Multi-factor centralized photovoltaic power station site selection optimization method

Abstract

The invention relates to the technical field of renewable energy infrastructure planning, in particular to a centralized photovoltaic power station site selection optimization method under multiple factors. Comprising the steps of collecting and processing multi-source geographic space data including meteorological conditions and climate disaster risk data; rejecting unsuitable areas through constraint rejection analysis; performing normalization and weighted aggregation on the meteorological conditions and the climate disaster risk data, and constructing a climate suitability index layer; establishing a multi-standard evaluation framework, determining each criterion weight by adopting a multi-standard decision algorithm, and generating a preliminary site power generation potential map through weighted stacking; correcting the power generation potential map by adopting a bivariate local Moran index analysis and geographically weighted regression model; and identifying and dividing an optimal candidate site on the corrected power generation potential map. According to the method, meteorological conditions and climate disaster risks are quantified and integrated, and the evaluation result is corrected by using the spatial statistical model, so that the scientificity and reliability of site selection decision are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of renewable energy infrastructure planning technology, specifically to a method for optimizing the site selection of centralized photovoltaic power plants under multiple factors. Background Technology

[0002] Centralized photovoltaic (PV) power plants, as a key form of large-scale solar energy utilization, rely heavily on the site selection decisions for PV panel deployment, which significantly impact the project's economic benefits, long-term stable operation, and environmental impact. Currently, the site selection process for PV power plants generally relies on a combination of Geographic Information Systems (GIS) and Multi-Criterion Decision Analysis (MCDA). GIS, as a powerful spatial data management and analysis platform, can layer, overlay, and analyze various geographic reference data, including climate, topography, land use, and infrastructure.

[0003] Among various MCDA techniques, the Analytic Hierarchy Process (AHP) is the most widely used method. AHP decomposes complex problems into multiple levels, such as objectives, criteria, and alternatives, and systematically assigns weights to each criterion using pairwise comparisons to achieve decision-making. In AHP-based models, evaluation criteria have formed a relatively mature system, typically including:

[0004] Photovoltaic resource potential: mainly refers to the total horizontal irradiance, which is the basis for determining the power generation capacity of a power plant.

[0005] Topographical constraints: Slope and topographic relief are key factors. Excessive slope and topographic relief will significantly increase construction costs and construction and maintenance difficulties. At the same time, areas with a north-facing slope have poor radiation resources.

[0006] Land use: By analyzing land cover types, we can avoid areas such as nature reserves, urban built-up areas, and high-value farmland.

[0007] Climate suitability: This mainly considers the climate conditions of the photovoltaic power station area, such as temperature, precipitation, relative humidity, cloud cover, etc.

[0008] However, these AHP-based models rely on static and historical data for optimization. These methods assess site suitability based on past averages (such as annual average solar radiation and existing land use). They fail to systematically and quantitatively integrate forward-looking, dynamic risks, particularly natural disaster risks related to climate change. Even when existing models consider "risk," they often use simplified static factors (such as seismic zone delineation) without adequately accounting for the changing trends in the frequency and intensity of disasters such as floods, wildfires, extreme winds, and droughts over the next few decades. These dynamic risks pose a serious long-term threat to the long-lifecycle infrastructure investment of photovoltaic power plants.

[0009] Based on the above analysis, it can be seen that existing technologies generally adopt a "static constraint optimization" paradigm, that is, to find the optimal solution under the current unchanged conditions. This paradigm ignores the dynamic environmental evolution brought about by climate change, which means that today's "optimal" site may become a high-risk, high-cost, or even unoperational area in the future. Summary of the Invention

[0010] This invention provides a method for optimizing the site selection and photovoltaic panel deployment of centralized photovoltaic power plants under various factors. This method overcomes the shortcomings of existing technologies that rely solely on static data and cannot assess long-term operational risks by integrating a comprehensive and forward-looking climate suitability assessment into a multi-criteria decision-making framework.

[0011] A method for optimizing the site selection and photovoltaic panel deployment of a centralized photovoltaic power plant under multiple factors includes the following steps:

[0012] S1. Data Acquisition and Preprocessing: Acquire and process multi-source geospatial data within the study area using a unified coordinate system and raster size. The data includes: topographic data, land use data, and meteorological conditions and climate disaster risk data.

[0013] S2. Constraint Exclusion Analysis: Based on a set of preset absolute constraints, including legal nature reserves, water bodies, urban areas, and thresholds for slope, aspect, and topographic relief, a binary exclusion zone layer is identified and generated within the study area to shield unsuitable areas and obtain areas that are technically developable and constructible.

[0014] S3. Climate Suitability Layer Construction: Based on the meteorological conditions and the types of climate disaster risks, multiple climate layers are constructed and normalized. Then, the multiple climate layers are aggregated by weighted linear combination to construct a climate suitability index layer.

[0015] S4. Evaluation Framework Construction and Weighting: Establish a hierarchical evaluation framework that includes photovoltaic resource potential and engineering feasibility, and use the analytic hierarchy process (AHP) to determine the weight of each criterion by constructing pairwise comparison matrices and calculating their principal eigenvectors.

[0016] S5. Comprehensive assessment of power generation potential: All criterion layers are weighted and superimposed according to the determined weights, and the power generation potential score of each grid cell is calculated to generate the final site power generation potential map covering the study area.

[0017] S6. Spatial correlation analysis and optimization: Perform spatial statistical analysis on the final site power generation potential map, identify the spatial pattern, and revise the final site power generation potential map;

[0018] S7. Optimal Site Identification and Division: On the corrected final site power generation potential map, based on the preset target installed capacity and the land area required per megawatt, the region growth algorithm is used to merge adjacent high-scoring units starting from the unit with the highest suitability score until a coherent plot that meets the area requirements is formed, and this plot is output as the optimal candidate site.

[0019] Preferably, the topographic data includes slope, aspect, and topographic relief; meteorological conditions include total horizontal irradiance, average annual temperature, annual precipitation, total annual cloud cover, and average annual relative humidity; climate disaster risk data includes frost risk, strong wind risk, snow accumulation risk, high temperature risk, and lightning disaster risk; and land use data includes nature reserves and land use types.

[0020] Preferably, the steps for constructing the climate suitability index layer include: normalizing the data values ​​of each independent climate layer to a general scale of 0-1 using a linear function, and then aggregating the normalized climate layers by weighted linear combination.

[0021] Preferably, the analytic hierarchy process used in step S4 further includes: calculating a consistency ratio to verify the logical consistency of the pairwise comparison matrix, wherein the consistency ratio is less than a preset ratio threshold.

[0022] Preferably, the spatial statistical analysis performed in step S6 is a bivariate local Moran index analysis.

[0023] Preferably, the bivariate local Moran index analysis specifically includes:

[0024] The final site power generation potential map generated in step S5 is used as the first variable, and the climate suitability index map generated in step S3 is used as the second variable.

[0025] For each grid cell within the study area, a local Moran index is calculated. This index measures the spatial covariance between the value of the first variable in that cell and the spatial lag value of the second variable in the neighboring region.

[0026] Based on the calculation results and statistical significance of the local Moran index, the local spatial correlation pattern of each unit is divided into one of four types: high power generation potential-high suitability, high power generation potential-low suitability, low power generation potential-high suitability, and low power generation potential-high suitability.

[0027] Preferably, the spatial statistical analysis performed in step S6 is a geographically weighted regression analysis.

[0028] Preferably, the geographically weighted regression analysis specifically includes:

[0029] Establish a regression model with power generation potential as the dependent variable and various evaluation criteria as independent variables;

[0030] For each grid cell within the study area, an independent local regression equation is fitted.

[0031] Generate a series of local regression coefficient layers, where each layer shows the spatial distribution of the influence of an independent variable on the dependent variable;

[0032] Based on the magnitude and statistical significance of the coefficients in the local regression coefficient layer, the final site power generation potential map is revised.

[0033] Compared with the prior art, the advantages of this invention are:

[0034] By comprehensively considering meteorological factors affecting the power generation efficiency and safe operation of photovoltaic facilities, and conducting statistical analysis in a yearly manner, regional meteorological conditions and climate risk status can be more accurately reflected, thus avoiding meteorological risks in later operation.

[0035] By integrating various land-related constraints associated with photovoltaic project development, the technical feasibility of site selection can be improved.

[0036] Bivariate local autocorrelation analysis can effectively reflect spatial synergistic relationships. Combined with significance characteristics, it is more conducive to identifying regions with good climate suitability and high photovoltaic power generation potential, which is beneficial to maximizing photovoltaic resources while minimizing climate risks. Attached Figure Description

[0037] Figure 1 This is a flowchart of a multi-factor site selection optimization method for centralized photovoltaic power plants proposed in this invention. Detailed Implementation

[0038] 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 merely illustrative and not intended to limit the invention.

[0039] This method is implemented on a computer system that includes at least a processor, memory, and a non-transitory computer-readable storage medium. The storage medium contains instructions that, when executed by the processor, complete the steps described in this invention. The system integrates a GIS engine (such as an ArcGIS or QGIS-based library), a data processing module, and an MCDA calculation module.

[0040] Example 1:

[0041] refer to Figure 1This embodiment provides a detailed description of the proposed method for optimizing photovoltaic panel deployment in a centralized photovoltaic power plant under multiple factors:

[0042] S1. Multi-source data acquisition and preprocessing:

[0043] First, for a predefined research area, multiple digital datasets are collected. These datasets include:

[0044] Topographic data: Digital elevation model (DEM) used to derive slope, aspect, and topographic relief layers.

[0045] Land use / land cover (LULC) data: LULC maps derived from satellite remote sensing or government departments are used to identify woodlands, water bodies, towns, farmland, etc.

[0046] Reserve data: Vector data defining national parks, nature reserves, and other legally restricted areas.

[0047] Meteorological data: total horizontal irradiance, annual average temperature, annual precipitation, annual total cloud cover, annual average relative humidity, etc.

[0048] Climate disaster risk data: This includes explicit spatial data representing future or probabilistic risks, such as frost risk, strong wind risk, snow accumulation risk, high temperature risk, and lightning disaster risk data.

[0049] It should be noted that all data were reprojected to a unified coordinate system and rasterized to a uniform cell size (e.g., 30m × 30m) to facilitate subsequent spatial analysis.

[0050] S2. Constraint Analysis and Exclusion Region Generation:

[0051] This step aims to identify areas where photovoltaic power plant construction is absolutely impossible or prohibited. Specifically, based on a set of preset absolute constraints, a binary exclusion region layer is identified and generated within the study area. These areas where photovoltaic power plant construction is impossible or prohibited include:

[0052] Legally protected areas (such as national parks).

[0053] Water bodies.

[0054] Towns and built-up areas.

[0055] The area with a north-facing slope.

[0056] Areas where the slope exceeds a preset threshold (e.g., greater than 60 degrees).

[0057] Understandably, areas with slopes exceeding the preset threshold are impractical and too costly to construct on, and are therefore excluded.

[0058] S3, Construction of Climate Suitability Index Layer:

[0059] Normalization: Normalize the data values ​​of each individual climate layer to a common scale (e.g., 0 to 1), where 1 represents the most suitable.

[0060] Weighting and Aggregation: A weighted linear combination method is used to aggregate the normalized climate layers. Weights can be allocated based on expert opinion (e.g., determined using the AHP method) or based on the potential economic losses to photovoltaic power plants caused by different meteorological conditions and climate disaster types. The calculation formula is as follows:

[0061]

[0062] Where N is the number of risk layers, R i w is the quantization index of the i-th climate layer data value. i This is the corresponding weight value.

[0063] The output of this step is a single raster layer, where the value of each cell represents the climate suitability index for that location.

[0064] S4. Assessment Framework Construction and Weighted Assessment of Power Generation Potential:

[0065] The core of this step is modeling based on expert knowledge, which involves constructing a hierarchical evaluation framework and using a multi-criteria decision-making algorithm to determine weights. In this embodiment, the Analytic Hierarchy Process (AHP) is preferred.

[0066] Weight determination (AHP method):

[0067] Domain experts were invited to conduct pairwise comparisons of each main criterion and its sub-criterions, using a 1-9 scale for scoring. The results were used to construct a comparison matrix.

[0068] In this embodiment, the main criteria are photovoltaic resource potential and engineering feasibility. The sub-criteria for photovoltaic resource potential are solar radiation and ambient temperature, and the sub-criteria for engineering feasibility are slope, aspect, topographic relief, and land use type.

[0069] By calculating the principal eigenvectors of each matrix, the relative weights (priorities) of each criterion can be obtained. Simultaneously, the consistency ratio (CR) is calculated to verify the logical consistency of expert judgments, requiring a CR < 0.10.

[0070] Power generation potential calculation:

[0071] First, the data values ​​of each sub-criteria layer are normalized to a common power generation potential scale (e.g., 0 to 1).

[0072] Then, the final power generation potential score S for each grid cell is calculated using a weighted linear combination:

[0073]

[0074] Where n is the number of sub-criteria, w i Let c be the global weight of sub-criterion i. i This is the normalized score of the sub-criterion in the corresponding grid cell.

[0075] The output of this step is the final site power generation potential map.

[0076] S5. Spatial Pattern Analysis and Optimization:

[0077] Traditional MCDA methods process each pixel independently when calculating the power generation potential score for each location, without considering the spatial context of the surrounding environment or the spatial variations in the influence of various factors. This step addresses this deficiency by introducing spatial statistical methods.

[0078] First, the final power generation potential map was analyzed using the global Moran index. This index provides a single value to determine whether high-power generation potential areas exhibit a clustered, discrete, or random distribution pattern throughout the study area. A significantly positive Moran index indicates that high-power generation potential areas tend to be adjacent to each other, which is advantageous for large-scale contiguous development.

[0079] Then, a bivariate local spatial autocorrelation analysis was performed. Input variable 1 for this analysis was the final power generation potential map generated in step S4, and input variable 2 was the climate suitability index map generated in step S3. This analysis identified four statistically significant spatial correlation patterns:

[0080] High-high concentration: A region with high power generation potential is surrounded by other regions with high power generation potential, and at the same time, it is located within a highly climate-suitable region and its concentration zone. This is an ideal target region that combines high power generation potential with high long-term climate security.

[0081] High-low clustering: Areas with high power generation potential are located within low-climate suitable areas and their clustering zones. Although such areas have high power generation potential, they are located within low-climate suitable areas and cannot be used as power generation areas; therefore, these areas should be excluded.

[0082] Low-high clustering: Areas with low power generation potential are located in areas with high climate suitability. Although the climate suitability is high, the power generation potential is low, so they are not the priority, but can be considered as alternative areas.

[0083] Low-low clustering: Areas with low power generation potential that are located in low climate suitability zones should also be explicitly excluded.

[0084] Then, geographically weighted regression (GWR) was introduced to model local regression relationships.

[0085] Model construction: The final power generation potential score is used as the dependent variable, and each evaluation criterion is used as the independent variable.

[0086] Local regression: GWR fits an independent local regression equation for each grid cell within the study area, and the construction of the equation takes into account the neighboring observation data.

[0087] Results Analysis and Optimization: The output of GWR is a series of local coefficient raster plots, each showing the spatial distribution of the influence of a specific criterion on power generation potential. By analyzing these coefficient plots, we can identify:

[0088] In which regions is solar radiation the most critical factor determining power generation potential?

[0089] In which regions is engineering feasibility considered the dominant factor?

[0090] Then, for regions where the key positive factors revealed by GWR have a particularly strong impact, their power generation potential score can be further improved.

[0091] Optimization and final site allocation:

[0092] The revised final site power generation potential map will serve as input to the optimization algorithm. The algorithm aims to identify and delineate one or more candidate sites.

[0093] Algorithm input:

[0094] Final power generation potential map.

[0095] Target power generation capacity (e.g., 100MW).

[0096] An estimate of the land area required per megawatt (e.g., 2 hectares / MW).

[0097] Constraints on the final plot shape or connectivity.

[0098] Algorithm execution: The algorithm starts with the highest-scoring unit on the power generation potential map and continuously merges adjacent high-scoring units through region growth or similar methods until a coherent plot that meets the minimum area requirement is formed.

[0099] Algorithm output: The final output is a set of sorted polygons with delineated boundaries, representing the best candidate sites for building centralized photovoltaic power plants.

[0100] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0101] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for optimizing the site selection of centralized photovoltaic power plants under multiple factors, characterized in that, Includes the following steps: S1. Data Acquisition and Preprocessing: Acquire and process multi-source geospatial data within the study area using a unified coordinate system and raster size. The data includes: topographic data, land use data, and meteorological conditions and climate disaster risk data. S2. Constraint Exclusion Analysis: Based on a set of preset absolute constraints, including legal nature reserves, water bodies, urban areas, and thresholds for slope, aspect, and topographic relief, a binary exclusion zone layer is identified and generated within the study area to shield unsuitable areas and obtain areas that are technically developable and constructible. S3. Construction of Climate Suitability Index Layer: Based on the meteorological conditions and the types of climate disaster risks, multiple climate layers are constructed and normalized. Then, the multiple climate layers are aggregated by weighted linear combination to construct a climate suitability index layer. S4. Evaluation Framework Construction and Weighting: Establish a hierarchical evaluation framework that includes photovoltaic resource potential and engineering feasibility, and use the analytic hierarchy process (AHP) to determine the weight of each criterion by constructing pairwise comparison matrices and calculating their principal eigenvectors. S5. Comprehensive assessment of power generation potential: All criterion layers are weighted and superimposed according to the determined weights, and the power generation potential score of each grid cell is calculated to generate the final site power generation potential map covering the study area. S6. Spatial correlation analysis and optimization: Perform spatial statistical analysis on the final site power generation potential map, identify the spatial pattern, and revise the final site power generation potential map; S7. Optimal Site Identification and Division: On the corrected final site power generation potential map, based on the preset target installed capacity and the land area required per megawatt, the region growth algorithm is used to merge adjacent high-scoring units starting from the unit with the highest suitability score until a coherent plot that meets the area requirements is formed, and this plot is output as the optimal candidate site.

2. The method for optimizing the site selection of centralized photovoltaic power plants under multiple factors according to claim 1, characterized in that, The topographic data includes slope, aspect, and topographic relief; meteorological conditions include total horizontal irradiance, average annual temperature, annual precipitation, total annual cloud cover, and average annual relative humidity; climate disaster risk data includes frost risk, strong wind risk, snow accumulation risk, high temperature risk, and lightning disaster risk; land use data includes nature reserves and land use types.

3. The method for optimizing the site selection of centralized photovoltaic power plants under multiple factors according to claim 1, characterized in that, The steps for constructing the climate suitability index layer include: normalizing the data values ​​of each independent climate layer to a general scale of 0-1 using a linear function, and then aggregating the normalized climate layers by weighted linear combination.

4. The method for optimizing the site selection of centralized photovoltaic power plants under multiple factors according to claim 1, characterized in that, The analytic hierarchy process used in step S4 further includes: calculating a consistency ratio to verify the logical consistency of the pairwise comparison matrix, wherein the consistency ratio is less than a preset ratio threshold.

5. The method for optimizing the site selection of centralized photovoltaic power plants under multiple factors according to claim 1, characterized in that, The spatial statistical analysis performed in step S6 is a bivariate local Moran index analysis.

6. The method for optimizing the site selection of a centralized photovoltaic power station under multiple factors according to claim 5, characterized in that, The bivariate local Moran index analysis specifically includes: The final site power generation potential map generated in step S5 is used as the first variable, and the climate suitability index map generated in step S3 is used as the second variable. For each grid cell within the study area, a local Moran index is calculated. This index measures the spatial covariance between the value of the first variable in that cell and the spatial lag value of the second variable in the neighboring region. Based on the calculation results and statistical significance of the local Moran index, the local spatial correlation pattern of each unit is divided into one of four types: high power generation potential-high suitability, high power generation potential-low suitability, low power generation potential-high suitability, and low power generation potential-high suitability.

7. The method for optimizing the site selection of centralized photovoltaic power plants under multiple factors according to claim 1, characterized in that, The spatial statistical analysis performed in step S6 is a geographically weighted regression analysis.

8. The method for optimizing the site selection of a centralized photovoltaic power station under multiple factors according to claim 7, characterized in that, The geographically weighted regression analysis specifically includes: Establish a regression model with power generation potential as the dependent variable and various evaluation criteria as independent variables; For each grid cell within the study area, an independent local regression equation is fitted. Generate a series of local regression coefficient layers, where each layer shows the spatial distribution of the influence of an independent variable on the dependent variable; Based on the magnitude and statistical significance of the coefficients in the local regression coefficient layer, the final site power generation potential map is revised.

Images