Method for optimizing layout of photovoltaic equipment for road engineering based on solar energy conversion efficiency

Abstract

The present application relates to a highway engineering photovoltaic device layout optimization method based on solar energy conversion efficiency, relates to the field of photovoltaic device layout optimization, comprising: randomly arranging a preset photovoltaic component in a constraint area to obtain two groups of layout information and corresponding position and angle parameters. Comparing the two positions, a difference irradiance prediction model is obtained, and two groups of predicted irradiance are analyzed. The two groups of layout information are sorted according to irradiance, and the sorting result is updated. When the convergence condition is met, the layout scheme is configured according to the sorting result, and the highway photovoltaic device layout is managed. The present application solves the technical problem of low optimization efficiency in the layout optimization of photovoltaic devices in highway engineering due to the complexity of irradiance analysis. By comparing the photovoltaic components in different layout modes, only the difference irradiation type is analyzed, the calculation complexity is simplified under the premise of ensuring the optimization effect of photovoltaic device layout, and the technical effect of improving the optimization efficiency of photovoltaic device layout in highway engineering is achieved.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic equipment layout optimization, and in particular to a method for optimizing the layout of photovoltaic equipment in highway engineering based on solar energy conversion efficiency. Background Technology

[0002] In existing technologies, optimizing the layout of photovoltaic (PV) equipment in highway engineering typically requires a comprehensive irradiance analysis of the entire PV array to assess its solar energy conversion efficiency. However, irradiance analysis is highly nonlinear, complex in its calculations, and demands significant computing power. Especially in large-scale highway engineering PV equipment layout optimization, traditional methods often result in low processing efficiency, making it difficult to meet the optimization speed requirements of practical applications.

[0003] To address the aforementioned issues, existing technologies have proposed several improvement schemes. For example, a simplified irradiance model can be employed, sacrificing some accuracy for improved computational efficiency; or parallel computing techniques can be used to perform irradiance analysis simultaneously through multiple processing units, thereby shortening the overall optimization time. However, these schemes have limitations in optimization effectiveness or in terms of implementation cost and complexity.

[0004] Therefore, how to improve optimization efficiency and simplify calculation complexity while ensuring the optimization effect of photovoltaic equipment layout has become an urgent problem to be solved in the field of photovoltaic equipment layout optimization in highway engineering. Summary of the Invention

[0005] This invention addresses the technical problem of low optimization efficiency in the layout optimization of photovoltaic equipment in existing highway engineering projects due to the complexity of irradiance analysis, and provides a method for optimizing the layout of photovoltaic equipment in highway engineering based on solar energy conversion efficiency.

[0006] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0007] This invention provides a method for optimizing the layout of photovoltaic (PV) equipment in highway engineering based on solar energy conversion efficiency. The method includes: randomly arranging preset-model PV modules based on horizontal and vertical layout constraint areas within a constrained region for PV equipment layout in the highway engineering project, obtaining first and second PV module layout information; obtaining the first layout position, first layout tilt angle, and first layout azimuth angle of the first PV module layout information, and the second layout position, second layout tilt angle, and second layout azimuth angle of the second PV module layout information; configuring differential irradiance types based on the first and second layout positions to obtain a differential irradiance prediction model; and configuring the differential irradiance type based on the differential irradiance type. An irradiance prediction model performs irradiance analysis based on the first layout location, the first layout tilt angle, and the first layout azimuth angle to obtain a first predicted irradiance. A differential irradiance prediction model performs irradiance analysis based on the second layout location, the second layout tilt angle, and the second layout azimuth angle to obtain a second predicted irradiance. Based on the first and second predicted irradiances, the layout information of the first and second photovoltaic modules is sorted from largest to smallest, and the layout information ranking result is updated. When the photovoltaic module layout information meets a preset convergence condition, a photovoltaic layout scheme is configured based on the layout information ranking result to manage the layout of photovoltaic equipment in highway engineering.

[0008] The beneficial effects of this invention are as follows: Based on the horizontal and vertical layout constraints of the photovoltaic equipment layout area in highway engineering, photovoltaic modules of a preset model are randomly laid out to obtain first and second photovoltaic module layout information, resulting in two layout schemes for individual photovoltaic modules. These schemes can be two photovoltaic modules in different locations or two modules in the same location but with different placement angles, providing a basis for subsequent comparison and optimization. The invention obtains the first layout position, first layout tilt angle, and first layout azimuth angle of the first photovoltaic module layout information, and the second layout position, second layout tilt angle, and second layout azimuth angle of the second photovoltaic module layout information, preparing for subsequent irradiance analysis. Based on the first and second layout positions, differential irradiance types are configured to obtain a differential irradiance prediction model. By comparing the positional differences between the two layout schemes, the differential areas requiring irradiance analysis are determined, and corresponding prediction models are established, reducing unnecessary computation. Based on the differential irradiance prediction model, irradiance analysis is performed using the first layout location, first layout tilt angle, and first layout azimuth angle to obtain the first predicted irradiance. Using the differential irradiance prediction model, irradiance prediction is then performed on the layout information of the first photovoltaic module to obtain the corresponding estimated irradiance value. Similarly, based on the second layout location, second layout tilt angle, and second layout azimuth angle, irradiance analysis is performed using the same model to obtain the second predicted irradiance. Likewise, irradiance prediction is performed on the layout information of the second photovoltaic module to obtain the corresponding estimated value. Based on the first and second predicted irradiance values, the layout information of the first and second photovoltaic modules is ranked from largest to smallest, and the layout information ranking is updated. By comparing the irradiance prediction results of the two layout schemes, they are ranked, and the layout scheme with the larger irradiance and higher solar energy conversion efficiency is selected. When the photovoltaic module layout information meets the preset convergence condition, a photovoltaic layout scheme is configured based on the layout information ranking result for the management of photovoltaic equipment layout in highway engineering. By comparing photovoltaic modules with different layouts pairwise and analyzing only the differences in irradiance types, the computational complexity is significantly reduced and the optimization efficiency is improved while ensuring the optimization effect of photovoltaic equipment layout. Attached Figure Description

[0009] Figure 1 A flowchart illustrating the method for optimizing the layout of photovoltaic equipment in highway engineering based on solar energy conversion efficiency provided by this invention;

[0010] Figure 2 This is a schematic diagram of the structure of the electronic device provided by the present invention;

[0011] Figure 3 This is a schematic diagram of the structure of a computer-readable storage medium provided by the present invention.

[0012] In the attached diagram, the components represented by each number are as follows:

[0013] Electronic device 100, memory 110, processor 120, first computer program 111, computer-readable storage medium 200, second computer program 211. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0016] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0017] Example 1:

[0018] like Figure 1 As shown, this embodiment of the invention provides a method for optimizing the layout of photovoltaic equipment in highway engineering based on solar energy conversion efficiency, including:

[0019] S100: Based on the horizontal and vertical layout constraints of the photovoltaic equipment layout constraint area in the highway engineering, the photovoltaic modules of the preset model are randomly laid out to obtain the layout information of the first photovoltaic module and the layout information of the second photovoltaic module.

[0020] Specifically, the photovoltaic (PV) equipment layout constraint area in highway engineering is a three-dimensional spatial area defined based on the actual conditions of the highway project, within which PV modules need to be rationally arranged. First, based on the PV equipment layout constraint area, horizontal and vertical layout constraint areas are delineated as optional installation spaces for PV module layout optimization. The horizontal and vertical layout constraint areas are determined according to the actual conditions of the highway project and can be areas suitable for PV module installation, such as both sides of the highway, roadbed slopes, and interchange areas. By clearly defining the layout boundaries in the horizontal and vertical directions, the layout constraints become more specific and quantifiable.

[0021] After determining the horizontal and vertical layout constraints, a preset model of photovoltaic (PV) modules is selected as the objects to be laid out. These preset models of PV modules have standard dimensions, rated power, conversion efficiency, and other parameters. Selecting a specific model of module as the layout object allows for the evaluation of the advantages and disadvantages of different layout methods under uniform specifications.

[0022] Then, the photovoltaic modules of the preset model are randomly arranged within the constrained area, automatically generating two different photovoltaic module layout methods. These can be two photovoltaic modules in different locations, or two modules in the same location but with different installation angles. For example, using a random number generation function, within the horizontal and vertical layout constraint areas, two sets of photovoltaic module installation position coordinates, orientation angles (azimuth angles), tilt angles, and other layout parameters are randomly generated, thus forming the first photovoltaic module layout information and the second photovoltaic module layout information.

[0023] By pre-delineating horizontal and vertical layout constraint areas and randomly arranging specific types of photovoltaic modules within them, two sets of photovoltaic module layout information are automatically generated, providing diverse alternative solutions for subsequent comparison and optimization. This is the foundation for achieving high-efficiency photovoltaic array layout.

[0024] S200: Obtain the first layout position, first layout tilt angle, and first layout azimuth angle of the first photovoltaic module layout information, and obtain the second layout position, second layout tilt angle, and second layout azimuth angle of the second photovoltaic module layout information.

[0025] Specifically, after obtaining the layout information of the first and second photovoltaic modules, the following steps are taken: First, the first layout position coordinates, first layout tilt angle, and first layout azimuth angle data of the photovoltaic modules under this layout method are extracted from the first photovoltaic module layout information. The first layout position refers to the spatial coordinates of the photovoltaic module within the three-dimensional constrained area, expressed in latitude, longitude, elevation, or relative coordinate system, used to determine the horizontal and vertical position of the module. The first layout tilt angle is the angle between the photovoltaic module and the horizontal plane. In actual installation, the tilt angle can be optimized according to local latitude, terrain, and other factors to allow the module to receive more solar radiation. The first layout azimuth angle is the deflection angle of the photovoltaic module's projection onto the horizontal plane, indicating the orientation of the module. Generally, due south is taken as 0°, and clockwise directions are successively positive.

[0026] Similarly, from the second photovoltaic module layout information, the second layout position, second layout tilt angle, and second layout azimuth angle of the photovoltaic modules under this layout method also need to be extracted. Through information extraction, the two sets of random layout information are further analyzed into a series of specific position and angle parameters, describing the spatial state of the photovoltaic modules within the three-dimensional layout constraint area.

[0027] By extracting the specific location and angle parameters of photovoltaic modules from the photovoltaic module layout information, the spatial attributes of photovoltaic modules are quantitatively described. Two sets of alternative random layout schemes are characterized in a structured data form, providing the necessary data foundation for subsequent steps such as irradiance prediction and layout optimization.

[0028] S300: Configure the differential irradiance type according to the first layout position and the second layout position to obtain the differential irradiance prediction model.

[0029] Specifically, the first and second layout positions are key parameters describing the spatial distribution of photovoltaic (PV) modules in the first and second PV module layout information, respectively. By comparing the position coordinates of PV modules in the two layouts, differences in spatial distribution can be observed. In actual PV power generation applications, the type and intensity of solar irradiance received by PV modules at different locations may vary, mainly manifested in direct irradiance, diffused irradiance, and reflected irradiance. Direct irradiance refers to sunlight directly hitting the surface of the PV module at a certain angle of incidence; the intensity of direct irradiance received by PV modules at different positions and angles will vary. Diffuse irradiance refers to sunlight being scattered in the atmosphere and clouds; scattered light from various directions eventually hits the surface of the PV module. The spatial distribution of diffused irradiance is relatively uniform, but there will still be some differences between different locations. Reflected irradiance refers to sunlight hitting the ground or other object surfaces and then reflecting onto the PV modules; the intensity distribution of reflected irradiance may vary between PV modules at different locations due to differences in the surrounding environment and terrain.

[0030] Therefore, based on the differences between the first and second layout positions, the differences in direct irradiance, diffused irradiance, and reflected irradiance between the two sets of photovoltaic modules are determined. The types of irradiance differences that significantly impact the efficiency of the two layouts are identified, resulting in configuration results for these irradiance types. After clarifying the irradiance difference types of the two layouts, a corresponding irradiance difference prediction model is selected. A irradiance difference prediction model is a pre-established parameterized mathematical model for a specific type of irradiance difference. By inputting parameters such as the position and angle of the photovoltaic modules, it can quickly predict the irradiance distribution characteristics of the modules.

[0031] By leveraging the spatial differences in the two sets of random layouts, the main types of irradiance variations are identified, allowing for the invocation of targeted irradiance prediction models and simplifying the subsequent irradiance analysis process. This approach ensures prediction accuracy while avoiding redundant calculations, thus improving the efficiency of layout scheme evaluation.

[0032] S400: Based on the differential irradiance prediction model, irradiance analysis is performed on the first layout position, the first layout tilt angle, and the first layout azimuth angle to obtain the first predicted irradiance.

[0033] Specifically, the three parameters—first layout position, first layout tilt angle, and first layout azimuth angle—completely describe the spatial attributes of the photovoltaic modules in the first group of random layouts. The first layout position determines the installation coordinates of the photovoltaic modules within the three-dimensional constrained area; the first layout tilt angle represents the angle between the photovoltaic modules and the horizontal plane, affecting the effective solar irradiance receiving area; and the first layout azimuth angle represents the orientation of the photovoltaic modules projected onto the horizontal plane, relating to the amount of irradiance at different times.

[0034] The parameters of the first layout (first layout position, first layout tilt angle, first layout azimuth angle) are input into the differential irradiance prediction model. For photovoltaic modules with this layout, quantitative analysis and calculations are performed from three aspects: direct irradiance, diffuse irradiance, and reflected irradiance. The irradiance intensity received by the photovoltaic modules under this layout condition is predicted, resulting in the first predicted irradiance. The differential irradiance prediction model is selected based on the differences in the positions of the two layouts. Because this model fully considers the differences in irradiance distribution at different layout positions, it can specifically predict key irradiance influencing factors, thus efficiently completing the irradiance analysis task for the first layout with high accuracy.

[0035] By utilizing the differential irradiance prediction model, based on the specific layout parameters such as the location, tilt angle, and azimuth angle of the first group of random layouts, the predicted irradiance value of the photovoltaic modules under this layout is obtained, providing an important basis for subsequent layout evaluation and selection.

[0036] S500: Based on the differential irradiance prediction model, irradiance analysis is performed on the second layout position, the second layout tilt angle, and the second layout azimuth angle to obtain the second predicted irradiance.

[0037] Specifically, the differential irradiance prediction model uses parameters such as the position coordinates, tilt angle, and orientation of the photovoltaic modules in the second layout (i.e., the position of the second layout, the tilt angle of the second layout, and the azimuth angle of the second layout) to derive the irradiance distribution prediction of the photovoltaic modules and obtain the second predicted irradiance.

[0038] Because the two random layouts differ in spatial distribution characteristics, using the specific parameters of the second layout as input will result in a certain deviation between the predicted irradiance and the first predicted irradiance. This reflects the impact of different layout methods on photovoltaic power generation efficiency and provides a reference for subsequent comparison and selection of optimal layout schemes. In subsequent steps, the first and second predicted irradiance indicators can be used to determine which layout scheme achieves better photovoltaic power generation efficiency, thereby guiding the optimization and selection of layout schemes.

[0039] By predicting two different random layouts using a differential irradiance prediction model, we can better adapt to different layout conditions, improve the accuracy of irradiance prediction, and provide key data support for the optimized layout of photovoltaic arrays.

[0040] S600: Based on the first predicted irradiance and the second predicted irradiance, update the layout information sorting result of the first photovoltaic module layout information and the second photovoltaic module layout information from largest to smallest.

[0041] Specifically, the first and second predicted irradiance reflect the photovoltaic power generation efficiency corresponding to the two layout schemes. Therefore, the predicted irradiance is used as the criterion to rank the two random layouts and select the relatively superior layout scheme.

[0042] First, the order of preference for the two layout schemes is determined based on the magnitudes of the first and second predicted irradiance values. If the first predicted irradiance is greater than the second, the first random layout scheme achieves higher photovoltaic power generation efficiency and thus has a higher priority; conversely, the second layout scheme has a higher priority. Based on this comparison, the layout information for the first and second photovoltaic modules is sorted in descending order of predicted irradiance. Layout information corresponding to larger predicted irradiance is ranked first, and layout information corresponding to smaller predicted irradiance is ranked last. Through this sorting process, the order of preference for the two layout schemes is quantified into an ordered sequence, denoted as the layout information sorting result.

[0043] By comparing the first and second predicted irradiance, two quantitative indicators, a set of evaluation criteria for the merits of two random layout schemes was established. The layout information ranking results were updated, which intuitively reflected the differences in photovoltaic power generation efficiency obtained by using different photovoltaic module layout methods. This laid the foundation for finally determining the optimal layout scheme to be used in actual projects.

[0044] S700: When the photovoltaic module layout information meets the preset convergence conditions, a photovoltaic layout scheme is configured based on the sorting result of the layout information to manage the layout of photovoltaic equipment in highway engineering.

[0045] Specifically, firstly, a preset convergence condition is set as the standard for judging whether the alternative layout scheme has reached the optimal level. Only when the photovoltaic module layout information meets the preset convergence condition (e.g., the photovoltaic module layout information has covered all possible layouts within the constrained area of ​​the photovoltaic equipment layout in the highway project) is the scheme considered sufficiently optimized and can be used as the final layout selection. Here, the photovoltaic module layout information refers to all photovoltaic module layout information generated by randomly placing preset model photovoltaic modules.

[0046] After determining that the photovoltaic (PV) module layout information meets the preset convergence conditions, the overall PV layout scheme is further optimized based on the obtained layout information ranking results. Since the layout information ranking results already include the merit levels of different individual PV module layouts, this ranking can be used to determine the PV layout scheme with the best overall evaluation. Once the PV layout scheme is obtained, it is applied to the actual PV equipment layout management work in highway engineering. Specifically, based on the key parameter information such as location and angle provided by the PV layout scheme, the actual installation tasks of each PV module are planned and arranged, thereby achieving efficient optimization of the PV equipment layout.

[0047] Furthermore, embodiments of this application also include:

[0048] S310: Obtain the first direct irradiation influencing factor, the first diffuse irradiation influencing factor, and the first reflected irradiation influencing factor at the first layout location;

[0049] S320: Obtain the second direct irradiance influence factor, the second diffuse irradiance influence factor, and the second reflected irradiance influence factor for the second layout location;

[0050] S330: Compare the first direct irradiation influencing factor and the second direct irradiation influencing factor to obtain the direct irradiation factor deviation;

[0051] S340: Compare the first scattered radiation influencing factor and the second scattered radiation influencing factor to obtain the scattered radiation factor deviation;

[0052] S350: Compare the first reflected irradiance influencing factor and the second reflected irradiance influencing factor to obtain the reflected irradiance factor deviation;

[0053] S360: When the deviation of the direct irradiance factor is greater than or equal to the direct irradiance factor deviation threshold, add direct irradiance to the differential irradiance type; when the deviation of the scattered irradiance factor is greater than or equal to the scattered irradiance factor deviation threshold, add scattered irradiance to the differential irradiance type; when the deviation of the reflected irradiance factor is greater than or equal to the reflected irradiance factor deviation threshold, add reflected irradiance to the differential irradiance type.

[0054] S370: Based on the differential irradiance type, retrieve the differential irradiance prediction model.

[0055] In a preferred embodiment, firstly, the influencing factors of direct irradiance, diffuse irradiance, and reflected irradiance at the first and second layout locations are obtained, namely, the first direct irradiance influencing factors, the first diffuse irradiance influencing factors, and the first reflected irradiance influencing factors at the first layout location, and the second direct irradiance influencing factors, the second diffuse irradiance influencing factors, and the second reflected irradiance influencing factors at the second layout location. These influencing factors comprehensively reflect the illumination conditions and radiation characteristics of a specific layout location. For the influencing factors of direct irradiance, parameters such as solar position (including azimuth and altitude angles), atmospheric transparency, photovoltaic module installation angle (tilt angle and azimuth angle), solar constant, day length, and solar altitude angle are mainly considered, which together determine the incident intensity and angular distribution of direct sunlight at different layout locations. For the influencing factors of diffuse irradiance, key attention needs to be paid to atmospheric composition parameters (such as aerosol and water vapor content), cloud cover and cloud type, solar altitude angle, surface albedo, and the specific installation location and angle of the photovoltaic modules. These factors work together to affect the intensity distribution and reception efficiency of diffuse radiation at different layout locations. Factors influencing reflected irradiance mainly include the reflectivity of the surrounding environment (such as the ground, buildings, and water bodies), the installation angle of the photovoltaic modules, the distribution of surrounding topography and features, and the solar altitude and azimuth angles. These factors collectively affect the directionality, uniformity, and amount of reflected irradiance received at different layout locations. By acquiring these three types of irradiance influencing factors, the differences in irradiance conditions at different layout locations can be described more comprehensively.

[0056] Then, the differences between the first and second layout locations in terms of three types of irradiance influencing factors were compared separately to quantitatively assess the irradiance condition deviation between the two locations. This yielded three key parameters: deviation of direct irradiance, deviation of diffuse irradiance, and deviation of reflected irradiance. This enabled a quantitative characterization and comparison of the differences in layout locations, providing a basis for subsequent determination of the differential irradiance type. Next, by setting deviation thresholds (direct irradiance, diffuse irradiance, and reflected irradiance), the significance of the deviation for each type of irradiance influencing factor was determined. When the deviation of a certain type of irradiance factor exceeded the threshold, it indicated a significant difference between the two layout locations in that type of irradiance. This was included in the differential irradiance type and became the focus of subsequent differential analysis, thereby identifying the factors that contributed most to the irradiance difference between the two layout locations and laying the foundation for targeted application of the prediction model. Specifically, when the deviation of direct irradiance is greater than or equal to the direct irradiance deviation threshold, direct irradiance is added to the differential irradiance type; when the deviation of scattered irradiance is greater than or equal to the scattered irradiance deviation threshold, scattered irradiance is added to the differential irradiance type; and when the deviation of reflected irradiance is greater than or equal to the reflected irradiance deviation threshold, reflected irradiance is added to the differential irradiance type. Then, based on the determined differential irradiance type, the corresponding differential irradiance prediction model is selectively retrieved, providing customized calculation tools for subsequent irradiance analysis and layout optimization. This avoids comprehensive prediction for all irradiance types, reduces unnecessary computation, and improves overall solution efficiency.

[0057] By accurately identifying key irradiance difference factors between different layout locations and selecting matching prediction models, a foundation is laid for improving the targeting and computational efficiency of irradiance prediction and layout optimization.

[0058] Furthermore, the process for obtaining the influencing factors of the first reflected irradiation includes:

[0059] S311: The first layout position includes the first grounding position of the first photovoltaic equipment support, and a first ground feature with the first grounding position as the center and a preset length as the radius is obtained;

[0060] S312: Set the first ground feature as the first reflected radiation influencing factor.

[0061] In a preferred embodiment, firstly, the first grounding position of the first photovoltaic device support corresponding to the first layout position is determined. The support can be a photovoltaic module mounting frame, pile foundation, etc., and its grounding position is the connection point between the support and the ground, reflecting the spatial installation position of the photovoltaic module. A circular area, called the first ground feature area, is delineated with the first grounding position as the center and a predetermined length as the radius. This area characterizes the ground environment characteristics within a certain range around the photovoltaic module and has a direct impact on the reflected irradiance incident on the photovoltaic module.

[0062] Then, the obtained first ground feature is directly set as the first influencing factor of reflected irradiance. This is because the physical properties of the ground environment, such as material, roughness, and color, determine the reflectivity and reflection direction distribution of sunlight after it hits the ground, thus affecting the intensity of reflected irradiance received by the photovoltaic module. The first ground feature area centered on the first grounding location comprehensively reflects the reflected irradiance characteristics of the surrounding surface and can be used as an important parameter for calculating the intensity of reflected irradiance.

[0063] By establishing a direct link between the local layout characteristics of photovoltaic modules and the intensity of reflected irradiance, a convenient and efficient scheme for obtaining the influencing factors of reflected irradiance is proposed. Compared with traditional methods, this scheme can more precisely characterize the differences in reflected irradiance under different layout configurations, providing support for improving the overall power generation efficiency of photovoltaic arrays.

[0064] Furthermore, the differential irradiance prediction model includes a direct irradiance prediction model, a diffuse irradiance prediction model, and a reflected irradiance prediction model. The training methods for the direct irradiance prediction model, the diffuse irradiance prediction model, and the reflected irradiance prediction model are the same. Specifically, the training steps for the direct irradiance prediction model include:

[0065] S371: Based on the rated range of irradiation influencing factors, randomly configure the data recording of direct irradiation influencing factors;

[0066] S372: Using the data recorded on the influencing factors of direct irradiation, collect irradiation samples for photovoltaic module models to obtain a set of direct irradiance recorded values;

[0067] S373: Perform a central tendency assessment on the set of direct irradiance recorded values ​​to obtain direct irradiance identification data;

[0068] S374: Add the recorded data of direct irradiation influencing factors to the set of recorded data of direct irradiation influencing factors, and add the data of direct irradiance labeling to the set of direct irradiance labeling data;

[0069] S375: Train the direct irradiance prediction model based on the direct irradiance identification data set and the direct irradiance influencing factor record data set.

[0070] In a preferred embodiment, the differential irradiance prediction model includes a direct irradiance prediction model, a diffuse irradiance prediction model, and a reflected irradiance prediction model, and these three sub-models use the same training steps.

[0071] Taking the training of a direct irradiance prediction model as an example, firstly, based on the known nominal ranges of irradiance influencing factors, a set of recorded data on these factors is randomly generated. These influencing factors include solar position, atmospheric transparency, photovoltaic module tilt angle, and azimuth angle. By randomly sampling within their nominal ranges, a series of representative combinations of influencing factors can be obtained for subsequent irradiance sample collection and model training. Next, using the generated recorded data on direct irradiance influencing factors, actual irradiance samples are collected from photovoltaic modules of a specified model. This can be achieved, for example, through physical measurement or simulation, with the aim of obtaining data on the actual direct irradiance intensity received by the photovoltaic modules under different combinations of influencing factors, forming a set of recorded direct irradiance values.

[0072] Subsequently, statistical analysis was performed on the collected direct irradiance records to assess the central tendency of the data distribution, thereby obtaining comprehensive direct irradiance label data. This label data can be the mean, median, or other indicators of central tendency of the irradiance records, used to characterize the overall level of direct irradiance under the current combination of influencing factors. By summarizing a set of irradiance records into a representative label value, subsequent model training and application are facilitated. Next, the generated direct irradiance influencing factor records and the obtained direct irradiance label data were added to their respective datasets: the direct irradiance influencing factor records dataset and the direct irradiance label dataset. These two datasets are used to store the input feature data and target output data required for model training, respectively. By continuously accumulating more sample records, the coverage and diversity of the training data can be expanded, laying a data foundation for improving the model's generalization ability. Then, using the obtained direct irradiance influencing factor records dataset as input and the direct irradiance label dataset as supervision signal, a direct irradiance prediction model was trained using a machine learning algorithm. This model can predict the direct irradiance label value under corresponding conditions based on the combination of input influencing factors. Machine learning algorithms such as neural networks, support vector machines, and decision trees can all be used to train this model.

[0073] Through the above-mentioned training process for the differential irradiance prediction model, the influencing factors of irradiance and measured irradiance data are fully utilized to construct an efficient and accurate prediction model. This model can quantitatively analyze the irradiance differences of photovoltaic modules under different layout methods, providing an important basis for subsequent layout optimization decisions.

[0074] Furthermore, embodiments of this application also include:

[0075] S410: Based on the differential irradiance prediction model, extract a first differential irradiance prediction model, wherein the first differential irradiance prediction model has a first differential irradiance type label, and the first differential irradiance type label has a first weight.

[0076] S420: Based on the first differential irradiance type label, collect the first irradiance influencing factor characteristic data of the first layout location, process the first irradiance influencing factor characteristic data through the first differential irradiance prediction model, and generate the first type of predicted irradiance.

[0077] S430: Until the Nth differential irradiance prediction model is extracted according to the differential irradiance prediction model, wherein the Nth differential irradiance prediction model has an Nth differential irradiance type label and the Nth differential irradiance type label has an Nth weight;

[0078] S440: Based on the Nth differential irradiance type label, collect the Nth irradiance influencing factor characteristic data of the first layout location, process the Nth irradiance influencing factor characteristic data through the Nth differential irradiance prediction model, and generate the Nth type predicted irradiance, where 3≥N≥0, and N is an integer;

[0079] S450: Based on the first weight up to the Nth weight, the normalized values ​​of the first type of predicted irradiance up to the Nth type of predicted irradiance are weighted to obtain an N-ary irradiance feature array.

[0080] S460: The N-ary irradiance feature array is processed using the irradiance fusion analysis model to generate the first predicted irradiance.

[0081] In a preferred embodiment, firstly, based on the constructed differential irradiance prediction model, a first differential irradiance prediction model is extracted for subsequent analysis. This model is specifically designed for specific types of irradiance differences (such as direct irradiance differences, diffuse irradiance differences, etc.), and therefore has a corresponding irradiance type label, i.e., the first differential irradiance prediction model. Simultaneously, a weight value is assigned to the type label of this model to characterize the relative importance of the irradiance difference type's contribution to the total irradiance. Next, based on the first differential irradiance type label of the first differential irradiance prediction model, irradiance influencing factor characteristic data for a matching first layout location is collected, i.e., first irradiance influencing factor characteristic data. This characteristic data can include information from various aspects such as solar position, meteorological conditions, and topography, acquired through various sensors or databases. Then, the collected first irradiance influencing factor characteristic data is input into the first differential irradiance prediction model for processing, thereby obtaining the irradiance prediction value for a specific irradiance difference type under this layout condition, i.e., the first type predicted irradiance.

[0082] Repeat the above process until all differential irradiance prediction models have been traversed. The total number of differential irradiance prediction models is N (3 ≥ N ≥ 0, N is an integer). For the Nth differential irradiance prediction model, its irradiance type label and corresponding weights are obtained, relevant feature data of the first layout position are collected, and the Nth type of predicted irradiance is generated through this model. After obtaining the predicted irradiance for all types, these predicted values ​​are weighted and fused. First, the predicted irradiance for types 1 to NT are normalized to unify their numerical range to the [0, 1] interval, facilitating subsequent weighting. Then, using the obtained type label weights, the normalized predicted irradiance is weighted and summed to obtain an irradiance feature array containing N elements. This array integrates the prediction results of different irradiance difference types, reflecting the relative contribution of each type of irradiance to the total irradiance. Finally, the obtained N-ary irradiance feature array is input into a pre-trained irradiance fusion analysis model for processing. This model is a machine learning model (such as neural networks, decision trees, etc.). By learning from a large amount of historical data, it has mastered the pattern of mapping feature combinations of different types of irradiance to the predicted value of total irradiance. After calculation by the fusion analysis model, the predicted value of total irradiance for the first layout method under the current parameter settings is obtained, that is, the first predicted irradiance.

[0083] By modeling and weighting different types of irradiance differences separately, the characteristics of irradiance distribution under complex environments are depicted more precisely. Normalization and weighted fusion are introduced, allowing for a comprehensive evaluation of irradiance predictions of different types on a unified scale, reflecting the relative importance of each type of irradiance. Furthermore, the irradiance fusion analysis model is used to perform secondary processing on the weighted features, further improving the accuracy of total irradiance prediction.

[0084] Furthermore, the steps for constructing the irradiance fusion analysis model include:

[0085] S461: Configure the M-element irradiance array to record the dataset and the irradiance detection true value dataset according to the photovoltaic module model;

[0086] S462: Using the irradiance detection true value dataset as supervision and the M-gram irradiance array record dataset as input, train a first-level irradiance fusion analysis model;

[0087] S463: Extract the set of output deviation vectors from the first-level irradiance fusion analysis model;

[0088] S464: Perform systematic error vector analysis on the set of output deviation vectors to obtain the first-level systematic error vector;

[0089] S465: When the magnitude of the first-level system error vector is greater than or equal to the magnitude threshold, the negative number of the first-level system error vector is set as the second-level irradiance label dataset.

[0090] S466: Using the secondary irradiance identifier dataset as supervision and the M-gram irradiance array record dataset as input, train the secondary irradiance fusion analysis model;

[0091] S467: Until the magnitude of the Q-level system error vector is less than the magnitude threshold, perform output summation and full connection on the first-level irradiance fusion analysis model, the second-level irradiance fusion analysis model, and up to the Q-level irradiance fusion analysis model to obtain the irradiance fusion analysis model.

[0092] In a preferred embodiment, firstly, two basic datasets are configured according to the photovoltaic module model to be analyzed. One is an M-ary irradiance array dataset, containing a large number of irradiance feature samples under different conditions, each sample represented by an irradiance feature vector composed of M components. The other is an irradiance detection ground truth dataset, recording the actual irradiance measurement value corresponding to each irradiance feature sample, serving as label data for model training. The configuration of these two datasets provides the necessary data foundation for subsequent model training. Next, an initial irradiance fusion analysis model, called the first-level irradiance fusion analysis model, is trained using the configured datasets. Specifically, the M-ary irradiance array dataset is used as the model input, and the irradiance detection ground truth dataset is used as the supervision signal. Parameter optimization and model training are performed using machine learning algorithms such as neural networks and decision trees. The trained first-level irradiance fusion analysis model can initially realize the mapping relationship from irradiance features to actual irradiance.

[0093] To further improve the model's prediction accuracy, an iterative optimization process is introduced. First, the prediction results of the first-level irradiance fusion analysis model on the training dataset are compared with the true labels, resulting in a series of output bias vectors, i.e., the set of output bias vectors, reflecting the current model's prediction error distribution. Then, systematic error analysis is performed on these output bias vectors, evaluating the overall bias characteristics of the model output using metrics such as mean and variance, resulting in a comprehensive first-level systematic error vector. This vector characterizes the average prediction bias of the first-level model across the entire sample space. Subsequently, it is determined whether the magnitude of the first-level systematic error vector exceeds a preset magnitude threshold. If it exceeds the threshold, it indicates that the current model's prediction error is large and further optimization is needed. At this point, the first-level systematic error vector is inverted, creating a new irradiance label dataset, called the second-level irradiance label dataset. This new label dataset actually reflects the prediction bias distribution of the first-level irradiance fusion analysis model and can be used to guide the training and optimization of the next-level model.

[0094] Subsequently, an M-ary irradiance array is used as input to record the dataset, and a secondary irradiance label dataset is used as the supervision signal to train a secondary irradiance fusion analysis model. This new model is structurally similar to the primary irradiance fusion analysis model, but the training objective is no longer to directly predict the true irradiance value, but rather to fit the prediction bias of the primary irradiance fusion analysis model. In this way, the secondary irradiance fusion analysis model can capture the pattern of the prediction error of the primary irradiance fusion analysis model and compensate for it. This process is repeated until the magnitude of the Q-level systematic error vector of the trained Q-level irradiance fusion analysis model is less than a magnitude threshold. This iterative optimization process can gradually reduce the model's prediction bias and improve the accuracy of irradiance prediction. Afterwards, all irradiance fusion analysis models from primary to Q levels are summed and fully connected to obtain the final irradiance fusion analysis model. Specifically, the outputs of each level of the model are weighted and summed, and the weights can be optimized using methods such as cross-validation. By combining and integrating these multi-level models, the predictive capabilities of each level can be fully utilized, and complementary advantages can be achieved under different error distributions, resulting in a more robust and better generalization irradiance fusion analysis model.

[0095] The irradiance fusion analysis model construction process fully considers the complexity and nonlinearity of photovoltaic irradiance prediction. By introducing multi-level iterative optimization and error compensation mechanisms, the model's prediction accuracy and adaptability can be continuously improved. Compared with traditional single-model methods, the multi-level fusion model adopted in this invention can better capture the inherent laws of irradiance data and exhibit stronger robustness under noise interference and abnormal deviations, providing a reliable decision-making basis for photovoltaic module layout.

[0096] Furthermore, embodiments of this application also include:

[0097] S710: Through the user terminal, configure the layout distance constraint variance, layout boundary constraint distance and layout constraint quantity for the layout constraint area of ​​the photovoltaic equipment in the highway project;

[0098] S720: Construct a fitness function, wherein the fitness function is the sum of the sequence numbers of the layout information sorting results;

[0099] S730: Using the layout distance constraint variance, the layout boundary constraint distance and the number of layout constraints as constraints, and based on the layout information sorting results, perform several random combinations to configure a set of photovoltaic equipment layout schemes;

[0100] S740: Based on the fitness function, perform particle swarm optimization on the set of photovoltaic equipment layout schemes to obtain a recommended photovoltaic equipment layout scheme;

[0101] S750: Manage the layout of photovoltaic equipment in highway engineering according to the recommended photovoltaic equipment layout scheme.

[0102] In a preferred embodiment, firstly, a series of constraint parameters are configured for the photovoltaic (PV) equipment layout constraint area of ​​the highway project via the user terminal, including the layout distance constraint variance, layout boundary constraint distance, and the number of layout constraints. The layout distance constraint variance limits the range of distance variation between adjacent PV modules, requiring the variance of the spacing between PV modules in the layout scheme to be greater than this constraint value. The layout boundary constraint distance defines the minimum distance between the PV module and the boundary of the layout area, ensuring that the module does not exceed the restricted area. The number of layout constraints specifies the total number of PV modules to be laid out. The introduction of these constraints provides necessary boundary restrictions for the subsequent generation of layout schemes. Next, a fitness function is constructed to evaluate the merits of candidate layout schemes. For example, the fitness function can be defined as the sum of the sequence numbers of the layout information ranking results. The layout information ranking results refer to the sequence obtained by ranking individual candidate PV modules from best to worst based on predicted irradiance. The smaller the sum of the sequence numbers, the more high-quality PV modules are placed at the top, and the higher the overall fitness.

[0103] Subsequently, under the set constraints, based on the obtained layout information ranking results, several candidate photovoltaic (PV) equipment layout schemes are generated through random combination, forming a set of PV equipment layout schemes. Random combination refers to randomly selecting several PV module layouts from the ranking results for combination to explore new layout possibilities. Simultaneously, it is ensured that the randomly generated schemes meet the requirements of layout distance constraint variance, layout boundary constraint distance, and the number of layout constraints to guarantee the feasibility of the schemes. Next, the obtained PV equipment layout scheme set is further optimized using a particle swarm optimization (PSO) algorithm. PSO is a heuristic search algorithm based on swarm intelligence, which searches for the optimal solution in the solution space by simulating the foraging behavior of a flock of birds. Each candidate PV module layout is regarded as a particle, and the value of the fitness function is used as the fitness evaluation index of the particle. By iteratively updating the position and velocity of the particles, they continuously move towards directions with higher fitness, eventually converging to a globally optimal or near-optimal layout scheme, which is then used as the recommended PV equipment layout scheme. Subsequently, the recommended photovoltaic equipment layout scheme will be applied to the actual management of photovoltaic equipment layout in highway engineering projects to ensure that the photovoltaic power generation system can operate efficiently and stably in accordance with the recommended photovoltaic equipment layout scheme, thereby achieving the goal of improving solar energy conversion efficiency and power generation.

[0104] Furthermore, embodiments of this application also include:

[0105] S761: The preset convergence conditions include a threshold for the proportion of the already laid-out area volume, a threshold for the proportion of the already laid-out tilt angle interval, and a threshold for the proportion of the already laid-out azimuth angle interval.

[0106] S762: When the proportion of the volume of the laid-out area of ​​the photovoltaic module layout information is greater than or equal to the threshold of the proportion of the volume of the laid-out area, and the proportion of the tilt angle interval of any layout position is greater than or equal to the threshold of the proportion of the tilt angle interval, and the proportion of the azimuth angle interval of any layout position is greater than or equal to the threshold of the proportion of the azimuth angle interval, it is considered that the preset convergence condition is met.

[0107] In a preferred embodiment, preset convergence conditions are defined, including a threshold for the proportion of the already deployed area volume, a threshold for the proportion of the already deployed tilt angle interval, and a threshold for the proportion of the already deployed azimuth angle interval. These three thresholds quantify and evaluate the convergence status of the photovoltaic module layout from three aspects: layout coverage, tilt angle optimization degree, and azimuth angle optimization degree, respectively. The proportion of the already deployed area volume refers to the ratio of the volume of the area where photovoltaic modules are currently randomly deployed to the volume of the entire layout constraint area. The proportions of the already deployed tilt angle interval and the proportion of the already deployed azimuth angle interval evaluate the layout from an angle perspective. Both tilt angle and azimuth angle affect the solar radiation reception efficiency of photovoltaic modules, and different combinations of angles may lead to significant differences in power generation.

[0108] Taking into account the above three convergence conditions, a standard is given for judging whether the photovoltaic module layout information meets the preset convergence conditions. Specifically, only when the proportion of the layout area volume in the photovoltaic module layout information is greater than or equal to the corresponding threshold, and the proportions of the tilt angle interval and azimuth angle interval of the layout both reach or exceed the corresponding thresholds, is the current photovoltaic module layout information considered to meet the preset convergence conditions, and thus the overall photovoltaic module layout planning can begin.

[0109] The method for optimizing the layout of photovoltaic equipment in highway engineering based on solar energy conversion efficiency provided in this invention has at least the following technical effects:

[0110] Based on the horizontal and vertical layout constraints of the photovoltaic (PV) equipment layout area in the highway engineering project, pre-defined PV modules are randomly laid out to obtain first and second PV module layout information. Different individual PV module layouts are generated within the constraint area, providing diverse candidate solutions for subsequent comparison and optimization. The first layout information (first layout position, first layout tilt angle, and first layout azimuth angle) and the second layout information (second layout position, second layout tilt angle, and second layout azimuth angle) of the first and second PV modules are obtained, identifying key factors affecting PV module irradiance and providing necessary input parameters for subsequent irradiance analysis. Differential irradiance type configurations are performed based on the first and second layout positions to obtain a differential irradiance prediction model. By comparing layout differences, areas requiring focused analysis are identified, and targeted irradiance prediction models are established to reduce unnecessary computation. Based on the differential irradiance prediction model, irradiance analysis is performed using the first layout position, first layout tilt angle, and first layout azimuth angle to obtain the first predicted irradiance. The irradiance level of the PV modules under the first layout scheme is evaluated, providing a basis for subsequent scheme selection. Based on the differential irradiance prediction model, irradiance analysis is performed using the second layout location, tilt angle, and azimuth angle to obtain the second predicted irradiance. This assesses the irradiance level of photovoltaic modules under the second layout method, providing a basis for subsequent scheme optimization. Based on the first and second predicted irradiances, the layout information of the first and second photovoltaic modules is ranked from largest to smallest, updating the layout information ranking results. Based on the irradiance level, the layout method of a single photovoltaic module with higher solar energy conversion efficiency is selected, providing a reference for the final layout optimization. When the photovoltaic module layout information meets the preset convergence conditions, a photovoltaic layout scheme is configured based on the layout information ranking results for highway engineering photovoltaic equipment layout management. Through iterative optimization, the layout scheme is continuously updated and screened, ultimately obtaining the optimal layout that meets the requirements, guiding the actual deployment of photovoltaic equipment, and improving optimization efficiency while ensuring optimization effect.

[0111] Example 2:

[0112] Please see Figure 2 , Figure 2 This is a schematic diagram illustrating an embodiment of the electronic device provided in this invention. For example... Figure 2 As shown, an electronic device 100 provided in this embodiment of the invention includes a memory 110, a processor 120, and a first computer program 111 stored in the memory 110 and executable on the processor 120. When the processor 120 executes the first computer program 111, it implements a method for optimizing the layout of photovoltaic equipment in highway engineering based on solar energy conversion efficiency.

[0113] Example 3:

[0114] Please see Figure 3 , Figure 3 This is a schematic diagram illustrating an embodiment of a computer-readable storage medium provided by an embodiment of the present invention. For example... Figure 3 As shown, this embodiment provides a computer-readable storage medium 200, on which a second computer program 211 is stored. When the second computer program 211 is executed by a processor, it implements a method for optimizing the layout of photovoltaic equipment in highway engineering based on solar energy conversion efficiency.

[0115] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0116] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0117] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0118] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0119] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0120] Although preferred embodiments of the invention have been described, those skilled in the art, once they have learned the basic inventive concept, can make other changes and modifications to these embodiments.

[0121] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for optimizing the layout of photovoltaic equipment in highway engineering based on solar energy conversion efficiency, characterized in that, include: Based on the horizontal and vertical layout constraints of the photovoltaic equipment layout area in highway engineering, photovoltaic modules of a preset model are randomly laid out to obtain the layout information of the first photovoltaic module and the layout information of the second photovoltaic module. Obtain the first layout position, first layout tilt angle, and first layout azimuth angle of the first photovoltaic module layout information, and obtain the second layout position, second layout tilt angle, and second layout azimuth angle of the second photovoltaic module layout information; Based on the first layout position and the second layout position, the differential irradiance type is configured to obtain the differential irradiance prediction model; Based on the differential irradiance prediction model, irradiance analysis is performed on the first layout position, the first layout tilt angle, and the first layout azimuth angle to obtain the first predicted irradiance. Based on the differential irradiance prediction model, irradiance analysis is performed on the second layout position, the second layout tilt angle, and the second layout azimuth angle to obtain the second predicted irradiance. Based on the first predicted irradiance and the second predicted irradiance, the layout information sorting results of the first photovoltaic module layout information and the second photovoltaic module layout information are updated from largest to smallest. When the photovoltaic module layout information meets the preset convergence conditions, a photovoltaic layout scheme is configured based on the sorting result of the layout information to manage the layout of photovoltaic equipment in highway engineering. Specifically, a differential irradiance prediction model is obtained by configuring the differential irradiance type based on the first and second layout positions, including: Obtain the first direct irradiance influence factor, the first diffuse irradiance influence factor, and the first reflected irradiance influence factor at the first layout location; Obtain the second direct irradiance influencing factors, the second diffuse irradiance influencing factors, and the second reflected irradiance influencing factors for the second layout location; By comparing the first direct irradiation influencing factor and the second direct irradiation influencing factor, the deviation of the direct irradiation factor is obtained; By comparing the first and second factors affecting scattered radiation, the deviation of the scattered radiation factors is obtained. By comparing the first and second factors affecting reflected irradiance, the deviation of reflected irradiance factors is obtained. When the deviation of the direct irradiance factor is greater than or equal to the direct irradiance factor deviation threshold, the direct irradiance is added to the differential irradiance type; when the deviation of the scattered irradiance factor is greater than or equal to the scattered irradiance factor deviation threshold, the scattered irradiance is added to the differential irradiance type; when the deviation of the reflected irradiance factor is greater than or equal to the reflected irradiance factor deviation threshold, the reflected irradiance is added to the differential irradiance type. Based on the differential irradiance type, retrieve the differential irradiance prediction model; The process for obtaining the first influencing factor of reflected irradiation includes: The first layout location includes the first grounding location of the first photovoltaic equipment support, and obtains a first ground feature with the first grounding location as the center and a preset length as the radius; The first ground feature is set as the first reflected radiation influencing factor.

2. The method as described in claim 1, characterized in that, The differential irradiance prediction model includes a direct irradiance prediction model, a diffuse irradiance prediction model, and a reflected irradiance prediction model. The training methods for the direct irradiance prediction model, the diffuse irradiance prediction model, and the reflected irradiance prediction model are the same. Specifically, the training steps for the direct irradiance prediction model include: Based on the rated range of irradiation influencing factors, record data of direct irradiation influencing factors are randomly configured. Using the data recorded on the influencing factors of direct irradiation, irradiation samples were collected for photovoltaic module models to obtain a set of direct irradiance recorded values. A central tendency assessment is performed on the set of direct irradiance records to obtain direct irradiance identification data; Add the recorded data of direct irradiation influencing factors to the set of recorded data of direct irradiation influencing factors, and add the direct irradiance label data to the set of direct irradiance label data; The direct irradiance prediction model is trained based on the set of direct irradiance identifiers and the set of direct irradiance influencing factors records.

3. The method as described in claim 2, characterized in that, Based on the differential irradiance prediction model, irradiance analysis is performed on the first layout position, the first layout tilt angle, and the first layout azimuth angle to obtain the first predicted irradiance, including: Based on the differential irradiance prediction model, a first differential irradiance prediction model is extracted, wherein the first differential irradiance prediction model has a first differential irradiance type label, and the first differential irradiance type label has a first weight. Based on the first differential irradiance type label, the first irradiance influencing factor characteristic data of the first layout location is collected, and the first irradiance influencing factor characteristic data is processed by the first differential irradiance prediction model to generate the first type of predicted irradiance. Until the Nth differential irradiance prediction model is extracted based on the differential irradiance prediction model, wherein the Nth differential irradiance prediction model has an Nth differential irradiance type label and the Nth differential irradiance type label has an Nth weight; Based on the Nth differential irradiance type label, collect the Nth irradiance influencing factor characteristic data of the first layout location, process the Nth irradiance influencing factor characteristic data through the Nth differential irradiance prediction model, and generate the Nth type predicted irradiance, where 3≥N≥0 and N is an integer; Based on the first weight up to the Nth weight, the normalized values ​​of the first type of predicted irradiance up to the Nth type of predicted irradiance are weighted to obtain an N-ary irradiance feature array. The N-ary irradiance feature array is processed using an irradiance fusion analysis model to generate the first predicted irradiance.

4. The method as described in claim 3, characterized in that, The steps for constructing the irradiance fusion analysis model include: Based on the photovoltaic module model, configure an M-element irradiance array to record the dataset and the irradiance detection true value dataset; Using the irradiance detection true value dataset as supervision and the M-gram irradiance array record dataset as input, a first-level irradiance fusion analysis model is trained. Extract the set of output deviation vectors from the first-level irradiance fusion analysis model; Perform systematic error vector analysis on the set of output deviation vectors to obtain the first-level systematic error vector; When the magnitude of the first-level system error vector is greater than or equal to the magnitude threshold, the negative number of the first-level system error vector is set as the second-level irradiance label dataset. Using the secondary irradiance identifier dataset as supervision and the M-gram irradiance array record dataset as input, a secondary irradiance fusion analysis model is trained. Until the magnitude of the Q-level system error vector is less than the magnitude threshold, the output summation fully connected is performed on the first-level irradiance fusion analysis model, the second-level irradiance fusion analysis model, and so on up to the Q-level irradiance fusion analysis model to obtain the irradiance fusion analysis model.

5. The method as described in claim 1, characterized in that, When the photovoltaic module layout information meets the preset convergence conditions, a photovoltaic layout scheme is configured based on the sorting results of the layout information to manage the layout of photovoltaic equipment in highway engineering, including: Through the user terminal, the layout distance constraint variance, layout boundary constraint distance and layout constraint quantity are configured for the layout constraint area of ​​the photovoltaic equipment in the highway project. Construct a fitness function, wherein the fitness function is the sum of the sequence numbers of the layout information sorting results; Using the layout distance constraint variance, the layout boundary constraint distance, and the number of layout constraints as constraints, a set of photovoltaic equipment layout schemes is configured by performing several random combinations based on the layout information sorting results. Based on the fitness function, particle swarm optimization is performed on the set of photovoltaic equipment layout schemes to obtain recommended photovoltaic equipment layout schemes; The photovoltaic equipment layout management for highway engineering shall be carried out in accordance with the recommended photovoltaic equipment layout scheme.

6. The method as described in claim 5, characterized in that, When the photovoltaic module layout information meets the preset convergence conditions, including: The preset convergence conditions include a threshold for the proportion of the already laid area volume, a threshold for the proportion of the already laid tilt angle interval, and a threshold for the proportion of the already laid azimuth angle interval. When the proportion of the volume of the laid-out area of ​​the photovoltaic module layout information is greater than or equal to the threshold of the proportion of the volume of the laid-out area, and the interval of the laid-out tilt angle of any layout position is greater than or equal to the threshold of the proportion of the laid-out tilt angle interval, and the interval of the laid-out azimuth angle of any layout position is greater than or equal to the threshold of the proportion of the laid-out azimuth angle interval, it is considered that the preset convergence condition is met.

Images