Thermal power plant photovoltaic layout method and system

Abstract

The application provides a thermal power plant photovoltaic layout method and system, BIM data and GIS data are acquired, preprocessed, and the preprocessed data is deeply fused to construct a fusion model; based on the fusion model, a shadow area is determined, and the shadow area is divided into static fixed shadow, dynamic temporary shadow, time sequence gradual change shadow and superimposed complex shadow; for the four types of shadows, an adaptive artificial intelligence model is constructed, and quantitative analysis of various shadows is realized according to the artificial intelligence model; the shadow analysis results are spliced to form the total shadow analysis result, and the best photovoltaic array layout is determined through multi-objective optimization algorithm combined with GIS spatial analysis, photovoltaic component technical parameters and engineering constraint conditions, specifically, the accurate quantification of shadow characteristics is realized through the creative shadow analysis logic, the best photovoltaic array layout is determined combined with fusion modeling and multi-objective optimization, and the photovoltaic power generation efficiency is improved.

Description

Technical Field

[0001] This invention belongs to the field of photovoltaic layout technology for thermal power plants, and specifically relates to a method and system for photovoltaic layout in thermal power plants. Background Technology

[0002] With the rapid development of the new energy industry, thermal power plants, as traditional energy bases, have made their idle spaces, such as rooftops and open spaces, important carriers for photovoltaic development. However, thermal power plant areas are densely built and have complex equipment, resulting in various types of shading (such as shadows from fixed buildings, temporary equipment, vegetation, and superimposed shadows). Shading can seriously affect the power generation efficiency of photovoltaic arrays. Therefore, accurate shading analysis is a core prerequisite for the rational layout of photovoltaic arrays in thermal power plants. In existing methods for photovoltaic (PV) layout in thermal power plants, shadow analysis often adopts a broad, "one-size-fits-all" approach, failing to differentiate between the formation mechanisms and variation patterns of different types of shadows. It typically uses only a single model or simple overlay method to analyze shadows, resulting in low accuracy and poor adaptability, and an inability to accurately quantify shadow characteristics under complex shading scenarios. Furthermore, existing methods do not achieve deep integration of the spatial analysis capabilities of GIS (Geographic Information System) and the detailed modeling advantages of BIM (Building Information Modeling), making it difficult to consider both site spatial attributes and detailed information about building equipment. This often leads to problems such as low power generation efficiency, poor engineering feasibility, and inconvenient operation and maintenance in PV array layouts. Summary of the Invention

[0003] Based on this, the present invention provides a method and system for photovoltaic layout in thermal power plants, which aims to solve the problems of crude shadow analysis and unreasonable layout in the prior art.

[0004] A first aspect of this invention provides a method for photovoltaic layout in a thermal power plant, the method comprising: Acquire BIM and GIS data, preprocess them, and then deeply integrate the preprocessed BIM and GIS data to build a fusion model. Based on the fusion model, the shadow region is determined and divided into four categories, including static fixed shadow, dynamic temporary shadow, temporal gradual shadow and superimposed complex shadow; For the four types of shadows, corresponding artificial intelligence models are constructed, and quantitative analysis of each type of shadow is achieved based on the artificial intelligence models. Specifically, a CNN model is used for the static fixed shadows; a YOLO+LSTM fusion model is used for the dynamic temporary shadows; a Transformer model is used for the temporally changing shadows; and an attention mechanism fusion model is used for the superimposed complex shadows. The shadow analysis results processed by each artificial intelligence model are stitched together to form the overall shadow analysis result; Based on the overall results of the shadow analysis, combined with GIS spatial analysis, photovoltaic module technical parameters, and engineering constraints, the optimal photovoltaic array layout is determined through a multi-objective optimization algorithm.

[0005] Furthermore, the static fixed shadow is formed by permanent buildings or fixed equipment, with a fixed position and shape, and only shifts with the time sequence of the sun's trajectory, resulting in a stable shading range; The dynamic temporary shadow is formed by movable devices and temporary facilities, with its position and shape changing randomly, and its duration and range of occlusion uncertain; The time-varying shadow is formed by vegetation and changes gradually with the seasons and day and night, with the intensity and range of shading fluctuating regularly. The superimposed complex shadow is formed by superimposing two or more single shadows from the static fixed shadow, the dynamic temporary shadow, and the temporal gradient shadow, resulting in a shadow with complex shape and uneven occlusion intensity.

[0006] Furthermore, in the step of constructing adapted artificial intelligence models for the four types of shadows and realizing quantitative analysis of various types of shadows based on the artificial intelligence models, for the static fixed shadows, the three-dimensional parameters of static buildings and fixed equipment, solar trajectory time series data, and component material reflectivity in the fusion model are obtained, input into the trained CNN model, and the first occlusion range and actual solar irradiance are output. Based on the actual solar irradiance, the occlusion duration and occlusion intensity are calculated. For the dynamic temporary shadow, the three-dimensional parameters, historical movement trajectory data, operation and maintenance scheduling plan, and solar trajectory time series data of the mobile device and temporary facility in the fusion model are obtained. The trained YOLO+LSTM fusion model is input and the mobile device position, device movement speed, solar altitude angle and azimuth angle are output. Based on the mobile device position, device movement speed, solar altitude angle and azimuth angle, the second occlusion range is determined and the occlusion probability and occlusion duration are calculated. The occlusion is emphasized by the occlusion probability. For the time-varying shading, time-series remote sensing images, GIS terrain data, meteorological data, vegetation parameters, and solar trajectory time-series data are acquired, input into the trained Transformer model, and output vegetation leaf area index and vegetation height. Based on the vegetation leaf area index, vegetation height, solar altitude angle, and azimuth angle, the third shading range is determined, the shading intensity is calculated, and the shading duration is statistically analyzed. For the superimposed complex shadow, obtain the occlusion range, occlusion duration, and occlusion intensity of the shadow output by the CNN model, YOLO+LSTM fusion model, and Transformer model, as well as the weight coefficients of static fixed shadow, dynamic temporary shadow, and temporally gradation shadow. Input the trained attention mechanism fusion model and output the fourth occlusion range, occlusion duration, and occlusion intensity of the superimposed shadow.

[0007] Furthermore, the step of stitching together the shadow analysis results processed by each artificial intelligence model to form the overall shadow analysis result includes: Weights are assigned to the shadow analysis results processed by each artificial intelligence model. Specifically, type weights are assigned based on the degree of impact of various types of shadows on photovoltaic power generation and the stability of shading; intensity weights are assigned based on the shading intensity of various types of shadows; and region adaptation weights are assigned based on the type and priority of photovoltaic layout areas. The comprehensive weight of each type of shadow in the corresponding region is calculated, and the shadow analysis results processed by each artificial intelligence model are spatially superimposed based on the fusion model to obtain the stitched occlusion range. For the overlapping areas in the stitched occlusion range, the shadow type with the highest comprehensive weight is selected as the core shadow, and the occlusion intensity of other overlapping shadows is merged to obtain the merged occlusion intensity. For non-overlapping areas within the stitched occlusion range, the corresponding shadow analysis results are directly retained; Based on the occlusion intensity after fusion, the total occlusion duration is adjusted to achieve linkage matching between occlusion intensity and occlusion duration.

[0008] Furthermore, in the step of calculating the comprehensive weight of each type of shadow in the corresponding region, the formula for calculating the comprehensive weight is as follows: ; in, Let be the combined weight of the i-th type of shading in the region (x,y). Let i be the type weight of the i-th type of shadow. Let i be the intensity weight of the i-th type of shadow. The adaptation weights for the region (x,y).

[0009] Furthermore, in the step of selecting the shadow type with the highest comprehensive weight as the core shadow for the overlapping area in the stitched occlusion range, and simultaneously fusing the occlusion intensity of other overlapping shadows to obtain the fused occlusion intensity, the calculation formula for the fusion of occlusion intensity in the overlapping area is as follows: ; in, Let represent the occlusion intensity of the overlapping region (x, y) after fusion at time t. Let be the combined weight of the i-th type of shading in the region (x,y). Let be the occlusion intensity of the i-th type of shadow.

[0010] Furthermore, in the step of correcting the total occlusion duration based on the fused occlusion intensity, the correction formula is as follows: ; ; in, This is the corrected total occlusion duration. The total duration of occlusion monitored in the field, where k is the correction coefficient. Let be the total annual occlusion duration for region (x, y). This represents the total monitoring time for the entire year. Let be the duration of occlusion of the i-th type of shadow at time t. The occlusion intensity after fusion.

[0011] A second aspect of this invention provides a photovoltaic layout system for a thermal power plant, used to implement the photovoltaic layout method for a thermal power plant described in the first aspect, the system comprising: The fusion module is used to acquire BIM data and GIS data, perform preprocessing, and then deeply fuse the preprocessed BIM data and GIS data to build a fusion model. The segmentation module is used to determine the shadow region based on the fusion model and divide the shadow region into four categories, including static fixed shadow, dynamic temporary shadow, temporal gradient shadow and superimposed complex shadow; The module is used to build suitable artificial intelligence models for four types of shadows, and to perform quantitative analysis of each type of shadow based on the artificial intelligence models. Specifically, a CNN model is used for static fixed shadows; a YOLO+LSTM fusion model is used for dynamic temporary shadows; a Transformer model is used for temporally changing shadows; and an attention mechanism fusion model is used for superimposed complex shadows. The stitching module is used to stitch together the shadow analysis results processed by each artificial intelligence model to form the overall shadow analysis result; The photovoltaic array layout determination module is used to determine the optimal photovoltaic array layout based on the overall results of the shadow analysis, combined with GIS spatial analysis, photovoltaic module technical parameters, and engineering constraints, through a multi-objective optimization algorithm.

[0012] A third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the photovoltaic layout method for thermal power plants provided in the first aspect.

[0013] A fourth aspect of the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the photovoltaic layout method for thermal power plants provided in the first aspect.

[0014] This invention provides a method and system for photovoltaic (PV) layout in a thermal power plant. It acquires and preprocesses BIM and GIS data, then deeply integrates the preprocessed BIM and GIS data to construct a fusion model. Based on this model, shadow areas are identified and categorized into four types: static fixed shadows, dynamic temporary shadows, time-varying shadows, and superimposed complex shadows. For each of the four types, a suitable artificial intelligence (AI) model is constructed, and quantitative analysis of each type of shadow is performed. The shadow analysis results processed by each AI model are then combined to form a total shadow analysis result. Based on this total result, and combined with GIS spatial analysis, PV module technical parameters, and engineering constraints, a multi-objective optimization algorithm determines the optimal PV array layout. Specifically, through innovative shadow analysis logic, precise quantification of shadow characteristics is achieved. Combined with GIS-BIM fusion modeling and multi-objective optimization, the optimal PV array layout is determined, thereby improving PV power generation efficiency. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the implementation of a photovoltaic layout method for a thermal power plant according to Embodiment 1 of the present invention. Figure 2 This is a structural block diagram of a photovoltaic layout system for a thermal power plant provided in Embodiment 2 of the present invention; Figure 3 This is a structural block diagram of an electronic device provided in Embodiment 3 of the present invention. Detailed Implementation

[0016] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0017] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0019] Example 1 According to an embodiment of the present invention, a method for photovoltaic layout in a thermal power plant is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0020] This first embodiment provides a method for photovoltaic layout in a thermal power plant, which can be used in electronic devices, such as computers. Please refer to [link / reference]. Figure 1 , Figure 1 The flowchart of a photovoltaic layout method for a thermal power plant provided in Embodiment 1 of the present invention is shown, specifically including steps S01 to S05.

[0021] Step S01: Obtain BIM data and GIS data, perform preprocessing, and then deeply integrate the preprocessed BIM data and GIS data to construct a fusion model.

[0022] In this embodiment of the invention, for BIM data, existing engineering drawings (civil engineering, electromechanical, steel structure) and BIM model files (preferably IFC format) of thermal power plants are collected. The three-dimensional geometric parameters of all buildings (factory buildings, chimneys, cooling towers, power distribution rooms, etc.), equipment foundations, and pipelines in the plant area are extracted, including length, width, height, cross-sectional dimensions, spatial coordinates (X, Y, Z), and component materials (reflectivity, absorptivity). Detailed information such as roof slope, parapet wall height, roof waterproofing level, open space flatness, and ground bearing capacity are supplemented. Subsequently, tools such as Revit and Navisworks were used to optimize the BIM model for lightweighting, removing redundant components unrelated to the photovoltaic layout (such as interior furniture and small pipelines), simplifying the geometry of complex components (while retaining the core outline), and ensuring that the model file size is suitable for subsequent GIS import and shadow analysis. The coordinate system of the BIM model is unified into an independent coordinate system for the factory area (consistent with the GIS data). Through coordinate translation, rotation, and scaling, the BIM model is ensured to accurately correspond to the actual factory area location.

[0023] For GIS data, 1:500 scale GIS vector data (topography, roads, green spaces, pipeline distribution, land use boundaries) of the thermal power plant area and surrounding areas were acquired. Regional meteorological data (average annual total solar radiation, average monthly sunshine duration, solar altitude / azimuth time series data, annual precipitation, and average annual wind speed over the past 5 years), latitude and longitude coordinates of the plant area, and elevation data (accuracy ±10cm) were collected. Satellite remote sensing imagery (resolution ≥0.5m) was acquired through GIS platforms (ArcGIS, QGIS) for subsequent vegetation shadow analysis. Furthermore, GIS tools are used to perform topological checks on vector data, correcting coordinate deviations and topological errors (such as overlaps and gaps); denoising is performed on meteorological data to remove outliers (such as extreme weather data), and linear interpolation is used to supplement missing data; elevation data is smoothed to generate a digital elevation model (DEM) of the plant area. Finally, the corrected GIS data (vector, image, DEM, meteorological data) are integrated to construct a GIS spatial database for the factory area, enabling the classification, storage, query, and retrieval of data, and ensuring that the data format is compatible with the BIM model.

[0024] It should be noted that the data fusion process involves using a 3D modeling plugin for a GIS platform (such as ArcGIS's BIMConnector) to import a lightweight BIM model (IFC format) into a GIS spatial database, achieving precise integration between the BIM model and the GIS terrain and coordinate system. Subsequently, through the GIS attribute association function, the component attributes (material, height, purpose, load-bearing capacity) of the BIM model are bound to the spatial attributes (location, elevation, land use type, solar radiation intensity) of the GIS data, forming a comprehensive 3D model containing "geometric information + attribute information + spatial information." Finally, the fused 3D model is validated to check its integrity (no missing components, no coordinate deviations) and attribute consistency (component attributes match reality), ensuring that the model accurately reflects the actual situation of the factory area and meets the requirements of shadow analysis and photovoltaic layout.

[0025] More specifically, auxiliary data was also collected, including photovoltaic module technical parameters (rated power, dimensions, installation angle adaptation range, photoelectric conversion efficiency, temperature coefficient), inverter parameters (rated capacity, conversion efficiency), land use planning documents for thermal power plant sites, equipment operation and maintenance access requirements (width ≥ 1.5m), cable laying path constraints (avoiding dense pipeline areas and equipment foundations), and roof load-bearing capacity test reports. This collected auxiliary data was converted into a unified format to clarify the engineering boundary conditions of the photovoltaic layout, establish a constraint database, and mark the coordinate range and constraint type of unusable areas, providing a basis for subsequent layout optimization.

[0026] Furthermore, based on the fusion model and combined with the constraints in the auxiliary data, unusable areas are explicitly excluded through the GIS spatial query function. Examples include equipment maintenance passages, dense pipeline areas, areas with insufficient roof load-bearing capacity, fire lanes, areas directly under tall equipment that are obstructed, and areas outside the land boundary line.

[0027] Furthermore, the remaining usable area is divided into rooftop usable area and open space usable area. Rooftop usable area is classified according to factory building type (main factory building, auxiliary factory building) and roof slope (flat roof ≤5°, gentle slope roof 5°-30°, steep slope roof >30°). Open space usable area is classified according to flatness (flatness error ≤5cm, 5-10cm) and shading risk (low risk ≤1h / day, medium risk 1-2h / day). A vector map layer of usable area is generated, and the core parameters of each area (area, slope, load-bearing capacity, shading risk) are marked.

[0028] Step S02: Based on the fusion model, determine the shadow region and divide the shadow region into 4 categories, including static fixed shadow, dynamic temporary shadow, temporal gradient shadow and superimposed complex shadow.

[0029] Specifically, the static fixed shadow is formed by permanent buildings and fixed equipment (such as chimneys, cooling towers, factory buildings, and transformer foundations), with a fixed position and shape, which only shifts with the time sequence of the sun's trajectory, and has a stable shading range. In addition, the shading duration has a regularity (such as the shading duration being longer in winter than in summer). The dynamic temporary shadow is formed by movable equipment (such as cranes, maintenance vehicles, forklifts) and temporary facilities (such as maintenance sheds, scaffolding), and its position and shape change randomly, with uncertain duration and range of occlusion. The aforementioned time-varying shadow is formed by vegetation (trees, shrubs), which gradually changes with the seasons (leaf fall, budding, growth) and day and night, with the intensity and range of shading fluctuating regularly (e.g., strong shading in summer when vegetation is lush, weak shading in winter when leaves fall; strong shading during the day, no shading at night). The superimposed complex shadow is formed by superimposing two or more single shadows from the static fixed shadow, the dynamic temporary shadow, and the time-varying shadow (such as superimposing factory shadow and tree shadow, or superimposing fixed equipment shadow and temporary equipment shadow), resulting in a shadow with complex shape and uneven occlusion intensity.

[0030] Step S03: For the four types of shadows, construct appropriate artificial intelligence models for each type, and perform quantitative analysis of each type of shadow based on the artificial intelligence models.

[0031] For the static fixed shadow, a CNN model is used. Specifically, the three-dimensional parameters (height H, dimensions L×W, spatial coordinates (X,Y,Z)) of the static building (structure) and fixed equipment, the solar trajectory time series data (solar altitude angle α, solar azimuth angle β), and the material reflectivity ρ of the components are obtained from the fusion model and input into the trained CNN model. The first occlusion range and the actual solar irradiance are output. Based on the actual solar irradiance, the occlusion duration and occlusion intensity are calculated. The formula for calculating the occlusion duration is as follows: ; The formula for calculating shading intensity is: ; in, The duration of static, fixed shadow occlusion in the region (x, y) at time t. Let be the actual solar irradiance of region (x, y) at time t. Let t be the standard solar irradiance at time t, and t1 and t2 be the times when the shadow appears and disappears, respectively. This represents the occlusion intensity under static, fixed shadow types. Understandably, several typical static shadow areas within the factory area are selected, and on-site shadow data (occlusion range, duration, and intensity) is collected via drone aerial photography and ground sensors. This data serves as the training set (70%) and validation set (30%). A CNN model is trained to learn the formation patterns of static shadows, and the model's convolutional kernel size (3×3), learning rate (0.001), and number of iterations (1000) are optimized to ensure model convergence.

[0032] For the aforementioned dynamic temporary shadow, a YOLO+LSTM fusion model is employed. Specifically, the 3D parameters (size, movement speed v, movement trajectory), historical movement trajectory data, operation and maintenance scheduling plan, and solar trajectory time series data of the movable device and temporary facility are obtained from the fusion model. These are input into the trained YOLO+LSTM fusion model, which outputs the movable device's position, movement speed, solar altitude angle, and azimuth angle. Based on the movable device's position, movement speed, solar altitude angle, and azimuth angle, the second occlusion range is determined, and the occlusion probability and occlusion duration are calculated. Here, occlusion is emphasized using the occlusion probability, and the formula for calculating the occlusion probability is: ; in, P represents the probability (0≤P≤1) that the region (x,y) is occluded by a dynamic temporary shadow at time t. Here, is the sigmoid activation function, used to map the output to the [0,1] interval; W is the model weight matrix; b is the model bias term; and v(t) is the moving speed of the mobile device at time t. , Let be the solar altitude angle and azimuth angle at time t, and L(t) be the position coordinates (X, Y) of the mobile device at time t. Understandably, using recent mobile device operation and maintenance records, temporary facility construction / dismantling records, and on-site shadow monitoring data from the factory area, a YOLO model (for real-time identification of equipment / facilities location and shape) and an LSTM model (for predicting dynamic shadow temporal changes) are trained. The confidence threshold of the YOLO model is set to 0.8, the number of hidden layer nodes of the LSTM model is set to 128, and the number of iterations is 800, to achieve real-time identification and prediction of dynamic shadows.

[0033] For the aforementioned temporally varying shading, a Transformer model is employed. Specifically, temporal remote sensing images (capturing seasonal vegetation changes), GIS topographic data, meteorological data (monthly average temperature T, monthly precipitation R, affecting vegetation growth), vegetation parameters (height h, canopy radius r), and temporal solar trajectory data are acquired and input into the trained Transformer model. The model outputs the vegetation leaf area index and vegetation height. Based on the vegetation leaf area index, vegetation height, solar altitude angle, and azimuth angle, the third shading range is determined, and the shading intensity and duration are calculated. The formula for calculating the solar irradiance after temporally varying shading is as follows: ; Let be the solar irradiance of region (x, y) after being shaded by a time-varying shading at time t, u be the vegetation shading coefficient, F(t) be the vegetation leaf area index at time t, and h(t) be the vegetation height at time t. The formula for calculating the shading intensity under the time-varying shading type is: ; Where 0≤S≤1, the larger S is, the stronger the occlusion. Understandably, several typical vegetation areas in the factory area were selected, and time-series remote sensing images, vegetation growth data (height, crown width), and shadow monitoring data from recent years were collected to train a Transformer model to capture the correlation between vegetation growth and shadow changes. The number of attention heads in the model was set to 8, the hidden layer dimension was set to 256, and the number of iterations was 1200 to ensure that the model can accurately predict vegetation shadow changes in different seasons.

[0034] For the aforementioned complex superimposed shadows, an attention mechanism fusion model is employed. Specifically, the occlusion range, occlusion duration, and occlusion intensity of the shadows output from the CNN model, the YOLO+LSTM fusion model, and the Transformer model, as well as the weight coefficients of static fixed shadows, dynamic temporary shadows, and time-varying shadows, are obtained. These are input into the trained attention mechanism fusion model, which outputs the fourth occlusion range, occlusion duration, and occlusion intensity of the superimposed shadows. Understandably, several typical shadow superimposed areas in the factory area are selected, and real-world shadow superimposed data is collected to train the attention mechanism fusion model. This automatically identifies the superimposed areas and proportions of various shadow types, optimizes the attention weight allocation, and ensures the quantification accuracy of the superimposed shadows. The formula for calculating the occlusion intensity of the superimposed shadows is: ; Let represent the intensity of superimposed complex shadow occlusion in region (x,y) at time t. The weight coefficient for the i-th type of single shadow is... Let S be the occlusion intensity of the i-th type of single shadow in the region (x,y) at time t (0≤S≤1).

[0035] Step S04: The shadow analysis results processed by each artificial intelligence model are spliced ​​together to form the overall shadow analysis result.

[0036] Specifically, weights are assigned to the shadow analysis results processed by each artificial intelligence model. Among them, type weights are assigned based on the degree of impact of various types of shadows on photovoltaic power generation and the stability of shading; intensity weights are assigned based on the shading intensity of various types of shadows; and region adaptation weights are assigned based on the type and priority of photovoltaic layout areas. The comprehensive weight of each type of shadow in the corresponding region is calculated, and based on the fusion model, the shadow analysis results processed by each artificial intelligence model are spatially superimposed to obtain the stitched occlusion range. The formula for calculating the comprehensive weight is as follows: ; in, Let be the combined weight of the i-th type of shading in the region (x,y). Let i be the type weight of the i-th type of shadow. Let i be the intensity weight of the i-th type of shadow. The adaptation weights for the region (x,y); For overlapping areas within the stitched occlusion range, the shadow type with the highest overall weight is selected as the core shadow. Simultaneously, the occlusion intensities of other overlapping shadows are fused to obtain the fused occlusion intensity. The formula for calculating the occlusion intensity fusion of overlapping areas is as follows: ; in, Let represent the occlusion intensity of the overlapping region (x, y) after fusion at time t. Let be the combined weight of the i-th type of shading in the region (x,y). Let be the occlusion intensity of the i-th type of shadow; For non-overlapping areas within the stitched occlusion range, the corresponding shadow analysis results are directly retained; Based on the occlusion intensity after fusion, the total occlusion duration is adjusted to achieve a linkage match between occlusion intensity and occlusion duration. The adjustment formula is as follows: ; ; in, This is the corrected total occlusion duration. The total duration of occlusion monitored in the field, where k is the correction coefficient. Let be the total annual occlusion duration for region (x, y). This represents the total monitoring time for the entire year. Let be the duration of occlusion of the i-th type of shadow at time t. The result is the occlusion intensity after merging. Understandably, the stitched shadow analysis results are corrected based on on-site monitoring data. A polynomial fitting method is used to correct boundary errors, and outlier data (such as data with occlusion duration exceeding a reasonable range) is removed. The boundaries of the shadow range are smoothed to ensure that the shadow range is consistent with the actual situation, and the corrected error is ≤5%.

[0037] Step S05: Based on the overall results of the shadow analysis, combined with GIS spatial analysis, photovoltaic module technical parameters, and engineering constraints, the optimal photovoltaic array layout is determined through a multi-objective optimization algorithm.

[0038] It should be noted that, based on the overall results of the shading analysis, combined with the roof load-bearing capacity, open ground flatness, and cable laying path, the priority of photovoltaic layout is determined. In this embodiment of the invention, the specific order is as follows: unshaded area (highest priority, P1=1.0) → lightly shaded area (second highest priority, P2=0.8) → moderately shaded area (careful layout, P3=0.5) → heavily shaded area (layout prohibited, P4=0).

[0039] Furthermore, the available areas are screened, prioritizing areas with high priority, large area, load-bearing capacity, and convenient cable laying, while excluding areas with low priority and unmet constraints, thus generating a candidate area layer for photovoltaic layout. Based on GIS spatial analysis functions, combined with solar radiation data and shading duration, the optimal installation angle of photovoltaic modules is determined to ensure that the modules receive the maximum solar irradiance. Based on the shadow analysis results, the optimal array spacing is calculated to ensure that there is no inter-array shading in winter (when the solar altitude angle is the smallest); The usable areas on the roof are arranged along the slope (consistent with the roof slope) to reduce the amount of roof renovation work; the usable areas on the open ground are arranged in rows and columns to balance space utilization and power generation efficiency, and the arrangement direction is perpendicular to the solar azimuth angle (to maximize the reception of solar radiation).

[0040] Furthermore, a genetic algorithm is employed, with the objective functions of "maximizing annual power generation, maximizing installed capacity per unit area, and minimizing engineering investment," and constraints including shading, roof load-bearing capacity, maintenance access, and cable laying, to optimize the photovoltaic array layout scheme. The objective function is expressed as: ; ; ; In this embodiment of the invention, the constraints are as follows: (Occlusion strength constraint, S) max =0.5); (Roof load-bearing constraints, W) max =2.5kN / m 2 ); (Operation and maintenance channel constraints, D) min =1.5m); Where M represents the number of photovoltaic deployment areas, and A i Let G be the area of ​​the i-th region. i Let S be the total annual solar radiation for the i-th region, η be the photovoltaic module conversion efficiency, and S be the total annual solar radiation. i Let P be the occlusion intensity of the i-th region. total For the total installed capacity, A total C represents the total layout area. component For component cost, C installation For installation costs, C cable For cable laying costs, S max W is the maximum permissible occlusion intensity threshold.i The total load actually borne by the i-th roof area (unit: kN / m) 2 ), W max D represents the maximum allowable load-bearing capacity of the roof structure. i D represents the width of the maintenance channel between the i-th photovoltaic arrays / between the array and the boundary. min This is the minimum width requirement for the operation and maintenance channel.

[0041] In summary, the photovoltaic layout method for thermal power plants in the above embodiments of the present invention acquires BIM data and GIS data, performs preprocessing, and deeply integrates the preprocessed BIM data and GIS data to construct a fusion model. Based on the fusion model, shadow areas are determined and divided into four categories: static fixed shadows, dynamic temporary shadows, time-varying shadows, and superimposed complex shadows. For each of the four types of shadows, an appropriate artificial intelligence model is constructed, and quantitative analysis of each type of shadow is achieved based on the artificial intelligence model. The shadow analysis results processed by each artificial intelligence model are stitched together to form a total shadow analysis result. Based on the total shadow analysis result, combined with GIS spatial analysis, photovoltaic module technical parameters, and engineering constraints, the optimal photovoltaic array layout is determined through a multi-objective optimization algorithm. Specifically, through innovative shadow analysis logic, the precise quantification of shadow features is achieved. By combining GIS-BIM fusion modeling and multi-objective optimization, the optimal photovoltaic array layout is determined, thereby improving photovoltaic power generation efficiency.

[0042] Example 2 Please see Figure 2 , Figure 2 This is a structural block diagram of a photovoltaic layout system for a thermal power plant according to Embodiment 2 of the present invention. This photovoltaic layout system 200 is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0043] Specifically, the photovoltaic layout system 200 for thermal power plants includes: a fusion module 21, a division module 22, a construction module 23, a splicing module 24, and a photovoltaic array layout determination module 25, wherein: The fusion module 21 is used to acquire BIM data and GIS data, perform preprocessing, and then deeply fuse the preprocessed BIM data and GIS data to build a fusion model. The segmentation module 22 is used to determine the shadow area based on the fusion model and divide the shadow area into 4 categories, including static fixed shadow, dynamic temporary shadow, time-varying shadow and superimposed complex shadow. The static fixed shadow is formed by permanent buildings and fixed equipment, with a fixed position and shape, and only moves with the time of the sun's trajectory, with a stable occlusion range. The dynamic temporary shadow is formed by movable devices and temporary facilities, with its position and shape changing randomly, and its duration and range of occlusion uncertain; The time-varying shadow is formed by vegetation and changes gradually with the seasons and day and night, with the intensity and range of shading fluctuating regularly. The superimposed complex shadow is formed by superimposing two or more single shadows from the static fixed shadow, the dynamic temporary shadow, and the time-varying shadow, resulting in a shadow with complex shape and uneven occlusion intensity. Module 23 is used to construct adapted artificial intelligence models for four types of shadows, and to perform quantitative analysis of each type of shadow based on the artificial intelligence models. Specifically, a CNN model is used for static fixed shadows; a YOLO+LSTM fusion model is used for dynamic temporary shadows; a Transformer model is used for temporally varying shadows; and an attention mechanism fusion model is used for superimposed complex shadows. In the step of constructing adapted artificial intelligence models for the four types of shadows and performing quantitative analysis of each type of shadow based on the artificial intelligence models, for static fixed shadows, the three-dimensional parameters of static buildings and fixed equipment, the temporal data of solar trajectory, and the reflectivity of component materials in the fusion model are obtained and input into the trained CNN model. The first occlusion range and the actual solar irradiance are output, and the occlusion duration and occlusion intensity are calculated based on the actual solar irradiance. For the dynamic temporary shadow, the three-dimensional parameters, historical movement trajectory data, operation and maintenance scheduling plan, and solar trajectory time series data of the mobile device and temporary facility in the fusion model are obtained. The trained YOLO+LSTM fusion model is input and the mobile device position, device movement speed, solar altitude angle and azimuth angle are output. Based on the mobile device position, device movement speed, solar altitude angle and azimuth angle, the second occlusion range is determined and the occlusion probability and occlusion duration are calculated. The occlusion is emphasized by the occlusion probability. For the time-varying shading, time-series remote sensing images, GIS terrain data, meteorological data, vegetation parameters, and solar trajectory time-series data are acquired, input into the trained Transformer model, and output vegetation leaf area index and vegetation height. Based on the vegetation leaf area index, vegetation height, solar altitude angle, and azimuth angle, the third shading range is determined, the shading intensity is calculated, and the shading duration is statistically analyzed. For the superimposed complex shadow, obtain the occlusion range, occlusion duration, and occlusion intensity of the shadow output by the CNN model, YOLO+LSTM fusion model, and Transformer model, as well as the weight coefficients of static fixed shadow, dynamic temporary shadow, and temporally gradation shadow. Input the trained attention mechanism fusion model and output the fourth occlusion range, occlusion duration, and occlusion intensity of the superimposed shadow. The stitching module 24 is used to stitch together the shadow analysis results processed by each artificial intelligence model to form the overall shadow analysis result; The photovoltaic array layout determination module 25 is used to determine the optimal photovoltaic array layout based on the overall results of the shadow analysis, combined with GIS spatial analysis, photovoltaic module technical parameters, and engineering constraints, through a multi-objective optimization algorithm.

[0044] Furthermore, in some optional embodiments of the present invention, the splicing module 24 includes: The weighting unit is used to assign weights to the shadow analysis results processed by each artificial intelligence model. Specifically, type weights are assigned based on the degree of impact of various types of shadows on photovoltaic power generation and the stability of shading; intensity weights are assigned based on the shading intensity of various types of shadows; and region adaptation weights are assigned based on the type and priority of photovoltaic layout areas. The calculation unit is used to calculate the comprehensive weight of each type of shadow in the corresponding region, and based on the fusion model, spatially superimposes the shadow analysis results processed by each artificial intelligence model to obtain the stitched occlusion range. The formula for calculating the comprehensive weight is: ; in, Let be the combined weight of the i-th type of shading in the region (x,y). Let i be the type weight of the i-th type of shadow. Let i be the intensity weight of the i-th type of shadow. The adaptation weights for the region (x,y); The blending unit is used to select the shadow type with the highest overall weight as the core shadow for the overlapping areas within the stitched occlusion range, while blending the occlusion intensities of other overlapping shadows to obtain the blended occlusion intensity. The calculation formula for the blending of occlusion intensity in overlapping areas is as follows: ; in, Let represent the occlusion intensity of the overlapping region (x, y) after fusion at time t. Let be the combined weight of the i-th type of shading in the region (x,y). Let be the occlusion intensity of the i-th type of shadow; The retained unit is used to directly retain the analysis results of the corresponding shadow in the non-overlapping areas of the stitched occlusion range; The correction unit is used to adjust the total occlusion duration based on the occlusion intensity after fusion, achieving linkage matching between occlusion intensity and occlusion duration. The correction formula is as follows: ; ; in, This is the corrected total occlusion duration. The total duration of occlusion monitored in the field, where k is the correction coefficient. Let be the total annual occlusion duration for region (x, y). This represents the total monitoring time for the entire year. Let be the duration of occlusion of the i-th type of shadow at time t. The occlusion intensity after fusion.

[0045] Example 3 In another aspect, the present invention also proposes an electronic device, please refer to [link to relevant documentation]. Figure 3 The image shows an electronic device according to Embodiment 3 of the present invention, including a memory 20, a processor 10, and a computer program 30 stored in the memory and executable on the processor. When the processor 10 executes the computer program 30, it implements the photovoltaic layout method of the thermal power plant as described above.

[0046] In some embodiments, the processor 10 may be a central processing unit (CPU), controller, microcontroller, microprocessor or other data processing chip, used to run program code stored in memory 20 or process data, such as executing access restriction programs.

[0047] The memory 20 includes at least one type of readable storage medium, such as flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 20 can be an internal storage unit of an electronic device, such as the hard disk of the electronic device. In other embodiments, the memory 20 can also be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. Furthermore, the memory 20 can include both internal and external storage units of the electronic device. The memory 20 can be used not only to store application software and various types of data of the electronic device, but also to temporarily store data that has been output or will be output.

[0048] It should be pointed out that, Figure 3 The structure shown does not constitute a limitation on the electronic device. In other embodiments, the electronic device may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0049] This invention also proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the photovoltaic layout method for thermal power plants as described above.

[0050] Those skilled in the art will understand that the logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0051] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0052] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0053] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," 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, the 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.

[0054] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this patent should be determined by the appended claims.

Claims

1. A method for photovoltaic layout in a thermal power plant, characterized in that, The method includes: Acquire BIM and GIS data, preprocess them, and then deeply integrate the preprocessed BIM and GIS data to build a fusion model. Based on the fusion model, the shadow region is determined and divided into four categories, including static fixed shadow, dynamic temporary shadow, temporal gradual shadow and superimposed complex shadow; For the four types of shadows, corresponding artificial intelligence models are constructed, and quantitative analysis of each type of shadow is achieved based on the artificial intelligence models. Specifically, a CNN model is used for the static fixed shadows; a YOLO+LSTM fusion model is used for the dynamic temporary shadows; a Transformer model is used for the temporally changing shadows; and an attention mechanism fusion model is used for the superimposed complex shadows. The shadow analysis results processed by each artificial intelligence model are stitched together to form the overall shadow analysis result; Based on the overall results of the shadow analysis, combined with GIS spatial analysis, photovoltaic module technical parameters, and engineering constraints, the optimal photovoltaic array layout is determined through a multi-objective optimization algorithm.

2. The photovoltaic layout method for thermal power plants according to claim 1, characterized in that, The static fixed shadow is formed by permanent buildings and fixed equipment, with a fixed position and shape, and only shifts with the time sequence of the sun's trajectory, with a stable occlusion range. The dynamic temporary shadow is formed by movable devices and temporary facilities, with its position and shape changing randomly, and its duration and range of occlusion uncertain; The time-varying shadow is formed by vegetation and changes gradually with the seasons and day and night, with the intensity and range of shading fluctuating regularly. The superimposed complex shadow is formed by superimposing two or more single shadows from the static fixed shadow, the dynamic temporary shadow, and the temporal gradient shadow, resulting in a shadow with complex shape and uneven occlusion intensity.

3. The photovoltaic layout method for thermal power plants according to claim 2, characterized in that, In the step of constructing adaptive artificial intelligence models for the four types of shadows and performing quantitative analysis of each type of shadow based on the artificial intelligence models, for the static fixed shadow, the three-dimensional parameters of the static building and fixed equipment, the time series data of the solar trajectory, and the reflectivity of the component material in the fusion model are obtained, input into the trained CNN model, and the first occlusion range and the actual solar irradiance are output. Based on the actual solar irradiance, the occlusion duration and occlusion intensity are calculated. For the dynamic temporary shadow, the three-dimensional parameters, historical movement trajectory data, operation and maintenance scheduling plan, and solar trajectory time series data of the mobile device and temporary facility in the fusion model are obtained. The trained YOLO+LSTM fusion model is input and the mobile device position, device movement speed, solar altitude angle and azimuth angle are output. Based on the mobile device position, device movement speed, solar altitude angle and azimuth angle, the second occlusion range is determined and the occlusion probability and occlusion duration are calculated. The occlusion is emphasized by the occlusion probability. For the time-varying shading, time-series remote sensing images, GIS terrain data, meteorological data, vegetation parameters, and solar trajectory time-series data are acquired, input into the trained Transformer model, and output vegetation leaf area index and vegetation height. Based on the vegetation leaf area index, vegetation height, solar altitude angle, and azimuth angle, the third shading range is determined, the shading intensity is calculated, and the shading duration is statistically analyzed. For the superimposed complex shadow, obtain the occlusion range, occlusion duration, and occlusion intensity of the shadow output by the CNN model, YOLO+LSTM fusion model, and Transformer model, as well as the weight coefficients of static fixed shadow, dynamic temporary shadow, and temporally gradation shadow. Input the trained attention mechanism fusion model and output the fourth occlusion range, occlusion duration, and occlusion intensity of the superimposed shadow.

4. The photovoltaic layout method for thermal power plants according to claim 3, characterized in that, The step of stitching together the shadow analysis results processed by each artificial intelligence model to form the overall shadow analysis result includes: Weights are assigned to the shadow analysis results processed by each artificial intelligence model. Specifically, type weights are assigned based on the degree of impact of various types of shadows on photovoltaic power generation and the stability of shading; intensity weights are assigned based on the shading intensity of various types of shadows; and region adaptation weights are assigned based on the type and priority of photovoltaic layout areas. The comprehensive weight of each type of shadow in the corresponding region is calculated, and the shadow analysis results processed by each artificial intelligence model are spatially superimposed based on the fusion model to obtain the stitched occlusion range. For the overlapping areas in the stitched occlusion range, the shadow type with the highest comprehensive weight is selected as the core shadow, and the occlusion intensity of other overlapping shadows is merged to obtain the merged occlusion intensity. For non-overlapping areas within the stitched occlusion range, the corresponding shadow analysis results are directly retained; Based on the occlusion intensity after fusion, the total occlusion duration is adjusted to achieve linkage matching between occlusion intensity and occlusion duration.

5. The photovoltaic layout method for thermal power plants according to claim 4, characterized in that, In the step of calculating the comprehensive weight of each type of shadow in the corresponding region, the formula for calculating the comprehensive weight is as follows: ； in, Let be the combined weight of the i-th type of shading in the region (x,y). Let i be the type weight of the i-th type of shadow. Let i be the intensity weight of the i-th type of shadow. The adaptation weights for the region (x,y).

6. The photovoltaic layout method for thermal power plants according to claim 5, characterized in that, In the step of selecting the shadow type with the highest overall weight as the core shadow for the overlapping area within the stitched occlusion range, and simultaneously fusing the occlusion intensity of other overlapping shadows to obtain the fused occlusion intensity, the calculation formula for the fused occlusion intensity of the overlapping area is as follows: ； in, Let represent the occlusion intensity of the overlapping region (x, y) after fusion at time t. Let be the combined weight of the i-th type of shading in the region (x,y). Let be the occlusion intensity of the i-th type of shadow.

7. The photovoltaic layout method for thermal power plants according to claim 6, characterized in that, In the step of correcting the total occlusion duration based on the occlusion intensity after fusion, the correction formula is as follows: ； ； in, This is the corrected total occlusion duration. The total duration of occlusion monitored in the field, where k is the correction coefficient. Let be the total annual occlusion duration for region (x, y). This represents the total monitoring time for the entire year. Let be the duration of occlusion of the i-th type of shadow at time t. The occlusion intensity after fusion.

8. A photovoltaic layout system for a thermal power plant, characterized in that, For implementing the photovoltaic layout method of a thermal power plant as described in any one of claims 1-7, the system comprises: The fusion module is used to acquire BIM data and GIS data, perform preprocessing, and then deeply fuse the preprocessed BIM data and GIS data to build a fusion model. The segmentation module is used to determine the shadow region based on the fusion model and divide the shadow region into four categories, including static fixed shadow, dynamic temporary shadow, temporal gradient shadow and superimposed complex shadow; The module is used to build suitable artificial intelligence models for four types of shadows, and to perform quantitative analysis of each type of shadow based on the artificial intelligence models. Specifically, a CNN model is used for static fixed shadows; a YOLO+LSTM fusion model is used for dynamic temporary shadows; a Transformer model is used for temporally changing shadows; and an attention mechanism fusion model is used for superimposed complex shadows. The stitching module is used to stitch together the shadow analysis results processed by each artificial intelligence model to form the overall shadow analysis result; The photovoltaic array layout determination module is used to determine the optimal photovoltaic array layout based on the overall results of the shadow analysis, combined with GIS spatial analysis, photovoltaic module technical parameters, and engineering constraints, through a multi-objective optimization algorithm.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the photovoltaic layout method for thermal power plants as described in any one of claims 1-7.

10. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the photovoltaic layout method for a thermal power plant as described in any one of claims 1-7.

Images