Layout method of distributed photovoltaic meteorological monitoring stations and related device

Abstract

The invention discloses a layout method of distributed photovoltaic meteorological monitoring stations and a related device, and belongs to the technical field of new energy power generation. The method comprises the steps that multi-source heterogeneous data are collected and preprocessed; constructing a spatio-temporal characteristic spectrum according to the preprocessed multi-source heterogeneous data; establishing a multi-target dynamic game optimization model based on the spatio-temporal characteristic spectrum; and solving the multi-target dynamic game optimization model to obtain the position layout of the distributed photovoltaic meteorological monitoring stations. According to the method, a multi-target dynamic game mechanism is established, conflict between coverage rate and cost is automatically balanced under real-time constraint, rigid uniform distribution or single-target optimization is replaced, a Pareto optimal strategy is generated through a dynamic game model (coverage range, cost and monitoring precision), and resource allocation flexibility is realized.

Description

Technical Field

[0001] This invention belongs to the field of new energy power generation technology, and relates to a layout method and related devices for distributed photovoltaic meteorological monitoring stations. Background Technology

[0002] In recent years, the new energy industry has developed rapidly, with photovoltaic power generation becoming a core growth engine. As an important component of clean energy, distributed photovoltaic power generation is crucial for grid dispatch and management due to its power generation efficiency and stability.

[0003] The fragmented nature of distributed photovoltaic (PV) power generation presents new challenges to meteorological monitoring: First, data representativeness is insufficient. PV module efficiency is dynamically affected by micro-meteorological parameters such as irradiance, temperature, and wind speed, and uniform grid deployment does not consider the distribution density of PV power plants and the impact of terrain on meteorological parameters. Second, dynamic adaptability is poor. Traditional layout methods cannot respond to the needs of PV power plant expansion and equipment upgrades, and data standards among meteorological, power grid, and PV companies are not unified. Most sites have not achieved multi-source data fusion, failing to support the demand for high-frequency, accurate meteorological data services such as safe operation and maintenance of power generation equipment and ultra-short-term power forecasting. Third, there is an economic imbalance. High-density deployment of meteorological observation stations leads to a surge in costs, while low-density deployment makes it difficult to capture key meteorological changes. However, since meteorological conditions (such as cloud dynamics and irradiance) have a significant impact on PV power output, accurate meteorological monitoring and forecasting are crucial to improving the accuracy of PV power generation prediction. Therefore, optimizing the layout of observation stations for distributed PV meteorological monitoring and forecasting to more accurately capture meteorological characteristics has become an urgent technical problem to be solved.

[0004] Currently, the layout of observation stations mainly employs traditional meteorological methods, statistical models, and static optimization models. Traditional meteorological methods plan station placement based solely on historical climate data, failing to fully consider the characteristics and needs of distributed photovoltaic (PV) power plants. Statistical model methods primarily rely on fixed station number thresholds, unable to quantify the nonlinear impact of station additions or subtractions on forecast accuracy. Static optimization models lack the ability to respond to dynamic changes in PV power plants, leading to a mismatch between station layout and real-time needs. This results in observational data failing to meet the requirements for distributed PV meteorological monitoring and forecasting in terms of spatiotemporal resolution and accuracy. Furthermore, the existing station layout may lack targeted optimization for distributed PV power plants, leading to insufficient data representativeness and affecting the accuracy and generalization ability of prediction models. Summary of the Invention

[0005] The purpose of this invention is to provide a layout method and related apparatus for distributed photovoltaic meteorological monitoring stations, so as to solve the technical problem that the layout of existing stations is difficult to meet the observation requirements.

[0006] To achieve the above objectives, the present invention employs the following technical solution:

[0007] In a first aspect, the present invention provides a method for laying out distributed photovoltaic meteorological monitoring stations, comprising the following steps:

[0008] Collect heterogeneous data from multiple sources and perform preprocessing;

[0009] Based on the preprocessed multi-source heterogeneous data, a spatiotemporal feature map is constructed;

[0010] A multi-objective dynamic game optimization model is established based on spatiotemporal feature maps;

[0011] The location layout of distributed photovoltaic meteorological monitoring stations is obtained by solving a multi-objective dynamic game optimization model.

[0012] Furthermore, the step of collecting multi-source heterogeneous data and performing preprocessing specifically includes:

[0013] Collect multi-source heterogeneous data, including satellite remote sensing data, meteorological data, photovoltaic engineering parameters, and terrain parameters;

[0014] The collected data undergoes quality control and missing value handling, and spatial benchmarks and time series are unified to complete data preprocessing.

[0015] Furthermore, the step of constructing a spatiotemporal feature map based on the preprocessed multi-source heterogeneous data specifically includes:

[0016] Data fusion is performed on the preprocessed multi-source heterogeneous data. The specific fusion formula is as follows:

[0017] F(x,y,t)=α·AOD_norm+β·Cloud_norm+γ·PV_density×(1-slope / 30°)

[0018] In the formula, α, β, and γ are constants; AOD_norm is satellite remote sensing data; Cloud_norm is meteorological data; PV_density is photovoltaic engineering parameters; and slope is terrain parameters.

[0019] The meteorological sensitivity is calculated using the sliding window method, and the specific calculation formula is as follows:

[0020] Sensitivity = (AOD variance × Cloud gradient) / (Terrain roughness + ε)

[0021] In the formula, Sensitivity represents meteorological sensitivity; ε is a constant.

[0022] OPTICS density clustering is performed on regions with meteorological sensitivity greater than the threshold to output highly sensitive sub-regions and obtain spatiotemporal feature maps.

[0023] Furthermore, the steps for establishing the multi-objective dynamic game optimization model specifically include:

[0024] Define the set of site locations and the set of mobile station scheduling variables:

[0025] S={s_i|s_i=(x_i,y_i,z_i),z_i∈{0,1}}

[0026] M={m_j|m_j=(x_j^t,y_j^t),t∈T}

[0027] In the formula, S represents the set of station locations; s_i = (x_i, y_i, z_i) represents the coordinates of a single station location; M represents the set of mobile station scheduling variables; T is the scheduling time window; m_j = (x_j^t, y_j^t) represents the coordinates of a single mobile station;

[0028] A multi-objective dynamic game optimization model is established, with three objective functions: maximizing meteorological coverage, minimizing photovoltaic associated costs, and optimizing terrain adaptability. Hard and soft constraints are set for the model.

[0029] Furthermore, the objective function of the multi-objective dynamic game optimization model is:

[0030] f_1=∑_{k=1}^Kw_k·tanh(∑_{i=1}^Nz_i·e^{-d_{ik} / D})

[0031] f_2=∑_{p=1}^PC_p·min(d_{ip})+λ·N

[0032] f_3=1 / N·∑_{i=1}^Nz_i·(1-slope_i / 25°)·exp(-|h_i-h_{avg}| / 200m)

[0033] In the formula, f_1 is the objective function for maximizing meteorological coverage; f_2 is the objective function for minimizing photovoltaic associated costs; f_3 is the objective function for optimizing terrain adaptability; w_k is the weight of the k-th highly sensitive sub-region; D is the resolution; C_p is the installed capacity of power station p; and λ is the construction cost of a single station.

[0034] The hard constraints of the model are: the distance between adjacent stations is ≥2D; stations are prohibited in areas with a slope >25°.

[0035] The soft constraint of the model is that each highly sensitive sub-region must be covered by at least one site.

[0036] Furthermore, the step of solving the multi-objective dynamic game optimization model to obtain the location layout of distributed photovoltaic meteorological monitoring stations specifically includes:

[0037] An improved NSGA-III algorithm is used to generate an initial population within the buffer zone of a photovoltaic power plant.

[0038] A targeted crossover strategy is adopted, prioritizing the retention of individuals covering highly sensitive areas when selecting the parent generation, and setting the crossover point in the dimension with the largest meteorological sensitivity gradient.

[0039] Adaptive mutation is implemented, and the mutation probability increases with the number of iterations;

[0040] The Pareto front solution is partitioned using a Voronoi diagram, and mobile stations are inserted into the coverage hole regions to form new solutions.

[0041] Finally, gradient-guided optimization was used to adjust the site location:

[0042] Furthermore, the method also includes: deploying a dynamic feedback mechanism to achieve dynamic adjustment of the site layout through digital twin verification and online learning optimization;

[0043] The digital twin verification includes: simulating extreme events in a virtual weather field to evaluate the station data capture rate and the improvement in early warning timeliness;

[0044] The online learning optimization includes: dynamically adjusting the objective function weights based on monitoring errors; and scheduling mobile stations to form a monitoring triangular network when the hourly change rate of regional cloud exceeds a threshold.

[0045] Secondly, the present invention provides a layout system for distributed photovoltaic meteorological monitoring stations, comprising:

[0046] The data acquisition module is used to collect multi-source heterogeneous data and perform preprocessing.

[0047] The feature map construction module is used to construct spatiotemporal feature maps based on preprocessed multi-source heterogeneous data;

[0048] The model building module is used to establish a multi-objective dynamic game optimization model based on spatiotemporal feature maps;

[0049] The layout solution module is used to solve a multi-objective dynamic game optimization model to obtain the location layout of distributed photovoltaic meteorological monitoring stations.

[0050] Thirdly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for the layout of a distributed photovoltaic meteorological monitoring station.

[0051] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method for the layout of a distributed photovoltaic meteorological monitoring station.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] This invention discloses a layout method and related devices for distributed photovoltaic meteorological monitoring stations. Traditional methods are mostly based on static meteorological data or uniform distribution strategies, which are difficult to adapt to the spatiotemporal heterogeneity of solar radiation fluctuations, terrain complexity, and meteorological changes, and resource allocation is rigid (e.g., insufficient coverage in high-sensitivity areas and redundant construction in low-sensitivity areas). This invention integrates multi-source dynamic data (satellite remote sensing, terrain, historical meteorology) through spatiotemporal feature maps to accurately identify sensitive areas; combined with a dynamic game optimization model, it balances the conflicts of multiple objectives such as coverage, cost, and monitoring accuracy in real time, realizing intelligent dynamic adjustment of station layout. Experiments show that, under the same cost, this scheme can improve the monitoring coverage of high-sensitivity areas (such as areas with frequent cloud cover and rugged terrain), while reducing the risk of monitoring blind spots caused by sudden environmental changes through dynamic feedback mechanisms (such as real-time correction by edge computing). This invention establishes a multi-objective dynamic game mechanism to automatically balance the conflict between coverage and cost under real-time constraints, replacing rigid uniform distribution or single-objective optimization. Through a dynamic game model (coverage, cost, monitoring accuracy), it generates Pareto optimal strategies, achieving flexibility in resource allocation. Attached Figure Description

[0054] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 This is a flowchart of the method of the present invention;

[0056] Figure 2 This is a schematic diagram of the system of the present invention;

[0057] Figure 3 This is a schematic diagram of the computer device structure of the present invention. Detailed Implementation

[0058] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0059] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used in this invention is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0060] See Figure 1 This invention discloses a method for laying out distributed photovoltaic meteorological monitoring stations, comprising the following steps:

[0061] S1: Collect multi-source heterogeneous data and perform preprocessing;

[0062] Collect multi-source heterogeneous data. The sources and types of the multi-source heterogeneous data are as follows:

[0063] 1. Satellite remote sensing data

[0064] Source: MODIS AOD (1km resolution, 1h time step)

[0065] 2. Meteorological data

[0066] Source: China Meteorological Administration Land Surface Data Assimilation System (CLDAS) - Temperature, Precipitation, Relative Humidity (1km resolution, 1h time step)

[0067] ART-3DCloudA V2.0, a 1km / 1h 3D cloud real-time analysis product for the China region (5km resolution, 1h time step).

[0068] 3. Photovoltaic Project Parameters

[0069] Source: Photovoltaic power plant coordinates, installed capacity, expansion plans

[0070] 4. Terrain parameters

[0071] Source: DEM elevation model (accuracy 10m)

[0072] The collected data underwent quality control and missing value processing, spatial benchmark unification and coordinate system transformation, time series alignment, and multimodal data association coding, as shown in Table 1 below.

[0073] Table 1

[0074] Data layer Spatial accuracy Time resolution coordinate system Numerical range MODIS AOD 1km 1h CGCS2000 0-2 (dimensionless) Cloud 1km 1h CGCS2000 0-1 (dimensionless) DEM-derived features 1km static CGCS2000 Standardized value Photovoltaic dynamic parameters Site-level Annual update CGCS2000 actual physical quantity

[0075] S2, construct a spatiotemporal feature map based on the preprocessed multi-source heterogeneous data;

[0076] Data fusion is performed on the preprocessed multi-source heterogeneous data. The specific fusion formula is as follows:

[0077] F(x,y,t)=α·AOD_norm+β·Cloud_norm+γ·PV_density×(1-slope / 30°)

[0078] In the formula, α, β, and γ are all constants (α = 0.6, β = 0.3, γ = 0.1, calibrated by ridge regression); AOD_norm is satellite remote sensing data; Cloud_norm is meteorological data; PV_density is photovoltaic engineering parameters; slope is topographic parameters;

[0079] The meteorological sensitivity is calculated using the sliding window method, and the specific calculation formula is as follows:

[0080] Sensitivity = (AOD variance × Cloud gradient) / (Terrain roughness + ε)

[0081] In the formula, Sensitivity is the meteorological sensitivity; ε is a constant (ε = 0.01 to prevent division by zero);

[0082] OPTICS density clustering is performed on regions with meteorological sensitivity greater than the threshold (Sensitivity>0.8) to output highly sensitive sub-regions and obtain spatiotemporal feature maps.

[0083] S3, based on spatiotemporal feature maps, establishes a multi-objective dynamic game optimization model;

[0084] Define the set of site locations and the set of mobile station scheduling variables:

[0085] S={s_i|s_i=(x_i,y_i,z_i),z_i∈{0,1}}

[0086] M={m_j|m_j=(x_j^t,y_j^t),t∈T}

[0087] In the formula, S represents the set of station locations; s_i = (x_i, y_i, z_i) represents the coordinates of a single station location; M represents the set of mobile station scheduling variables; T is the scheduling time window; m_j = (x_j^t, y_j^t) represents the coordinates of a single mobile station;

[0088] A multi-objective dynamic game optimization model is established, with three objective functions: maximizing meteorological coverage, minimizing photovoltaic associated costs, and optimizing terrain adaptability. Hard and soft constraints are set for the model.

[0089] The objective function of the multi-objective dynamic game optimization model is:

[0090] f_1=∑_{k=1}^Kw_k·tanh(∑_{i=1}^Nz_i·e^{-d_{ik} / D})

[0091] f_2=∑_{p=1}^PC_p·min(d_{ip})+λ·N

[0092] f_3=1 / N·∑_{i=1}^Nz_i·(1-slope_i / 25°)·exp(-|h_i-h_{avg}| / 200m)

[0093] In the formula, f_1 is the objective function for maximizing meteorological coverage; f_2 is the objective function for minimizing photovoltaic associated costs; f_3 is the objective function for optimizing terrain adaptability; w_k is the weight of the k-th highly sensitive sub-region; D is the resolution, D=5km; C_p is the installed capacity of power station p; λ is the construction cost of a single station.

[0094] The hard constraints of the model are: the distance between adjacent stations is ≥2D; stations are prohibited in areas with a slope >25°.

[0095] The soft constraint of the model is that each highly sensitive sub-region must be covered by at least one site.

[0096] S4, the location layout of distributed photovoltaic meteorological monitoring stations is obtained by solving the multi-objective dynamic game optimization model.

[0097] ① Improved NSGA-III algorithm

[0098] Population initialization: Generating initial solutions within a 5km buffer zone of the photovoltaic power station

[0099] Targeted crossover: When selecting the parent generation, individuals covering highly sensitive areas are prioritized, and the crossover point is set on the dimension with the largest meteorological sensitivity gradient.

[0100] Adaptive mutation:

[0101] P_mut = 0.2 + 0.1 * (t / T_max)

[0102] σ = 0.5D·(1 + terrain complexity index)

[0103] ② Local search enhancement

[0104] Voronoi-GA hybrid strategy: partition the Pareto front solution into a Voronoi graph; insert mobile stations in the coverage void area to form a new solution.

[0105] Gradient-guided optimization:

[0106]

[0107] (η = 0.1D, maximum iterations 50)

[0108] S5 deploys a dynamic feedback mechanism to dynamically adjust the site layout through digital twin verification and online learning optimization;

[0109] ① Digital twin verification

[0110] Simulate extreme events (dust storms, thunderstorms) in a virtual weather field:

[0111] Input: Historical event dataset (time resolution 15 minutes)

[0112] Outputs: Site data capture rate, improvement in early warning timeliness

[0113] Verification criteria:

[0114] Capture rate ≥ 90% and timeliness improvement ≥ 50%

[0115] ② Online learning optimization

[0116] Weights are dynamically adjusted:

[0117] Δα=0.1·(MAE_current-MAE_target) / MAE_target

[0118] Mobile station scheduling rules: When the hourly change rate of Cloud in a certain area is >50%, the three most recent mobile stations are scheduled to form a monitoring triangular network.

[0119] See Figure 2 This invention discloses a layout system for distributed photovoltaic meteorological monitoring stations, including a data acquisition module, a feature map construction module, a model construction module, and a layout solution module.

[0120] The data acquisition module is used to collect multi-source heterogeneous data and perform preprocessing.

[0121] The feature map construction module is used to construct spatiotemporal feature maps based on preprocessed multi-source heterogeneous data. In this step, dynamic relationship modeling based on graph neural networks can also be used, replacing traditional map construction methods with a spatiotemporal graph neural network (STGNN). This automatically extracts spatiotemporal correlation features of meteorology, topography, and illumination through end-to-end learning. This technology can reduce reliance on manually designed feature maps and improve generalization ability for unknown areas, but it reduces model interpretability and may affect the reliability of the layout strategy. Bayesian networks are used to model the causal relationship between meteorological sensitivity and station coverage, dynamically adjusting priorities through probabilistic inference and quantifying uncertainties (such as cloud movement probability) to support risk-sensitive decision-making. However, accurate modeling of conditional probabilities between variables is required, and data sparsity in real-world scenarios may lead to errors.

[0122] The model building module is used to establish a multi-objective dynamic game optimization model based on spatiotemporal feature maps. In this step, deep reinforcement learning (DRL) can also be used instead of the dynamic game model. Through interactive training between environmental states (such as meteorological data and station layout) and actions (such as adding or removing stations and adjusting their locations), the layout strategy can be autonomously optimized. This technology does not require pre-setting game rules and can automatically explore the equilibrium point between complex objectives. Its disadvantages are that it requires a large amount of historical data during the training phase, and the stability of real-time decisions may be insufficient.

[0123] The layout solution module is used to solve a multi-objective dynamic game optimization model to obtain the location layout of distributed photovoltaic meteorological monitoring stations. Alternatively, evolutionary algorithms such as the National Natural Science Foundation of China (NSGA-II) can be used to directly perform Pareto front search on objectives such as coverage, cost, and monitoring accuracy, replacing the strategy game process in game theory. This technique avoids the complex equilibrium solutions in game models and is suitable for multi-objective nonlinear optimization. Its drawbacks include high computational complexity and weak real-time update capability in dynamic environments.

[0124] Specifically, such as Figure 1 As shown, the system of the present invention comprises:

[0125] ① Data Fusion Layer: Construct a spatiotemporal feature map of AOD / Cloud-PV demand-terrain.

[0126] ② Intelligent decision-making layer: Design a multi-objective game optimization model and a hybrid solution algorithm.

[0127] ③ Dynamic execution layer: Deploy mobile monitoring stations and online learning feedback mechanisms.

[0128] Data flow:

[0129] ① Data preprocessing → ② Meteorological sensitivity → ③ Highly sensitive sub-regions → ④ Multi-objective game optimization model → ⑤ Hybrid solution algorithm design

[0130] In one embodiment of the invention, see [link to embodiment]. Figure 3A computer device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, 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, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions from the computer storage medium to achieve a corresponding method flow or function. The processor described in this embodiment can be used in the operation of a distributed photovoltaic meteorological monitoring station layout method.

[0131] This invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the layout method for a distributed photovoltaic meteorological monitoring station in the above embodiments.

[0132] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0133] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0134] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0135] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for laying out distributed photovoltaic meteorological monitoring stations, characterized in that, Includes the following steps: Collect heterogeneous data from multiple sources and perform preprocessing; Based on the preprocessed multi-source heterogeneous data, a spatiotemporal feature map is constructed; A multi-objective dynamic game optimization model is established based on spatiotemporal feature maps; The location layout of distributed photovoltaic meteorological monitoring stations is obtained by solving a multi-objective dynamic game optimization model.

2. The method for laying out distributed photovoltaic meteorological monitoring stations according to claim 1, characterized in that, The steps of collecting multi-source heterogeneous data and performing preprocessing specifically include: Collect multi-source heterogeneous data, including satellite remote sensing data, meteorological data, photovoltaic engineering parameters, and terrain parameters; The collected data undergoes quality control and missing value handling, and spatial benchmarks and time series are unified to complete data preprocessing.

3. The layout method of a distributed photovoltaic meteorological monitoring station according to claim 1, characterized in that, The step of constructing a spatiotemporal feature map based on the preprocessed multi-source heterogeneous data specifically includes: Data fusion is performed on the preprocessed multi-source heterogeneous data. The specific fusion formula is as follows: F(x,y,t)=α·AOD_norm+β·Cloud_norm+γ·PV_density×(1-slope / 30°) In the formula, α, β, and γ are constants; AOD_norm is satellite remote sensing data; Cloud_norm is meteorological data; PV_density is photovoltaic engineering parameters; and slope is terrain parameters. The meteorological sensitivity is calculated using the sliding window method, and the specific calculation formula is as follows: Sensitivity = (AOD variance × Cloud gradient) / (Terrain roughness + ε) In the formula, Sensitivity represents meteorological sensitivity; ε is a constant. OPTICS density clustering is performed on regions with meteorological sensitivity greater than the threshold to output highly sensitive sub-regions and obtain spatiotemporal feature maps.

4. The layout method of a distributed photovoltaic meteorological monitoring station according to claim 1, characterized in that, The steps for establishing a multi-objective dynamic game optimization model specifically include: Define the set of site locations and the set of mobile station scheduling variables: S={s_i|s_i=(x_i,y_i,z_i),z_i∈{0,1}} M={m_j|m_j=(x_j^t,y_j^t),t∈T} In the formula, S represents the set of station locations; s_i = (x_i, y_i, z_i) represents the coordinates of a single station location; M represents the set of mobile station scheduling variables; T is the scheduling time window; m_j = (x_j^t, y_j^t) represents the coordinates of a single mobile station; A multi-objective dynamic game optimization model is established, with three objective functions: maximizing meteorological coverage, minimizing photovoltaic associated costs, and optimizing terrain adaptability. Hard and soft constraints are set for the model.

5. The layout method of a distributed photovoltaic meteorological monitoring station according to claim 4, characterized in that, The objective function of the multi-objective dynamic game optimization model is: f_1=∑_{k=1}^Kw_k·tanh(∑_{i=1}^Nz_i·e^{-d_{ik} / D}) f_2=∑_{p=1}^PC_p·min(d_{ip})+λ·N f_3=1 / N·∑_{i=1}^Nz_i·(1-slope_i / 25°)·exp(-|h_i-h_{avg}| / 200m) In the formula, f_1 is the objective function for maximizing meteorological coverage; f_2 is the objective function for minimizing photovoltaic associated costs; f_3 is the objective function for optimizing terrain adaptability; w_k is the weight of the k-th highly sensitive sub-region; D is the resolution; C_p is the installed capacity of power station p; and λ is the construction cost of a single station. The hard constraints of the model are: the distance between adjacent stations is ≥2D; stations are prohibited in areas with a slope >25°. The soft constraint of the model is that each highly sensitive sub-region must be covered by at least one site.

6. The method for laying out distributed photovoltaic meteorological monitoring stations according to claim 1, characterized in that, The steps for obtaining the location layout of distributed photovoltaic meteorological monitoring stations by solving the multi-objective dynamic game optimization model specifically include: An improved NSGA-III algorithm is used to generate an initial population within the buffer zone of a photovoltaic power plant. A targeted crossover strategy is adopted, prioritizing the retention of individuals covering highly sensitive areas when selecting the parent generation, and setting the crossover point in the dimension with the largest meteorological sensitivity gradient. Adaptive mutation is implemented, and the mutation probability increases with the number of iterations; The Pareto front solution is partitioned using a Voronoi diagram, and mobile stations are inserted into the coverage hole regions to form new solutions. Finally, gradient-guided optimization was used to adjust the site location:

7. The method for laying out distributed photovoltaic meteorological monitoring stations according to claim 1, characterized in that, It also includes: deploying a dynamic feedback mechanism to achieve dynamic adjustments to the site layout through digital twin verification and online learning optimization; The digital twin verification includes: simulating extreme events in a virtual weather field to evaluate the station data capture rate and the improvement in early warning timeliness; The online learning optimization includes: dynamically adjusting the objective function weights based on monitoring errors; and scheduling mobile stations to form a monitoring triangular network when the hourly change rate of regional cloud exceeds a threshold.

8. A layout system for distributed photovoltaic meteorological monitoring stations, characterized in that, include: The data acquisition module is used to collect multi-source heterogeneous data and perform preprocessing. The feature map construction module is used to construct spatiotemporal feature maps based on preprocessed multi-source heterogeneous data; The model building module is used to establish a multi-objective dynamic game optimization model based on spatiotemporal feature maps; The layout solution module is used to solve a multi-objective dynamic game optimization model to obtain the location layout of distributed photovoltaic meteorological monitoring stations.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the layout method for a distributed photovoltaic meteorological monitoring station as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the layout method of a distributed photovoltaic meteorological monitoring station as described in any one of claims 1-7.

Images