key meteorological element data from the European Centre for Medium-Range Weather Forecasts (ECMWF) from January 1 to January 31, 2024, including wind field (U, V), temperature (T), water vapor mixing ratio (Qvap) and high-resolution terrain information. At the same time, satellite cloud top brightness temperature data, radar echo data and ground meteorological observation data for the same period were obtained as multi-source observation inputs. Through the WRF preprocessing system (WPS), these data were interpolated to a mesoscale grid with a horizontal resolution of 3 kilometers and 65 vertical layers to generate a basic initial field, providing accurate initial conditions for subsequent simulations.

[0096] Next, the system performs relative humidity (RH) on the generated basic initial field. original ) calculation, identify the water vapor distribution of each grid and vertical layer. In view of the problem of excessive cloud cover in winter in some high-altitude areas, the basic dehumidification ratio ΔRH is set base The water vapor mixing ratio (Qvap) was initially corrected to 12%, and the initially adjusted water vapor mixing ratio (Qvap base This adjustment is intended to reduce the amount of excess water vapor condensation in the WRF model under overcast conditions in winter, thereby suppressing the formation of excessive cloud cover.

[0097] Based on the preliminary drying correction, a deep learning model based on the Transformer architecture was introduced as an AI submodule. This model uses multi-source input features, including historical cloud cover deviation, precipitation probability, terrain characteristics, satellite cloud top brightness temperature, radar echo data, current wind field and temperature distribution, etc., and extracts and learns features through a multi-layer attention mechanism to generate a spatially distributed correction factor f. AI(t,x,y) This correction factor is used to dynamically adjust the basic drying ratio to obtain the final drying amplitude ΔRH final , ensuring that the drying process can adaptively reflect real-time and historical meteorological conditions and terrain characteristics.

[0098] Then, the system adjusts the water vapor mixing ratio (Qvap adj ) is written to the WRF intermediate file, and grid assimilation functions (such as FDDA) are enabled in the WRF configuration file. The assimilation coefficients are set to 0.0002 (for U, V, and T) and 0.00012 (for Qvap), and the application altitude is above 700hPa. Through grid assimilation, the model periodically pulls meteorological elements such as Qvap, U, V, and T back to the adjusted target field during the integration process, thereby continuously suppressing excessive cloud condensation under cloudy conditions. This process ensures the integration of physical processes with data-driven corrections, maintaining the physical consistency and continuity of model calculations.

[0099] After the simulation is completed, the system compares the measured GHI with the simulated GHI and calculates the deviation indicators, including Bias and Root Mean Square Error (RMSE). For example, on January 15, 2024, the Bias before adjustment is -250W / m 2 , RMSE is 320W / m 2 ; The adjusted Bias is -30W / m 2 , RMSE is 130W / m 2 . These deviation data are fed back to the AI ​​model as new training samples to further optimize its weights and parameters. Through multiple rounds of iteration, the AI ​​model gradually corrects and optimizes the deviation, improves its prediction and correction capabilities under different meteorological conditions, and forms a closed-loop optimization mechanism of "model calculation-result evaluation-AI correction-model re-simulation". Specifically, through verification on five winter cloudy weather days (January 5, January 10, January 15, January 20 and January 25, 2024) at a photovoltaic power station in certain high-altitude areas, the results show that the technical solution of the present invention significantly improves the accuracy of GHI forecast. The average Bias before adjustment is -220W / m 2 , after adjustment, it dropped to -25W / m 2 ; Average RMSE from 310W / m 2 Reduced to 125W / m 2 These results show that the comprehensive technical solution proposed in this paper can effectively reduce the systematic deviation of the WRF model under cloudy conditions in winter in practical applications, significantly improve the accuracy and reliability of GHI forecasts, and is particularly suitable for weather forecasting and photovoltaic power station site selection assessment in high-altitude, dry and cold regions.

[0100] In addition, with Figure 1 Corresponding to the method, the embodiments of the present application also provide a global horizontal irradiance forecast system, including:

[0101] A first processing unit is used to obtain key meteorological element data and model characteristic data;

[0102] A second processing unit is configured to determine a basic initial water vapor field according to the key meteorological element data, and to run the basic initial water vapor field to determine an initial water vapor mixing ratio;

[0103] a third processing unit, configured to input the model feature data into the first neural network model to obtain a spatial distribution correction factor;

[0104] a fourth processing unit, configured to determine a target water vapor mixing ratio based on the spatial distribution correction factor and the initial water vapor mixing ratio, and write the target water vapor mixing ratio into a WRF intermediate file;

[0105] a fifth processing unit, configured to run the WRF intermediate file and generate a simulated global horizontal irradiance, and determine a deviation index of the second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance;

[0106] a sixth processing unit, configured to input the deviation indicator into the second neural network model and adjust the network weights and network parameters of the second neural network model to obtain a target neural network;

[0107] The seventh processing unit is used to run the target neural network to obtain the global horizontal irradiance.

[0108] It should be noted that the first processing unit may be any integrated circuit unit or microprocessor unit obtained by integrating a processing chip and its peripheral circuits using existing integration technologies. The first and second processing units may also be any integrated circuit modules or microprocessor modules obtained by integrating a processing chip and its peripheral circuits using existing integration technologies. The first and second processing units may also include one or more memories.

[0109] It should be noted that the contents of the above-mentioned global horizontal irradiance forecast method embodiment are all applicable to the present global horizontal irradiance forecast system embodiment. The functions specifically implemented by the present global horizontal irradiance forecast system embodiment are the same as those of the above-mentioned global horizontal irradiance forecast method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned global horizontal irradiance forecast method embodiment.

[0110] In addition, the present invention also provides a global horizontal irradiance forecasting device, including:

[0111] at least one processor 1011;

[0112] at least one memory 1012, configured to store at least one program;

[0113] When the at least one program is executed by the at least one processor, the at least one processor implements the global horizontal irradiance forecasting method.

[0114] The contents of the above method embodiments are all applicable to the present device embodiments. The functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0115] and Figure 1Corresponding to the method, an embodiment of the present application further provides a computer-readable storage medium, which stores processor-executable instructions, and the processor-executable instructions are used to execute the global horizontal irradiance forecasting method when executed by the processor.

[0116] The contents of the above-mentioned global horizontal irradiance forecast method embodiment are all applicable to the present storage medium embodiment. The functions specifically implemented by the present storage medium embodiment are the same as those of the above-mentioned global horizontal irradiance forecast method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned global horizontal irradiance forecast method embodiment.

[0117] In some optional embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiments presented and described in the flow chart of the present application are provided in an exemplary manner for the purpose of providing a more comprehensive understanding of the technology. The disclosed method is not limited to the operations and logical flows presented herein. Optional embodiments are contemplated in which the order of the various operations is changed and the sub-operations described as a part of a larger operation are performed independently.

[0118] In addition, although the present application is described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It is also understood that a detailed discussion of the actual implementation of each module is not necessary for understanding the present application. More specifically, given the properties, functions, and internal relationships of the various functional modules in the devices disclosed herein, the actual implementation of the module will be understood within the routine skills of an engineer. Therefore, a person skilled in the art can implement the present application as set forth in the claims using ordinary techniques without undue experimentation. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present application, which is determined by the full scope of the appended claims and their equivalents.

[0119] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several programs for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0120] The logic and / or steps represented in a flowchart or otherwise described herein, for example, may be considered as an ordered list of executable programs for implementing the logical functions, and may be embodied in any computer-readable medium for use by, or in conjunction with, a program execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can retrieve and execute a program from a program execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" may be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, a program execution system, apparatus, or device.

[0121] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0122] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable program execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0123] In the above description of this specification, reference to the terms "one embodiment / example," "another embodiment / example," or "certain embodiments / examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples.

[0124] Although the embodiments of the present application have been shown and described, those skilled in the art will appreciate that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and intent of the present application, and that the scope of the present application is defined by the claims and their equivalents.

[0125] The above is a specific description of the preferred implementation of the present application, but the present application is not limited to the embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present application. These equivalent modifications or substitutions are all included in the scope defined by the claims of the present application.

Claims

1. A global horizontal irradiance forecasting method, characterized in that: The following steps are involved: Obtain key meteorological element data and model characteristic data; Determine a basic initial water vapor field based on the key meteorological element data, and run the basic initial water vapor field to determine an initial water vapor mixing ratio; Inputting the model feature data into a first neural network model to obtain a spatial distribution correction factor; Determining a target water vapor mixing ratio based on the spatial distribution correction factor and the initial water vapor mixing ratio and writing the target water vapor mixing ratio into a WRF intermediate file; Running the WRF intermediate file and generating a simulated global horizontal irradiance, and determining a deviation index of the second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance; Inputting the deviation indicator into the second neural network model and adjusting the network weights and network parameters of the second neural network model to obtain a target neural network; The target neural network is run to obtain the global horizontal irradiance.

2. A global horizontal irradiance forecasting method according to claim 1, characterized in that: The key meteorological element data are obtained by the following steps: Obtain original meteorological element data; The WRF preprocessing system is run to perform data interpolation processing on the original meteorological element data to obtain key meteorological element data.

3. A global horizontal irradiance forecasting method according to claim 1, characterized in that: The determining of the target water vapor mixing ratio based on the spatial distribution correction factor and the initial water vapor mixing ratio specifically includes: Determining an adjusted water vapor mixing ratio according to the initial water vapor mixing ratio and a preset basic dehumidification ratio; determining a drying amplitude according to the spatial distribution correction factor; A target water vapor mixing ratio is determined according to the adjusted water vapor mixing ratio and the drying amplitude.

4. A global horizontal irradiance forecasting method according to claim 1, characterized in that: Determining the deviation index of the second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance specifically includes: Calculating the difference between the measured global horizontal irradiance and the simulated global horizontal irradiance; The difference is used as the deviation indicator.

5. A global horizontal irradiance forecasting method according to claim 1, characterized in that: Inputting the model feature data into the first neural network model to obtain the spatial distribution correction factor specifically includes: Standardizing the model feature data and generating time series data; Dividing the time series data into a training set, a validation set, and a test set, inputting the training set into a Transformer model and training the model to obtain a trained first neural network model; The test set is input into the trained first neural network model to obtain a spatial distribution correction factor.

6. A global horizontal irradiance forecasting method according to claim 3, characterized in that: The step of determining and adjusting the water vapor mixing ratio according to the initial water vapor mixing ratio and the preset basic dehumidification ratio specifically includes: Inputting the initial water vapor mixing ratio and the preset basic dehumidification ratio into a first formula to obtain an adjusted water vapor mixing ratio; The first formula is: Qvap_base=Qvap_original×(1-ΔRH_base) Qvap_base is the adjusted water vapor mixing ratio, Qvap_original is the initial water vapor mixing ratio, and ΔRH_base is the preset basic dehumidification ratio.

7. A global horizontal irradiance forecasting method according to claim 3, characterized in that: The step of determining a target water vapor mixing ratio according to the adjusted water vapor mixing ratio and the drying amplitude specifically includes: Inputting the adjusted water vapor mixing ratio and the drying amplitude into a second formula to obtain a target water vapor mixing ratio; Wherein, the second formula is: Qvap_adj=Qvap_base×(1-ΔRH_final) Qvap_adj is the target water vapor mixing ratio, Qvap_base is the adjusted water vapor mixing ratio, and ΔRH_final is the drying amplitude.

8. A global horizontal irradiance forecast system, characterized in that: include: A first processing unit is used to obtain key meteorological element data and model characteristic data; A second processing unit is configured to determine a basic initial water vapor field according to the key meteorological element data, and to run the basic initial water vapor field to determine an initial water vapor mixing ratio; a third processing unit, configured to input the model feature data into the first neural network model to obtain a spatial distribution correction factor; a fourth processing unit, configured to determine a target water vapor mixing ratio based on the spatial distribution correction factor and the initial water vapor mixing ratio, and write the target water vapor mixing ratio into a WRF intermediate file; a fifth processing unit, configured to run the WRF intermediate file and generate a simulated global horizontal irradiance, and determine a deviation index of the second neural network model based on the measured global horizontal irradiance and the simulated global horizontal irradiance; a sixth processing unit, configured to input the deviation indicator into the second neural network model and adjust the network weights and network parameters of the second neural network model to obtain a target neural network; The seventh processing unit is used to run the target neural network to obtain the global horizontal irradiance.

9. A global horizontal irradiance forecasting device, characterized in that include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the global horizontal irradiance forecasting method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing instructions executable by a processor, characterized in that: The processor-executable instructions are used to execute a global horizontal irradiance forecasting method as described in any one of claims 1 to 7 when executed by the processor.

Images