Numerical weather prediction optimization method and system based on meteorological resource coupling

Abstract

The application discloses a numerical weather prediction optimization method and system based on meteorological resource coupling, belongs to the technical field of numerical weather prediction, focuses on the multi-source data in the target area, and constructs a double-dimension correlation between the power grid and the meteorology, integrates the meteorological resources, and is targeted to gather the power grid output, so that the subsequent numerical weather prediction is more in line with the power prediction requirements of the power grid, and through the correlation matrix and the comprehensive influence coefficient, the qualitative coupling relationship is converted into quantitative indexes, linear and nonlinear are considered, and therefore the prediction accuracy under complex weather and complex terrain is greatly improved, high-precision meteorological driving data is provided for photovoltaic short-term power prediction, and the problem that the existing numerical weather prediction is mostly based on single dimension for prediction, so that it is difficult to guarantee the numerical accuracy when facing environmental changes is solved.

Description

Technical Field

[0001] This invention relates to the field of numerical weather prediction technology, specifically to a numerical weather prediction optimization method and system based on meteorological resource coupling. Background Technology

[0002] Guided by the "dual carbon" goal, distributed photovoltaic (PV) power, as a core force in the clean energy transformation, has entered the stage of large-scale grid connection. However, the output of distributed PV is affected by meteorological factors such as solar radiation, cloud cover, and temperature, exhibiting significant randomness, volatility, and intermittency. Large-scale integration poses severe challenges to grid dispatch optimization, safe and stable operation, and renewable energy consumption. Traditional power forecasting methods mostly rely on single-point measured data, failing to adequately adapt to the regional characteristics of complex terrains (such as mountains and coastlines) and special weather conditions (typhoons, sea fog, and severe convection). Furthermore, insufficient meteorological input accuracy and the lack of effective integration of nonlinear mapping models result in large short-term forecast errors, making it difficult to meet the needs of refined grid dispatching. Numerical weather forecasting can provide large-scale, high-precision gridded meteorological data, offering crucial support for solving these problems. However, most existing numerical weather forecasting products rely on a single dimension for prediction, resulting in inconsistent accuracy under different geographical environments and weather conditions, severely impacting the accuracy of power forecasting. Therefore, a numerical weather forecasting method with strong anti-interference capabilities and high reliability is needed.

[0003] Chinese Patent, Publication No. CN117748467A, Publication Date: March 22, 2024, discloses a wind power prediction method and system based on multi-source numerical weather forecast fusion. The method involves acquiring actual wind speed and power data from wind farms, as well as forecast wind speed data from different numerical weather forecasts. Based on the actual wind speed, actual power data, and forecast wind speed data, correlation analysis and error analysis are used to initially screen numerical weather forecast sources to be fused. A multi-source numerical weather forecast fusion method library is constructed based on these sources. Decision strategies for numerical weather forecast fusion methods under different scenarios are formulated based on this library, and a deep learning-based wind power prediction model is trained. The fused wind speed for the forecast period is input into the wind power prediction model to achieve wind power prediction. While this method addresses the low accuracy of wind power prediction caused by the low accuracy of existing single numerical weather forecasts to some extent, it only considers the relationship between wind speed and actual power, without taking into account the impact of complex terrain and complex weather conditions on the accuracy of numerical weather forecasts. Summary of the Invention

[0004] This invention addresses the problem that existing numerical weather predictions, which mostly rely on a single dimension, struggle to maintain accuracy in the face of environmental changes. It provides a numerical weather prediction optimization method and system based on meteorological resource coupling. By focusing on multi-source data within a target area and constructing a targeted two-dimensional correlation between the power grid and meteorology, it not only integrates meteorological resources but also specifically focuses on power grid output. This makes the subsequent numerical weather predictions more aligned with the power grid's power forecasting requirements. Furthermore, through correlation matrices and comprehensive influence coefficients, the qualitative coupling relationship is transformed into quantitative indicators, balancing linear and nonlinear factors. This significantly improves forecast accuracy under complex weather and terrain conditions, providing high-precision meteorological driving data for short-term photovoltaic power prediction.

[0005] In a first aspect, one technical solution provided in this embodiment of the invention is: a numerical weather prediction optimization method based on meteorological resource coupling, comprising the following steps: S1. Collect meteorological and geographical data within the target area and preprocess them to obtain the raw dataset; S2. Construct power grid correlation relationships by performing power grid topology coupling on geographic data; construct meteorological correlation relationships by performing numerical forecast correlation coupling on meteorological data; S3. Unify and quantify the power grid correlation and meteorological correlation to construct a correlation matrix and calculate the comprehensive influence coefficient; S4. Construct a numerical weather prediction model, using the comprehensive influence coefficient and correlation matrix as model inputs to obtain the target forecast results.

[0006] This solution addresses the issues of messy and spatiotemporally inconsistent traditional forecast data by collecting and preprocessing multi-source meteorological and geographic data from the target area, and standardizing formats, time series, and quality. This provides a solid data foundation for coupled modeling. By constructing a targeted two-dimensional correlation between the power grid and meteorology, the solution accurately captures the patterns of strongly coupled meteorological factors such as topography, power grid topology constraints, and irradiance, thus aligning with the complex characteristics of the region and avoiding redundant information interference. Furthermore, by constructing a correlation matrix and calculating the comprehensive influence coefficient, the qualitative coupling relationship is transformed into a quantitative indicator, balancing linear and nonlinear correlations. This clarifies the contribution of each factor and overcomes the pain point of ambiguous influence weights in traditional methods. By inputting quantitative parameters into the numerical forecasting model, the model deeply integrates regional characteristics, significantly improving forecast accuracy under complex weather and terrain conditions. This provides high-precision meteorological driving data for short-term photovoltaic power prediction, thereby supporting refined power grid scheduling and efficient renewable energy consumption.

[0007] Optionally, in S1, meteorological and geographical data within the target area are collected and preprocessed to obtain the raw dataset, including the following steps: Information on light shading is obtained by monitoring atmospheric cloud cover and cloud velocity using ground-based cloud mapping equipment; Solar irradiance, temperature, humidity, wind speed, and air pressure are collected from ground weather stations as measured environmental information; Surface downwave radiation information was obtained by inverting cloud and aerosol parameters based on meteorological satellites and atmospheric radiative transfer models. Sunlight shading information, measured environmental information, and surface downwave radiation information are used as meteorological data, while distributed power generation data and topographic data are used as geographic data. The raw dataset is obtained by preprocessing meteorological and geographical data.

[0008] In this scheme, meteorological data is collected from multiple sources, including ground-based cloud maps, ground meteorological stations, and meteorological satellites, comprehensively covering information such as sunlight shading, environmental parameters, and shortwave radiation. At the same time, it combines distributed power generation from the power grid and topographic data to ensure the integrity of the data dimensions, providing high-quality, barrier-free raw data for subsequent coupled modeling and laying a solid foundation for improving the accuracy of numerical forecasts.

[0009] Optionally, in S2, the grid topology coupling of geographically relevant data is used to construct grid correlation relationships, including the following steps: Feature extraction is performed on distributed power generation data of the power grid to obtain power grid topology features, including topology type, topology parameters, and grid connection parameters; Feature extraction is performed on terrain data to obtain geographic features, including terrain parameters, surface parameters, and location parameters; Using the grid connection node ID of the photovoltaic site in the grid connection parameters as the grid column code, the geographical features and grid topology features corresponding to the column code are associated sequentially to construct the grid association relationship.

[0010] In this scheme, by extracting power grid topology features and geographical features, the constraints of power grid structure and regional geographical differences can be fully captured. At the same time, by using the grid-connected node ID of photovoltaic sites as a link, the corresponding geographical features and power grid topology features are accurately associated, thereby breaking down the isolation barrier between the two and constructing a logically clear power grid association relationship. This lays a precise foundation for the subsequent unified quantification of the association relationship with meteorology, and also allows the coupling relationship to fully adapt to the linkage law between regional geography and power grid, reducing forecast errors caused by association distortion, and supporting numerical weather prediction models to be more in line with actual application scenarios.

[0011] Optionally, in S2, numerical forecasting correlation coupling is performed on meteorological data to construct meteorological correlation relationships, including the following steps: Historical power output data of the power grid under different meteorological scenarios were used as experimental samples. Based on the data types of meteorological data, corresponding meteorological factors are set, and linear coupling relationship analysis and nonlinear coupling relationship analysis are performed on various meteorological factors based on experimental samples to obtain linear coupling coefficients and nonlinear coupling coefficients. The influence relationship between meteorological data corresponding to various meteorological factors and power grid output is determined based on linear and nonlinear coupling coefficients as meteorological correlation relationships.

[0012] This scheme selects historical power grid output data under different meteorological scenarios as experimental samples, comprehensively covering complex weather conditions such as sunny days, typhoons, and sea fog, ensuring the representativeness and completeness of the analysis samples. By conducting linear and nonlinear coupling relationship analysis on various meteorological factors, the linear and nonlinear coupling coefficients are accurately obtained. This captures strong linear correlations such as irradiance and temperature, while also not overlooking nonlinear effects such as humidity and wind speed, avoiding correlation distortion caused by a single analysis mode. By clarifying the specific impact of meteorological data on power grid output, a clear basis is provided for the unified quantification of subsequent correlations, allowing the coupling relationship to accurately characterize the effects of meteorological factors, reducing errors driven by meteorology in numerical forecasting, and laying a solid foundation for improving the accuracy of short-term photovoltaic power prediction.

[0013] Optionally, in S3, the correlation matrix is ​​constructed by uniformly quantifying the power grid correlation and meteorological correlation, and the comprehensive influence coefficient is calculated, including the following steps: A correlation matrix is ​​constructed using the grid-connected node ID of the photovoltaic site as the row and geographical features, grid topology features, and meteorological factors as the columns; The matrix value is calculated based on the correlation matrix as the comprehensive influence coefficient, which represents the degree of influence of the parameters corresponding to the columns in the correlation matrix on the power grid output.

[0014] In this scheme, a correlation matrix is ​​constructed using the grid-connected node ID of photovoltaic sites as rows and geographical / grid / meteorological features as columns. This efficiently integrates the results of two-dimensional correlations, breaking down the isolated barriers between geographical, grid, and meteorological correlations, thereby achieving a systematic sorting of multi-source features. By calculating the comprehensive influence coefficient, qualitative influence relationships are transformed into quantitative indicators, accurately quantifying the degree of influence of each parameter on grid output. This solves the problem of ambiguous influence weights in traditional coupling. This process makes the coupling correlation relationship more in line with the actual situation in the region, providing standardized and high-precision quantitative input for subsequent numerical forecasting models, significantly reducing forecast errors caused by correlation distortion, and laying a solid quantitative foundation for its application in photovoltaic power prediction.

[0015] Optionally, in S4, the numerical weather prediction model includes a single-model prediction layer and a comprehensive analysis layer; The single-model prediction layer obtains individual numerical weather prediction values ​​by matching the input data with the corresponding prediction model based on the model input. The comprehensive analysis layer assigns influence weights to various types of individual numerical weather prediction values ​​based on the comprehensive influence coefficient, and then performs a weighted summation of the individual numerical weather prediction values ​​based on the influence weights to obtain the target forecast result.

[0016] In this scheme, a single-model prediction layer matches the input data with the corresponding prediction model, accurately adapting to different parameter characteristics, thereby ensuring the relevance and basic accuracy of individual numerical weather predictions. Through a comprehensive analysis layer, influence weights are allocated based on the comprehensive influence coefficient, quantifying the degree of influence of geographical, power grid, and meteorological parameters, avoiding biases caused by single models or equal weights. This fully integrates regional characteristics and multi-source correlation patterns, and the target forecast results obtained after weighted summation are more in line with the actual scenario, effectively improving the forecast accuracy under complex conditions, and providing high-quality meteorological driving data for subsequent application to photovoltaic short-term power prediction.

[0017] Optionally, the single-model prediction layer includes a geographic prediction model, a power grid topology prediction model, and a meteorological prediction model; The geographic prediction model predicts numerical weather based on geographic features in the correlation matrix; The power grid topology prediction model predicts numerical weather based on the power grid topology features in the correlation matrix; The meteorological forecasting model predicts numerical weather based on various meteorological factors in the correlation matrix.

[0018] In this scheme, three types of prediction models—geographic, power grid topology, and meteorological—are subdivided, and each model makes accurate predictions based on the corresponding features in the correlation matrix. This allows for targeted capture of the influence of each dimension, adapting to differences in regional terrain, power grid structure, and meteorological patterns. This avoids the problem of insufficient adaptation of a single model to multi-source features, significantly improving the targeting and accuracy of individual predictions. It also provides high-quality basic data for the weighted summation of the comprehensive analysis layer, further strengthening the accuracy support for numerical forecasting.

[0019] Secondly, one technical solution provided in this embodiment of the invention is: a numerical weather forecast optimization system based on meteorological resource coupling, including a data acquisition module, a data coupling module, a quantization module and a forecasting module; The data acquisition module collects meteorological and geographical data within the target area and preprocesses them to obtain the raw dataset. The data coupling module performs power grid topology coupling on geographic data to construct power grid correlations, and performs numerical forecast correlation coupling on meteorological data to construct meteorological correlations. The quantification module unifies and quantifies the power grid correlation and meteorological correlation to construct a correlation matrix and calculates the comprehensive influence coefficient; The prediction module is equipped with a numerical weather prediction model, which uses the comprehensive influence coefficient and correlation matrix as model inputs to obtain the target forecast results.

[0020] In this solution, a corresponding system is built to integrate the numerical weather forecasting optimization method, thereby enabling human-computer interaction and improving the user experience.

[0021] Thirdly, one technical solution provided in this embodiment of the invention is: a computer device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; the memory is used to store computer programs; the processor is used to execute the program stored in the memory to implement the steps of the numerical weather forecast optimization method based on meteorological resource coupling.

[0022] Fourthly, one technical solution provided in this embodiment of the invention is: a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps of a numerical weather forecast optimization method based on meteorological resource coupling.

[0023] The beneficial effects of this invention are as follows: By focusing on multi-source data within the target area and constructing a two-dimensional correlation between the power grid and meteorology, this invention not only integrates meteorological resources but also specifically focuses on power grid output, thereby making the subsequent numerical weather forecasts more in line with the power grid's power prediction requirements. At the same time, through the correlation matrix and comprehensive influence coefficient, the qualitative coupling relationship is transformed into quantitative indicators, taking into account both linear and nonlinear factors, thus significantly improving the forecast accuracy under complex weather and terrain conditions, and providing high-precision meteorological driving data for short-term photovoltaic power prediction.

[0024] The above description of the invention is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0025] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings. The drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings.

[0026] Figure 1This is a flowchart of the numerical weather prediction optimization method based on meteorological resource coupling according to the present invention; Figure 2 This is a schematic diagram of the numerical weather prediction optimization system based on meteorological resource coupling according to the present invention; Figure 3 This is a schematic diagram of a computer device provided by the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only one preferred embodiment of this invention and are only used to explain this invention. They do not limit the scope of protection of this invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0028] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the operations (or steps) as sequential processes, many of the operations (or steps) can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but it may also have additional steps not included in the figures; the process may correspond to a method, function, procedure, subroutine, subroutine, etc.

[0029] Example 1: To address the problem that existing numerical weather predictions, which mostly rely on a single dimension, struggle to maintain accuracy in the face of environmental changes, this example provides a numerical weather prediction optimization method based on meteorological resource coupling, such as... Figure 1 As shown, it includes the following steps: S1: Collect meteorological and geographical data within the target area and preprocess them to obtain the raw dataset.

[0030] In this embodiment, meteorological and geographical data within the target area are collected and preprocessed to obtain the raw dataset, including the following steps: Information on light shading is obtained by monitoring atmospheric cloud cover and cloud velocity using ground-based cloud mapping equipment; Solar irradiance, temperature, humidity, wind speed, and air pressure are collected from ground weather stations as measured environmental information; Surface downwave radiation information was obtained by inverting cloud and aerosol parameters based on meteorological satellites and atmospheric radiative transfer models. Sunlight shading information, measured environmental information, and surface downwave radiation information are used as meteorological data, while distributed power generation data and topographic data are used as geographic data. The raw dataset is obtained by preprocessing meteorological and geographical data.

[0031] This embodiment collects meteorological data from multiple sources, including ground-based cloud maps, ground meteorological stations, and meteorological satellites, comprehensively covering information such as sunlight shading, environmental parameters, and shortwave radiation. At the same time, it combines distributed power generation from the power grid and topographic data to ensure the integrity of the data dimensions, providing high-quality, barrier-free raw data for subsequent coupled modeling and laying a solid foundation for improving the accuracy of numerical forecasts.

[0032] S2: Construct power grid correlation by coupling the geographic data with the power grid topology; construct meteorological correlation by coupling the meteorological data with numerical forecast correlation.

[0033] In this embodiment, the construction of power grid correlation by coupling geographic data with power grid topology includes the following steps: Feature extraction is performed on distributed power generation data of the power grid to obtain power grid topology features, including topology type, topology parameters, and grid connection parameters; Feature extraction is performed on terrain data to obtain geographic features, including terrain parameters, surface parameters, and location parameters; Using the grid connection node ID of the photovoltaic site in the grid connection parameters as the grid column code, the geographical features and grid topology features corresponding to the column code are associated sequentially to construct the grid association relationship.

[0034] This embodiment extracts power grid topology features and geographical features to comprehensively capture power grid structural constraints and regional geographical differences. At the same time, it uses the grid-connected node ID of photovoltaic sites as a link to accurately associate the corresponding geographical features with the power grid topology features, thereby breaking down the isolation barrier between the two and constructing a logically clear power grid correlation relationship. This lays a precise foundation for the subsequent unified quantification of the correlation with meteorology, and also allows the coupling relationship to fully adapt to the linkage pattern of regional geography and power grid, reducing forecast errors caused by correlation distortion, and supporting numerical weather prediction models to be more in line with actual application scenarios.

[0035] In this embodiment, the meteorological correlation relationship is constructed by numerical forecasting correlation coupling of meteorological data, including the following steps: Historical power output data of the power grid under different meteorological scenarios were used as experimental samples. Based on the data types of meteorological data, corresponding meteorological factors are set, and linear coupling relationship analysis and nonlinear coupling relationship analysis are performed on various meteorological factors based on experimental samples to obtain linear coupling coefficients and nonlinear coupling coefficients. The influence relationship between meteorological data corresponding to various meteorological factors and power grid output is determined based on linear and nonlinear coupling coefficients as meteorological correlation relationships.

[0036] Specifically, linear coupling relationship analysis was performed on various meteorological factors, using the Pearson correlation coefficient for quantification. The calculation formula is as follows: Where X represents the numerical value of the meteorological factor. Let Y represent the value of the i-th meteorological factor in the experimental sample, and Y represent the photovoltaic output value. and The mean values ​​of meteorological factors and photovoltaic output corresponding to the experimental samples are used. During the calculation, the correlation coefficient between each meteorological factor (irradiance, temperature, etc.) and the output of the corresponding station is calculated point by point according to the sample to obtain the factor-output correlation coefficient of a single station. Then, the average value of all stations in the region is taken as the final linear coupling coefficient.

[0037] Nonlinear coupling relationship analysis was performed on various meteorological factors, specifically using the mutual information method. First, the meteorological factors and output data were discretized into several intervals. In this embodiment, 10 intervals were selected. The calculation formula is as follows: in, Let X be the information entropy. For joint probability, and Using marginal probabilities, the mutual information value between each meteorological factor and the output is calculated. The larger the value, the stronger the nonlinear correlation.

[0038] like or mutual information value A value of 0.3 indicates that the meteorological factor is strongly coupled. , or 0.1 A mutual information value <0.3 indicates that the meteorological factor has moderate coupling. Furthermore, if the mutual information value is <0.1, it indicates that the meteorological factor is weakly coupled. The following results were obtained after calculating some meteorological factors: Strong coupling factors include solar irradiance (r=0.7948, mutual information value=0.42) and ambient temperature (r=0.5251, mutual information value=0.18); moderate coupling factors include relative humidity (r=-0.2457, mutual information value=0.12) and wind speed (r=0.2169, mutual information value=0.11); weak coupling factors include atmospheric pressure (r=0.0926, mutual information value=0.05). In subsequent numerical weather prediction, strong coupling factors will be retained as the key targets for subsequent numerical prediction, moderate coupling factors will be used as auxiliary factors, and weak coupling factors will not be included in the numerical weather prediction model for the time being.

[0039] This embodiment selects historical power grid output data under different meteorological scenarios as experimental samples, comprehensively covering complex weather conditions such as sunny days, typhoons, and sea fog, ensuring the representativeness and completeness of the analysis samples. By conducting linear and nonlinear coupling relationship analysis on various meteorological factors, the linear and nonlinear coupling coefficients are accurately obtained. This captures strong linear correlations such as irradiance and temperature, while also not overlooking nonlinear effects such as humidity and wind speed, avoiding correlation distortion caused by a single analysis mode. By clarifying the specific impact of meteorological data on power grid output, a clear basis is provided for the unified quantification of subsequent correlations, allowing the coupling relationship to accurately characterize the effects of meteorological factors, reducing errors driven by meteorology in numerical forecasting, and laying a solid foundation for improving the accuracy of short-term photovoltaic power prediction.

[0040] S3: Unify and quantify the correlation between power grid and meteorological factors to construct a correlation matrix and calculate the comprehensive influence coefficient.

[0041] In this embodiment, the correlation matrix is ​​constructed by uniformly quantifying the correlation between power grid and meteorological factors, and the comprehensive influence coefficient is calculated. This includes the following steps: A correlation matrix is ​​constructed using the grid-connected node ID of the photovoltaic site as the row and geographical features, grid topology features, and meteorological factors as the columns; The matrix value is calculated based on the correlation matrix as the comprehensive influence coefficient, which represents the degree of influence of the parameters corresponding to the columns in the correlation matrix on the power grid output.

[0042] Specifically, implementation rules can be formulated based on the coupling matrix and comprehensive influence coefficient. For example, if the solar irradiance is ≥800 W / m², the terrain is flat, and the topology is ring-shaped, the photovoltaic output can be set to ≥80% of the installed capacity; if the ambient temperature is >35℃ and the topology is radial, the output attenuation can be set to 5%-8%, etc., so as to adjust the grid capacity and output.

[0043] This embodiment constructs a correlation matrix with the grid-connected node ID of the photovoltaic site as the row and geographical / grid / meteorological features as the column. This efficiently integrates the results of two-dimensional correlations, breaking down the isolated barriers between geographical, grid, and meteorological correlations, thereby achieving a systematic sorting of multi-source features. By calculating the comprehensive influence coefficient, qualitative influence relationships are transformed into quantitative indicators, accurately quantifying the degree of influence of each parameter on grid output. This solves the problem of ambiguous influence weights in traditional coupling. This process makes the coupling correlation relationship more in line with the actual situation in the region, providing standardized and high-precision quantitative input for subsequent numerical forecasting models, significantly reducing forecast errors caused by correlation distortion, and laying a solid quantitative foundation for its application in photovoltaic power prediction.

[0044] S4: Construct a numerical weather prediction model, using the comprehensive influence coefficient and correlation matrix as model inputs to obtain the target forecast results.

[0045] In this embodiment, the numerical weather prediction model includes a single-model prediction layer and a comprehensive analysis layer; The single-model prediction layer obtains individual numerical weather prediction values ​​by matching the input data with the corresponding prediction model based on the model input. The comprehensive analysis layer assigns influence weights to various types of individual numerical weather prediction values ​​based on the comprehensive influence coefficient, and then performs a weighted summation of the individual numerical weather prediction values ​​based on the influence weights to obtain the target forecast result.

[0046] This embodiment uses a single-model prediction layer to match the input data with the corresponding prediction model, accurately adapting to different parameter characteristics, thereby ensuring the relevance and basic accuracy of individual numerical weather predictions. Through the comprehensive analysis layer, influence weights are allocated based on the comprehensive influence coefficient, quantifying the degree of influence of geographical, power grid, and meteorological parameters, avoiding deviations caused by single models or equal weights. This fully integrates regional characteristics and multi-source correlation patterns, and the target forecast results obtained after weighted summation are more in line with the actual scenario, effectively improving the forecast accuracy under complex conditions, and providing high-quality meteorological driving data for subsequent applications in short-term photovoltaic power prediction.

[0047] In this embodiment, the single-model prediction layer includes a geographic prediction model, a power grid topology prediction model, and a meteorological prediction model; The geographic prediction model predicts numerical weather based on geographic features in the correlation matrix; The power grid topology prediction model predicts numerical weather based on the power grid topology features in the correlation matrix; The meteorological forecasting model predicts numerical weather based on various meteorological factors in the correlation matrix.

[0048] This embodiment uses three sub-categories of prediction models—geographic, power grid topology, and meteorological—to make accurate predictions based on corresponding features in the correlation matrix. This allows for targeted capture of the impact of each dimension, adapting to differences in regional terrain, power grid structure, and meteorological patterns. This avoids the problem of insufficient adaptation of a single model to multi-source features, significantly improving the targeting and accuracy of individual predictions. It also provides high-quality basic data for the weighted summation of the comprehensive analysis layer, further strengthening the accuracy support for numerical forecasting.

[0049] Example 2: This example also provides a numerical weather prediction optimization system based on meteorological resource coupling, such as... Figure 2 As shown, it includes a data acquisition module, a data coupling module, a quantization module, and a prediction module; The data acquisition module collects meteorological and geographical data within the target area and preprocesses them to obtain the raw dataset. The data coupling module performs power grid topology coupling on geographic data to construct power grid correlations, and performs numerical forecast correlation coupling on meteorological data to construct meteorological correlations. The quantification module unifies and quantifies the power grid correlation and meteorological correlation to construct a correlation matrix and calculates the comprehensive influence coefficient; The prediction module is equipped with a numerical weather prediction model, which uses the comprehensive influence coefficient and correlation matrix as model inputs to obtain the target forecast results.

[0050] This embodiment integrates the numerical weather forecasting optimization method in this scheme by constructing a corresponding system, realizing human-computer interaction and improving the user experience.

[0051] This embodiment also provides a computer device, such as... Figure 3 As shown, it includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory is used to store computer programs; When the processor executes the program stored in memory, it implements a numerical weather forecasting optimization method based on meteorological resource coupling.

[0052] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect Standard (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0053] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0054] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0055] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0056] This application also provides a computer-readable storage medium storing a computer program, which is executed by a processor as a numerical weather forecast optimization method based on meteorological resource coupling.

[0057] As can be seen from the above embodiments, it has at least the following substantial effects: (1) This invention solves the problems of messy and spatiotemporal inconsistency in traditional forecast data by collecting and preprocessing meteorological and geographical multi-source data of the target area, and unifying the format, time series and quality standards, thus laying a solid data foundation for coupled modeling; (2) By constructing a two-dimensional correlation between the power grid and meteorology, this invention accurately captures the patterns of strongly coupled meteorological factors such as topography, power grid topology constraints and irradiance, thereby conforming to the complex characteristics of the region and avoiding redundant information interference. (3) By constructing a correlation matrix and calculating the comprehensive influence coefficient, this invention transforms qualitative coupling relationships into quantitative indicators, taking into account both linear and nonlinear relationships, thereby clarifying the contribution of each factor and solving the pain point of ambiguous influence weights in traditional methods; (4) By inputting quantitative parameters into the numerical forecasting model, the present invention enables the model to deeply integrate regional characteristics, thereby significantly improving the forecast accuracy under complex weather and terrain conditions, providing high-precision meteorological driving data for short-term photovoltaic power prediction, and thus supporting the refined scheduling of the power grid and the efficient consumption of new energy.

[0058] The specific embodiments described above are preferred embodiments of the present invention and are not intended to limit the specific scope of the present invention. The scope of the present invention includes, but is not limited to, these specific embodiments. All equivalent changes made in accordance with the shape and structure of the present invention are within the protection scope of the present invention.

Claims

1. A numerical weather prediction optimization method based on meteorological resource coupling, characterized in that: Includes the following steps: S1. Collect meteorological and geographical data within the target area and preprocess them to obtain the raw dataset; S2. Construct power grid correlation relationships by performing power grid topology coupling on geographic data; construct meteorological correlation relationships by performing numerical forecast correlation coupling on meteorological data; S3. Unify and quantify the power grid correlation and meteorological correlation to construct a correlation matrix and calculate the comprehensive influence coefficient; S4. Construct a numerical weather prediction model, using the comprehensive influence coefficient and correlation matrix as model inputs to obtain the target forecast results.

2. The numerical weather prediction optimization method based on meteorological resource coupling according to claim 1, characterized in that: In S1, meteorological and geographical data within the target area are collected and preprocessed to obtain the raw dataset, including the following steps: Information on light shading is obtained by monitoring atmospheric cloud cover and cloud velocity using ground-based cloud mapping equipment; Solar irradiance, temperature, humidity, wind speed, and air pressure are collected from ground weather stations as measured environmental information; Surface downwave radiation information was obtained by inverting cloud and aerosol parameters based on meteorological satellites and atmospheric radiative transfer models. Sunlight shading information, measured environmental information, and surface downwave radiation information are used as meteorological data, while distributed power generation data and topographic data are used as geographic data. The raw dataset is obtained by preprocessing meteorological and geographical data.

3. The numerical weather prediction optimization method based on meteorological resource coupling according to claim 2, characterized in that: In S2, the grid topology coupling of geographically related data is used to construct grid correlation relationships, including the following steps: Feature extraction is performed on distributed power generation data of the power grid to obtain power grid topology features, including topology type, topology parameters, and grid connection parameters; Feature extraction is performed on terrain data to obtain geographic features, including terrain parameters, surface parameters, and location parameters; Using the grid connection node ID of the photovoltaic site in the grid connection parameters as the grid column code, the geographical features and grid topology features corresponding to the column code are associated sequentially to construct the grid association relationship.

4. The numerical weather prediction optimization method based on meteorological resource coupling according to claim 3, characterized in that: In S2, the meteorological correlation relationship is constructed by numerical forecasting coupling of meteorological data, including the following steps: Historical power output data of the power grid under different meteorological scenarios were used as experimental samples. Based on the data types of meteorological data, corresponding meteorological factors are set, and linear coupling relationship analysis and nonlinear coupling relationship analysis are performed on various meteorological factors based on experimental samples to obtain linear coupling coefficients and nonlinear coupling coefficients. The influence relationship between meteorological data corresponding to various meteorological factors and power grid output is determined based on linear and nonlinear coupling coefficients as meteorological correlation relationships.

5. The numerical weather prediction optimization method based on meteorological resource coupling according to claim 4, characterized in that: In S3, the correlation matrix is ​​constructed by uniformly quantifying the correlation between power grid and meteorological factors, and the comprehensive influence coefficient is calculated, including the following steps: A correlation matrix is ​​constructed using the grid-connected node ID of the photovoltaic site as the row and geographical features, grid topology features, and meteorological factors as the columns; The matrix value is calculated based on the correlation matrix as the comprehensive influence coefficient, which represents the degree of influence of the parameters corresponding to the columns in the correlation matrix on the power grid output.

6. The numerical weather prediction optimization method based on meteorological resource coupling according to claim 1, characterized in that: In S4, the numerical weather prediction model includes a single-model prediction layer and a comprehensive analysis layer; The single-model prediction layer obtains individual numerical weather prediction values ​​by matching the input data with the corresponding prediction model based on the model input. The comprehensive analysis layer assigns influence weights to various types of individual numerical weather prediction values ​​based on the comprehensive influence coefficient, and then performs a weighted summation of the individual numerical weather prediction values ​​based on the influence weights to obtain the target forecast result.

7. The numerical weather prediction optimization method based on meteorological resource coupling according to claim 6, characterized in that: The single-model prediction layer includes a geographic prediction model, a power grid topology prediction model, and a meteorological prediction model; The geographic prediction model predicts numerical weather based on geographic features in the correlation matrix; The power grid topology prediction model predicts numerical weather based on the power grid topology features in the correlation matrix; The meteorological forecasting model predicts numerical weather based on various meteorological factors in the correlation matrix.

8. A numerical weather prediction optimization system based on meteorological resource coupling, applicable to the numerical weather prediction optimization method based on meteorological resource coupling correction as described in any one of claims 1-7, characterized in that: It includes a data acquisition module, a data coupling module, a quantization module, and a prediction module; The data acquisition module collects meteorological and geographical data within the target area and preprocesses them to obtain the raw dataset. The data coupling module performs power grid topology coupling on geographic data to construct power grid correlations, and performs numerical forecast correlation coupling on meteorological data to construct meteorological correlations. The quantification module unifies and quantifies the power grid correlation and meteorological correlation to construct a correlation matrix and calculates the comprehensive influence coefficient; The prediction module is equipped with a numerical weather prediction model, which uses the comprehensive influence coefficient and correlation matrix as model inputs to obtain the target forecast results.

9. A computer device, characterized in that: It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; the memory is used to store computer programs; and the processor, when executing the program stored in the memory, implements the steps of the numerical weather prediction optimization method based on meteorological resource coupling as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the numerical weather prediction optimization method based on meteorological resource coupling as described in any one of claims 1-7.

Images