Multi-algorithm fusion-based complex terrain photovoltaic module arrangement optimization system and method

Abstract

The invention relates to the technical field of electric digital data processing, and particularly discloses a complex terrain photovoltaic module arrangement optimization system and method based on multi-algorithm fusion. The system is provided with an arrangement optimization judgment module, a time sequence-terrain shielding kernel digestion module, a time sequence-space shielding kernel analysis module, a hourly energy effective state judgment module and a photovoltaic module arrangement optimization module, so that a time sequence-space shielding kernel is obtained; the hourly energy effective state judgment module is coupled with the hourly meteorological parameters and the time sequence-space shielding kernel to obtain an effective incident irradiation evaluation value, and the hourly energy effective state is judged based on a particle swarm algorithm; and finally, a photovoltaic module arrangement optimization module adjusts an arrangement framework according to the configuration, and optimization is completed. Wherein real-time judgment ensures optimization timeliness, shielding digestion improves irradiation calculation precision, mutual shielding kernel analysis reduces shielding influence between assemblies, energy state judgment provides accurate basis for optimization, and final optimization remarkably improves power generation efficiency of a photovoltaic system.

Description

A Multi-Algorithm Fusion-Based Optimization System and Method for Photovoltaic Module Layout in Complex Terrain Technical Field

[0001] This invention relates to the field of electronic digital data processing technology, specifically to a system and method for optimizing the layout of photovoltaic modules in complex terrain based on multi-algorithm fusion. Background Technology

[0002] Existing optimization methods for photovoltaic (PV) deployments in complex terrain typically begin with high-precision DEM / point cloud modeling, overlaying prohibitions and micro-topographic zoning, performing solar path and temporal shading analysis, and then using multi-objective optimization to seek optimal results across tilt angle, azimuth, row spacing / GCR, table size, ground clearance, subarray boundaries, and electrical grouping, while also constraining earthwork, road / voltage drop / grid connection, and structural safety, using AEP / LCOE as deliverable indicators. However, the slope, aspect, curvature, and elevation differences of complex terrain can raise the horizon angle and exacerbate twilight and over-the-horizon effects. Winter shading induces mismatches between cables and between cables, and geological bearing capacity, drainage and accessibility limit the orientation of supports and roads. Wind and snow / ice zoning changes the ground clearance and structural form, thus forcing the deployment to change from "regular and neat" to a combination strategy of "following the shape of the zone, step-leveling, staggered / segmented orientation, variable tilt angle and dynamic row spacing, local single-axis back tracking, equivalent grouping of components and coordination with cables / roads". Through repeated verification by time-series energy simulation, the shading loss is controllable and the construction cost and power generation benefits are optimized globally.

[0003] For example, Chinese invention patent CN115994291A discloses a method for arranging photovoltaic power station modules adapted to mountainous terrain, including the following steps: A. Measured topographic map; B. Dividing the slope into different regions according to the slope and orientation, that is, grouping slopes with similar slopes and orientations into a region set; C. Calculating the east-west and north-south slope components of the region set, specifically the east-west and north-south slope components are calculated from the slope inclination and orientation; D. Orientation selection; E. Slope selection; F. Array length selection; G. Array longitudinal direction selection. Finally, a specialized simulation can be performed using software.

[0004] For example, Chinese invention patent CN115841010A discloses an application of a Revit-based terrain shadow area analysis algorithm in photovoltaic design. Based on Revit software, it sets the time range for shadow calculation, determines the solar altitude angle and azimuth angle data within the set time period according to the latitude and longitude of the input terrain location, and calculates the shadow area based on the solar altitude angle and azimuth angle data. All triangular faces of the terrain emit rays along the direction of sunlight according to the solar altitude angle and azimuth angle values ​​of the set time. If the rays emitted by the triangular face collide with the terrain model, it means that the sunlight of the triangular face will be blocked by the terrain at that time, that is, the triangular face is a shadow area. The area is marked as a shadow area area. All calculated shadow areas within the set time period are merged and displayed to generate the shadow area area for that time period.

[0005] Based on the above technical solutions, it was found that most existing photovoltaic module placement methods are based on geometric shading determination during representative time periods or derive row spacing using simplified rules. Because the design process and tools for photovoltaic module placement are biased towards static visualization, use representative moments to replace hourly energy throughout the year and do not incorporate the temporal coupling of tracking / anti-tracking, and lack hourly incident energy weighting, it will lead to mismatch in row spacing configuration, increased shading during uncovered periods, and a larger deviation in the annual power generation assessment. At the same time, the terrain self-shading and inter-row (inter-array) mutual shading are treated separately and not coupled and optimized in the same spatiotemporal domain. This results in the terrain ray and array geometric rules being divided and conquered, and not linked with the electrical grouping control strategy in a unified spatiotemporal domain, ultimately affecting the effective placement of photovoltaic modules. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a system and method for optimizing the layout of photovoltaic modules in complex terrain based on multi-algorithm fusion, which can effectively solve the problems mentioned in the background technology.

[0007] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of the present invention provides a complex terrain photovoltaic module layout optimization system based on multi-algorithm fusion, comprising: a layout optimization determination module, used to obtain the real-time output power generation of the photovoltaic modules, determine the effective state of the photovoltaic module layout based on a genetic algorithm, and if the state is determined to be in need of optimization, then proceeding to the time-series-terrain shading kernel resolution module; the time-series-terrain shading kernel resolution module, used to obtain the hourly solar altitude angle and the hourly solar azimuth angle of each grid point under the photovoltaic module layout architecture, perform terrain and / or obstacle shading resolution, and obtain the time-series-terrain shading kernel of each grid point; and a time-series-spatial shading kernel parsing module, used to obtain the photovoltaic module... The system employs a multi-dimensional (MDE) array to determine the hourly arrangement parameters of each grid point within the array architecture. It analyzes the projection of adjacent rows of photovoltaic (PV) modules to obtain the temporal mutual shading kernel of each grid point. Combined with the temporal-terrain shading kernel of each grid point, a minimum-coupling algorithm is used to obtain the temporal-spatial shading kernel of each grid point. An hourly energy effective state determination module acquires the hourly meteorological parameters of the PV module layout area and couples them with the temporal-spatial shading kernel of each grid point to obtain the effective incident irradiance assessment value of each grid point. A particle swarm optimization algorithm is used to determine the hourly energy effective state of the PV modules. A PV module layout optimization module adjusts the PV module layout architecture based on the hourly energy effective state of the PV modules, updates the PV module layout adaptation architecture, and completes the PV module layout optimization.

[0008] The second aspect of this invention provides a method for optimizing the layout of photovoltaic modules in complex terrain based on multi-algorithm fusion, comprising: S1. obtaining the real-time output power of the photovoltaic modules, determining the effective state of the photovoltaic module layout based on a genetic algorithm, and if the state is determined to be in need of optimization, proceeding to S2; S2. obtaining the hourly solar altitude angle and the hourly solar azimuth angle of each grid point under the photovoltaic module layout architecture, performing terrain and / or obstacle shading resolution, and obtaining the time-series-terrain shading kernel of each grid point; S3. obtaining the hourly arrangement parameters of each grid point under the photovoltaic module layout architecture, and analyzing the photovoltaic modules. S4. Project adjacent rows to obtain the temporal mutual shading kernel of each grid point, and combine it with the temporal-terrain shading kernel of each grid point to obtain the temporal-spatial shading kernel of each grid point based on the minimum value coupling algorithm; S5. Obtain the hourly meteorological parameters of the photovoltaic module deployment area, and couple them with the temporal-spatial shading kernel of each grid point to obtain the effective incident irradiance evaluation value of each grid point, and determine the hourly energy effective state of the photovoltaic module based on the particle swarm optimization algorithm; S6. Adjust the photovoltaic module deployment architecture according to the hourly energy effective state of the photovoltaic module, update the photovoltaic module deployment adaptation architecture, and complete the photovoltaic module deployment optimization.

[0009] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: (1) The present invention provides a complex terrain photovoltaic module layout optimization system and method based on multi-algorithm fusion. The layout optimization judgment module first outputs the power generation of the photovoltaic module in real time and uses a genetic algorithm to determine its effective layout state. When optimization is required, it enters the time-terrain shading kernel resolution module. This module obtains the hourly solar altitude angle and azimuth angle of each grid point, resolves terrain and obstacle shading, and obtains the time-terrain shading kernel. The time-space shading kernel analysis module combines the hourly arrangement parameters of each grid point to analyze the adjacent row projection, obtains the matrix time-series mutual shading kernel, and then obtains the time-space shading kernel by the minimum value coupling algorithm with the time-terrain shading kernel. The hourly energy effective state judgment module couples the hourly meteorological parameters with the time-space shading kernel to obtain the effective incident radiation evaluation value, and determines the hourly energy effective state based on the particle swarm algorithm. Finally, the photovoltaic module layout optimization module adjusts the layout architecture accordingly to complete the optimization. Among these, real-time judgment ensures timely optimization, shading mitigation improves the accuracy of irradiance calculation, mutual shading analysis reduces the impact of shading between components, and energy state judgment provides accurate basis for optimization. Ultimately, optimization significantly improves the power generation efficiency of photovoltaic systems.

[0010] (2) This approach overcomes the limitations of averaging and idealization in traditional irradiance assessments by evaluating the effective incident irradiance values ​​of each grid point. Through precise coupling of hourly meteorological parameters and temporal-spatial shading kernels, it considers both the natural fluctuations of solar radiation (such as cloud changes and atmospheric attenuation) and complex shading factors such as terrain undulations, surrounding obstacles, and mutual shading between modules. The resulting irradiance data accurately reflects the actual available light energy at each grid point. This process provides a precise micro-scale basis for determining the energy state of photovoltaic modules, avoiding optimization biases caused by overgeneralization in irradiance assessments. It ensures that subsequent layout adjustments can specifically improve the light capture efficiency in low-irradiance areas while maximizing the power generation potential in high-irradiance areas, fundamentally improving the balance and maximization of energy conversion in the entire photovoltaic system.

[0011] (3) The full utilization of multiple parameters in the process of this invention has constructed a multi-dimensional evaluation system covering real-time status, environmental characteristics, and spatial relationships: the real-time output power generation directly reflects the current energy efficiency level of the components, providing a quantitative benchmark for the determination of optimization needs; the hourly solar altitude angle and azimuth angle are the core driving parameters for analyzing terrain and obstacle shading, determining the temporal dynamic characteristics of the shading core; meteorological parameters (such as total irradiance, cloud cover, aerosols, etc.) are the basis for calculating the effective incident irradiance assessment value, directly affecting the theoretical upper limit of solar energy resources; the component layout parameters are related to the spatial distribution of mutual shading of the array, determining the magnitude of shading loss within the system. The synergistic effect of these parameters not only provides solid data support for the analysis of each link, but also achieves full-dimensional coverage from the macro environment to the micro components, from static terrain to dynamic temporal sequence, enabling the layout optimization to adapt to complex natural environments and match the real-time status of component operation, greatly improving the scientificity and adaptability of the scheme.

[0012] (4) Compared with the prior art, the significant advantages of the process of the present invention are reflected in three dimensions: First, it breaks through the bottleneck of static planning and dynamic adjustment in traditional photovoltaic layout optimization. Through the closed-loop design of the layout optimization judgment module and the photovoltaic module layout optimization module, combined with the intelligent iteration of genetic algorithm and particle swarm algorithm, it realizes the transformation from passive adaptation to active optimization and can respond to environmental changes and module state fluctuations in real time. Second, in terms of shading treatment, it innovatively incorporates terrain shading, obstacle shading and module mutual shading into a unified temporal-spatial shading kernel framework. Through the minimum value coupling algorithm, it realizes the superposition analysis of multiple types of shading, which solves the problem of fragmented processing and evaluation distortion of shading factors in the prior art. Third, the whole process takes the hourly energy effective state as the core link, deeply integrates meteorological data, spatial parameters and algorithm models, and forms a full-chain technical system from state perception, shading analysis to optimization execution. Compared with the traditional method that relies solely on empirical formulas or local parameters, it has made significant breakthroughs in improving power generation efficiency (especially in complex terrain or high-density layout scenarios) and adaptability to environmental changes. It has a wider range of applications and more stable optimization effects. Attached Figure Description

[0013] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0014] Figure 1 is a schematic diagram of the system module connection of the present invention.

[0015] Figure 2 is a schematic diagram of the method steps of the present invention.

[0016] Figure 3 is a simplified schematic diagram of the projected side view section.

[0017] Figure 4 is a flowchart of the photovoltaic module layout optimization scheme.

[0018] Figure 5 is a schematic diagram of component layout coordinate updates.

[0019] Figure label: The azimuth of the obstructing object, For projection angle, Let L be the tilt angle of the photovoltaic module plane where each grid point is located, and L be the length of the shaded area. Detailed Implementation

[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0021] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0022] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0023] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0024] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0025] Referring to Figure 1, the first aspect of the present invention provides a photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion, comprising: a layout optimization determination module, a time-series-terrain shading kernel resolution module, a time-series-spatial shading kernel parsing module, an hourly energy effective state determination module, a photovoltaic module layout optimization module, and a module layout database. The module layout database is used to store preset values ​​for various parameters.

[0026] The aforementioned components are arranged in a database containing the corresponding relationships, including but not limited to preset relationships, matching relationships, and mapping relationships. Taking the mapping table of the relationship between the solar azimuth angle and the horizon angle threshold as an example, the specific acquisition process first collects a large amount of sample data of the target parameter (such as the solar azimuth angle) and the associated parameter (such as the horizon angle threshold) through multi-scenario field measurements. The initial mapping relationship is then derived by combining theoretical models (such as celestial kinematics and the geometric relationship of terrain occlusion). Then, based on the verification data of typical scenarios, the initial relationship is subjected to error analysis and iterative optimization to correct the deviation under extreme working conditions. Finally, through dynamic calibration of long-term operating data, a stable mapping table covering all working conditions is formed to ensure that the corresponding relationships between different parameters have high accuracy and applicability in various actual scenarios.

[0027] The layout optimization judgment module is connected to the time-series-terrain shading kernel resolution module, the time-series-terrain shading kernel resolution module is connected to the time-series-spatial shading kernel parsing module, the time-series-spatial shading kernel parsing module is connected to the hourly energy effective state judgment module, the hourly energy effective state judgment module is connected to the photovoltaic module layout optimization module, and the layout optimization judgment module, the time-series-terrain shading kernel resolution module, the hourly energy effective state judgment module, and the photovoltaic module layout optimization module are all connected to the module layout database.

[0028] In this embodiment, the specific layout optimization process is based on Figure 4, which is a flowchart of the photovoltaic module layout optimization scheme. The process starts from the real-time power and uses a genetic algorithm to determine whether the current layout is effective. If optimization is required, the site is first processed to eliminate terrain and obstacle shading to obtain a temporal terrain shading kernel. Then, the adjacent row projections are analyzed to obtain the array temporal mutual shading kernel and coupled with the minimum value of the terrain kernel to form a spatiotemporal shading kernel. Subsequently, hourly meteorological parameters are combined to calculate the effective incident irradiance evaluation value of each grid point, and the hourly energy effective state is determined using a particle swarm optimization algorithm. Finally, the layout architecture is adjusted and updated according to the determination results and returned to the starting point to form a closed-loop optimization driven by real-time data.

[0029] The layout optimization judgment module is used to obtain the real-time output power of photovoltaic modules and determine the effective state of photovoltaic module layout based on genetic algorithm. If it is determined to be a state that needs optimization, it enters the time-terrain shading kernel resolution module.

[0030] Specifically, the effective state of the photovoltaic module arrangement is determined by the following process: the photovoltaic module monitoring unit is equipped with a genetic algorithm to obtain the expected output power based on the cumulative historical output power of the photovoltaic modules.

[0031] The real-time output power of the photovoltaic module, which can be extracted from the output record of the photovoltaic module, is compared with the expected output power. If the real-time output power of the photovoltaic module is greater than or equal to the expected output power, the photovoltaic module monitoring unit determines that the photovoltaic module arrangement status is valid.

[0032] If the real-time output power of the photovoltaic module is less than the expected output power, then the continuous window in which the real-time output power is less than the expected output power is counted and recorded as the output weakening window of the photovoltaic module. This window is compared with the predefined output weakening boundary window in the module layout database to obtain the weakening window comparison result. The weakening window comparison result includes a first weakening window comparison result and a second weakening window comparison result.

[0033] The first result of the weakening window comparison indicates that the output weakening window of the photovoltaic module is less than or equal to the output weakening boundary window, and the second result of the weakening window comparison indicates that the output weakening window of the photovoltaic module is greater than the output weakening boundary window.

[0034] If the weakened window comparison result shows the first result of the weakened window comparison, the photovoltaic module arrangement status is determined to be in an effective state. If the weakened window comparison result shows the second result of the weakened window comparison, the photovoltaic module arrangement status is determined to be in an optimization state, and photovoltaic module arrangement optimization is performed.

[0035] In this embodiment, by collecting the output power of photovoltaic modules in real time and determining whether the arrangement is in an effective state, the operating data is directly pulled into the design-operation closed loop: once shading, mismatch, dust accumulation or seasonal changes in sunlight are detected that the power generation is too low, the arrangement optimization is triggered (such as fine-tuning of row spacing / tilt angle / azimuth and back tracking threshold, equivalent group rearrangement of series-MPPT (maximum power point tracking), correction of local steps and ground clearance), thereby timely curbing long-term performance loss, reducing shading and mismatch losses and line losses, reducing hot spots and failure probability, increasing annual power generation and kWh / kWp, reducing LCOE (levelized cost of electricity), and making operation and maintenance more targeted and grid connection more stable.

[0036] The temporal-terrain shading kernel resolution module is used to obtain the hourly solar altitude angle and the hourly solar azimuth angle of each grid point under the photovoltaic module layout architecture, and to perform terrain and / or obstacle shading resolution to obtain the temporal-terrain shading kernel of each grid point.

[0037] Furthermore, the temporal-terrain occlusion kernel of each grid point is obtained. The specific analysis process is as follows: based on the hourly solar azimuth angle of each grid point, which can be extracted from the monitoring records of the photovoltaic module monitoring sensor group, and based on the relationship mapping table between the solar azimuth angle and the horizon angle threshold, which is extracted from the module layout database, the hourly solar azimuth angle of each grid point is substituted into the relationship mapping table to obtain the horizon angle threshold of each grid point, which is denoted as the horizon angle adaptation threshold of each grid point.

[0038] The hourly solar altitude angle of each grid point in the photovoltaic module layout architecture is extracted from the monitoring records of the photovoltaic module monitoring sensor group. It is compared with the horizon angle adaptation threshold of each grid point. The horizon angle adaptation threshold represents the maximum elevation angle of all terrain / obstacles in that azimuth. If the hourly solar altitude angle of a grid point is greater than the horizon angle adaptation threshold of all grid points, the hourly shading kernel of that grid point is recorded as 1, indicating that there is no obstruction in the direct sunlight path from the sun to that grid point. If the hourly solar altitude angle of a grid point is less than or equal to the horizon angle adaptation threshold of all grid points, the hourly shading kernel of that grid point is recorded as 0, indicating that the surrounding terrain / obstacles (ridges, canopies, buildings, etc.) in a certain direction are blocking the sunlight in that solar azimuth, causing the grid point to be shaded.

[0039] The hourly shading kernel of each grid point is statistically analyzed and denoted as the time-series-terrain shading kernel of each grid point. Specifically, it refers to the coefficient at a certain time when a spatial grid location in the photovoltaic module deployment area is blocked from direct sunlight by the surrounding terrain / obstacles, and is a dimensionless number with a value of 0-1.

[0040] In this embodiment, the horizon angle threshold method can significantly reduce computational load and instability, stably capture key shading windows during dawn / dusk and winter, and seamlessly couple with the following matrix mutual shading kernel on the same grid / at the same time, facilitating inclusion in the inner loop iteration of multi-objective optimization such as GA / PSO. The result is to reduce the installed capacity density loss caused by conservative row spacing, suppress shading and mismatch misjudgment, increase annual power generation and power generation per unit installed peak power (kWh / kWp), reduce LCOE, and provide accurate and reusable shading base plates for subsequent construction and operation and maintenance (back tracking parameters, cleaning / pruning strategies).

[0041] The temporal-spatial occlusion kernel parsing module is used to obtain the hourly arrangement parameters of each grid point under the photovoltaic module layout architecture, parse the adjacent row projection of the photovoltaic module, obtain the matrix temporal mutual occlusion kernel of each grid point, and combine the temporal-terrain occlusion kernel of each grid point with the minimum value coupling algorithm to obtain the temporal-spatial occlusion kernel of each grid point.

[0042] Specifically, the temporal-spatial shading kernel for each grid point is obtained. The specific analysis process involves the hourly arrangement parameters of each grid point under the photovoltaic module layout architecture, including the tilt angle of the photovoltaic module plane where each grid point is located, the azimuth angle of the photovoltaic module plane where each grid point is located, the row spacing between two adjacent rows of photovoltaic module arrays, and the equivalent shading height between two adjacent rows of photovoltaic module arrays. The arrangement parameters can be extracted from the photovoltaic module layout records.

[0043] A side view profile of the photovoltaic module array is formed based on the tilt angle of the photovoltaic module plane where each grid point is located, the azimuth angle of the photovoltaic module plane where each grid point is located, the row spacing between two adjacent rows of photovoltaic module arrays, and the equivalent shading height of two adjacent rows of photovoltaic module arrays. In the profile perpendicular to the row direction of the photovoltaic module array, the module in the upstream row of the photovoltaic module array is recorded as the shading body, the upper edge segment (top edge) of the shading body is used as the projection generatrix, and the current hourly solar azimuth angle is used as the projection angle. The azimuth angle corresponding to the center grid point of the shading body is recorded as the azimuth angle of the shading body. The shading body is projected onto the downstream row of module plane, and the shadow area of ​​the shading body on the downstream row of module plane is projected.

[0044] The shadow area is rasterized onto each grid point of the downstream row component. If a grid point falls on the shadow area, the matrix temporal mutual occlusion kernel of that grid point is recorded as 0, otherwise it is recorded as 1, thus obtaining the matrix temporal mutual occlusion kernel of each grid point.

[0045] The projection process described above is shown in Figure 3. Figure 3 is a simplified schematic diagram of the projection side view, showing the relevant angular relationships of the photovoltaic modules. It is expressed as the azimuth of the obstructing object. Represented as the projection angle, The tilt angle of the photovoltaic module plane where each grid point is located is represented. The corresponding tilt angle of the photovoltaic module plane may vary in actual layout due to different terrain, layout environment, etc. L represents the length of the shaded area. There is a certain spatial relationship between the modules. Through projection analysis, we can better understand the projection of light on different modules and the possibility of mutual shading between modules, which helps to optimize the design and layout of photovoltaic modules and improve the overall power generation efficiency.

[0046] The temporal mutual occlusion kernel of each grid point and the temporal-terrain occlusion kernel of each grid point are input into the minimum value coupling algorithm. The minimum value coupling algorithm outputs the temporal-spatial occlusion kernel of each grid point.

[0047] The minimum value coupling algorithm described above is as follows:

[0048] In the formula, Let r be the spatial occlusion kernel at time t for the r-th grid point, where r is the grid point number. R represents the total number of grid points in a single photovoltaic module, and t represents the hourly time number. T is the total amount at each time step, and K is the total amount at each time step. tr (r, t) represents the terrain occlusion kernel at time t for the r-th grid point, tr represents the terrain occlusion kernel, and K represents the terrain occlusion kernel. ar (r, t) is the mutual occlusion kernel of the square matrix of the r-th grid point at time t, ar is the mutual occlusion kernel, min( () is to take the minimum value.

[0049] The hourly energy effective state determination module is used to obtain hourly meteorological parameters of the photovoltaic module deployment area. It is coupled with the time-space shading kernel of each grid point to obtain the effective incident irradiance evaluation value of each grid point, and the hourly energy effective state of the photovoltaic module is determined based on the particle swarm algorithm.

[0050] Furthermore, the hourly energy availability status of the photovoltaic modules is determined. The specific process involves generating irradiance curves for each grid point based on the effective incident irradiance assessment values. The similarity between the irradiance curves of each grid point and the irradiance curves of the remaining grid points is calculated, and the standard deviation is processed to obtain the similarity standard deviation of the photovoltaic modules. Specifically, the irradiance curve is constructed with the horizontal axis representing time points in hours and the vertical axis representing the effective incident irradiance assessment values ​​in watts per square meter.

[0051] Based on the similarity standard deviation of photovoltaic modules, the similarity clustering threshold factor corresponding to each predefined similarity standard deviation interval in the module layout database is matched to determine the specific interval of the similarity standard deviation of photovoltaic modules, obtain the similarity clustering threshold factor corresponding to the interval, and couple it with the similarity clustering reference value. Specifically, the similarity clustering threshold factor is multiplied by the similarity clustering reference value to obtain the similarity clustering adaptation value, and the clustering process of each grid point is configured.

[0052] Clustering the effective incident irradiance assessment values ​​of each grid point belonging to the same photovoltaic module yields the effective incident irradiance assessment value of the photovoltaic module.

[0053] The effective incident irradiance assessment value of the photovoltaic module is compared with each predefined effective incident irradiance assessment value interval in the module layout database. The effective incident irradiance assessment value is screened based on the particle swarm optimization algorithm to determine the interval to which the effective incident irradiance assessment value of the photovoltaic module belongs. Each effective incident irradiance assessment value interval includes a first effective incident irradiance assessment value interval, a second effective incident irradiance assessment value interval, a third effective incident irradiance assessment value interval, and a fourth effective incident irradiance assessment value interval.

[0054] The effective incident irradiance assessment value interval is defined as a monotonically decreasing threshold: Assume that y1 belongs to the first interval; y2 belongs to the second interval; y3 belongs to the third interval; and y4 belongs to the fourth interval. For any y1, y2, y3, y4, y1>y2>y3>y4, the contribution of each interval to annual power generation and the rigidity of shading control decrease synchronously: The first interval must strictly ensure no mutual shading and prioritize back tracking / heat dissipation to avoid roof clipping; the second interval, under the premise of meeting MPPT and voltage drop, can moderately increase GCR (ground coverage / layout density) and use equivalent grouping to suppress mismatch; the third interval can relax the no-shading window, increase installed capacity with denser layout, and pay attention to start-up voltage and multi-peak MPPT; the fourth interval mainly gives way to construction accessibility such as roads / drainage, and has the least impact on layout decisions.

[0055] If the effective incident irradiance assessment value of the photovoltaic module falls within the first effective incident irradiance assessment value range, then the hourly energy of the photovoltaic module is determined to be in an effective state; otherwise, the hourly energy of the photovoltaic module is determined to be in a sub-healthy state.

[0056] Specifically, the effective incident irradiance assessment value for each grid point is obtained. The specific coupling process is as follows: the hourly meteorological parameters of the photovoltaic module deployment area include the hourly solar altitude angle, hourly water vapor content, and hourly cloud cover of each grid point. These meteorological parameters can be extracted from the photovoltaic module monitoring records. Specifically, the hourly solar altitude angle, hourly water vapor content, and hourly cloud cover of each grid point are obtained by gridding these parameters into the corresponding grid points.

[0057] The hourly water vapor content and hourly cloud cover of each grid point are coupled together to obtain the hourly atmospheric transmittance influence coefficient of each grid point.

[0058] Specifically, the above coupling involves multiplying the hourly water vapor content of each grid point by the weight element corresponding to the predefined water vapor content in the component layout database to obtain the first component of the hourly atmospheric transmittance influence coefficient of each grid point; multiplying the hourly cloud cover of each grid point by the weight element corresponding to the predefined cloud cover in the component layout database to obtain the second component of the hourly atmospheric transmittance influence coefficient of each grid point; and adding the first component influence coefficient and the second component influence coefficient of the hourly atmospheric transmittance of each grid point to obtain the hourly atmospheric transmittance influence coefficient of each grid point.

[0059] The hourly solar elevation angle and the hourly atmospheric transmittance influence coefficient of each grid point are normalized together with the temporal-spatial shading kernel of each grid point. The normalized results are then coupled together to obtain the effective incident irradiance assessment value of each grid point. The specific analysis process is as follows:

[0060] In the formula, Ge(r,t) is the effective incident irradiance assessment value at time t for the r-th grid point, and K(r,t) is the spatial shading kernel at time t for the r-th grid point. (r, t) is the solar elevation angle at time t of the r-th grid point, T(r, t) is the atmospheric transmittance influence coefficient at time t of the r-th grid point, h2 is the weight parameter corresponding to the predefined spatial shading kernel in the component layout database, h1 is the weight parameter corresponding to the predefined solar elevation angle in the component layout database, and h3 is the weight parameter corresponding to the predefined atmospheric transmittance influence coefficient in the component layout database.

[0061] Furthermore, the effective incident irradiance evaluation values ​​corresponding to each grid point belonging to the same photovoltaic module are clustered to obtain the effective incident irradiance evaluation value of the photovoltaic module. This also includes a clustering process based on the clustering adaptation value of similarity clustering to configure an unsupervised clustering algorithm, and semi-clustering to obtain each cluster of effective incident irradiance evaluation values ​​of the photovoltaic module.

[0062] The standard deviation and mean of each effective incident irradiance assessment value cluster of the photovoltaic module are calculated separately. The standard deviation and mean of the effective incident irradiance assessment value cluster of the photovoltaic module are coupled. Specifically, the standard deviation and mean of the effective incident irradiance assessment value cluster of the photovoltaic module are processed by ratio processing to obtain the characteristic variation coefficient of each effective incident irradiance assessment value cluster of the photovoltaic module. The mean is then processed to obtain the intra-group irradiance characteristic variation coefficient of the photovoltaic module.

[0063] The coefficient of variation of the irradiance characteristics within a group of photovoltaic modules is compared with the predefined coefficient of variation of irradiance characteristics in the module layout database. If the coefficient of variation of the irradiance characteristics within a group of photovoltaic modules is less than or equal to the coefficient of variation of irradiance characteristics, then the effective incident irradiance assessment value clusters of photovoltaic modules are fully clustered to obtain the effective incident irradiance assessment value of the photovoltaic modules. If the coefficient of variation of the irradiance characteristics within a group of photovoltaic modules is greater than the coefficient of variation of irradiance characteristics, then a similarity threshold correction factor is matched based on the difference between the coefficient of variation of the irradiance characteristics within a group of photovoltaic modules and the coefficient of variation of irradiance characteristics. This factor is coupled with the similarity clustering adaptation value to obtain the similarity clustering correction value. The similarity clustering adaptation value is then updated to reconfigure the clustering process.

[0064] The above matching yields a similarity threshold correction factor. Specifically, the difference between the coefficient of variation of the irradiance characteristics within a group of photovoltaic modules and the coefficient of variation of the irradiance characteristics is recorded as the deviation of the coefficient of variation of the irradiance characteristics within a group of photovoltaic modules. This deviation is then matched with the similarity threshold correction factor corresponding to the predefined deviation interval of the coefficient of variation of the irradiance characteristics within each group in the module layout database. This determines the specific interval of the deviation of the coefficient of variation of the irradiance characteristics within a group of photovoltaic modules, and the similarity threshold correction factor corresponding to that interval is obtained.

[0065] The photovoltaic module layout optimization module is used to adjust the photovoltaic module layout architecture according to the hourly energy availability status of the photovoltaic modules, update the photovoltaic module layout adaptation architecture, and complete the photovoltaic module layout optimization.

[0066] The above-mentioned layout architecture update is specifically based on Figure 5, which is a schematic diagram of component layout coordinate update. The process first calculates the effective incident irradiance assessment value on the site grid, and then calculates the component efficiency based on it. The hourly power estimate is compared with the real-time output and cascade data to verify the model. At the same time, the irradiance assessment value samples are clustered and the mean, standard deviation and within-group coefficient of variation of each cluster are calculated to quantify the light reception consistency and mismatch risk. On the spatial side, a set of peripheral candidate layout areas is generated based on terrain and obstacle data, and effective candidate areas are selected by center point distance and buffer threshold to form a micro-site selection set with low modification cost. Then, the energy efficiency improvement ratio is jointly evaluated by the candidate areas and the variability of each cluster. If the improvement ratio of the optimal candidate area reaches the preset expected threshold, the coordinate update and parameter correction suggestions are output to realize conditional iterative optimization of component position and layout parameters.

[0067] Specifically, the photovoltaic module layout architecture is adjusted according to the hourly energy effective status of the photovoltaic modules. The specific adjustment process is as follows: when the effective incident irradiance assessment value of the photovoltaic module belongs to the second effective incident irradiance assessment value interval, the third effective incident irradiance assessment value interval, and the fourth effective incident irradiance assessment value interval, the hourly energy of the photovoltaic module is determined to be in a sub-healthy state.

[0068] When the effective incident irradiance assessment value of the photovoltaic module falls within the second effective incident irradiance assessment value range, the difference between the effective incident irradiance assessment value of the photovoltaic module and the minimum value of the second effective incident irradiance assessment value range is processed to obtain the effective incident irradiance assessment value deviation of the photovoltaic module. The photovoltaic module elevation angle adjustment factor is then obtained by matching the effective incident irradiance assessment value deviation of the photovoltaic module with the photovoltaic module elevation angle adjustment factor corresponding to each predefined effective incident irradiance assessment value deviation range. This determines the specific range of the effective incident irradiance assessment value deviation of the photovoltaic module and obtains the photovoltaic module elevation angle adjustment factor corresponding to that range.

[0069] The photovoltaic module elevation angle adjustment factor is multiplied by the current elevation angle of the photovoltaic module to obtain the adaptive elevation angle of the photovoltaic module. The photovoltaic module layout architecture parameters are then updated. The adaptive elevation angle of the photovoltaic module satisfies the constraint that it is less than or equal to the defined elevation angle of the photovoltaic module. If there is an adaptive elevation angle of the photovoltaic module that is greater than the defined elevation angle of the photovoltaic module, the layout coordinates of the photovoltaic module are then updated.

[0070] When the effective incident irradiance assessment value of the photovoltaic module falls within the third effective incident irradiance assessment value range, the layout coordinates of the photovoltaic module are updated.

[0071] When the effective incident irradiance assessment value of a photovoltaic module falls within the fourth effective incident irradiance assessment value range, an early warning needs to be issued for the current layout of the photovoltaic module.

[0072] Furthermore, the layout coordinates of the photovoltaic modules are updated. The specific update process is as follows: obtain the current layout coordinates of the center point of the photovoltaic module, and calculate the straight-line distances between the current layout coordinates of the center point of the photovoltaic module and the layout coordinates of the center point of the surrounding candidate layout area set, which are recorded as buffer distances.

[0073] The aforementioned fine-tuning of the layout introduces the distance between the current component center and the center of the peripheral candidate area, which can be used as a unified proxy for the local search radius and engineering costs: on the one hand, it reduces the search space, accelerates convergence, and avoids significant changes in the solution; on the other hand, it prioritizes retaining nearby candidates (minimizes changes to pile positions / steps / drainage / roads, shorter cables with lower voltage drop, and avoids triggering compliance setback zones), while maintaining the similarity of the shading / POA (planar incident radiation) characteristics of components in the same string, reducing the risk of mismatch and hot spots; by setting buffer boundaries or distance weights, it prioritizes nearby components and only moves distant ones if they require greater benefits, achieving a feasible comprehensive optimal solution among power generation gain, construction modifications, and electrical losses.

[0074] Each buffer distance is compared with a predefined buffer boundary distance, and several peripheral candidate layout areas with a straight-line distance less than or equal to the buffer boundary distance are selected and recorded as the effective candidate area set.

[0075] The set of effective candidate regions is extracted as input to the particle swarm optimization-gravity search hybrid algorithm, which outputs the simulated energy efficiency value of the optimal candidate region.

[0076] Extract the current energy efficiency value of the photovoltaic module corresponding to the third effective incident irradiance evaluation value interval. Ratio the simulated energy efficiency value of the optimal candidate region with the current energy efficiency value of the photovoltaic module to obtain the energy efficiency improvement ratio of the optimal candidate region. Compare this ratio with the predefined expected energy efficiency improvement ratio. If the energy efficiency improvement ratio of the optimal candidate region is greater than or equal to the expected energy efficiency improvement ratio, then the coordinates corresponding to the optimal candidate region are used as the layout coordinates of the photovoltaic module for updating. If the energy efficiency improvement ratio of the optimal candidate region is less than the expected energy efficiency improvement ratio, then several peripheral candidate layout regions with a straight-line distance equal to the buffer boundary distance are removed, the effective candidate region set is updated, and the optimal candidate region is screened again until the energy efficiency improvement ratio of the optimal candidate region is greater than or equal to the expected energy efficiency improvement ratio.

[0077] In this embodiment, updating the coordinates to the optimal candidate point where the energy efficiency improvement ratio is greater than or equal to the expected threshold is equivalent to substantially "moving" the component to a locally better micro-site. This immediately results in a higher POA (less shading, better incident angle), more consistent irradiance distribution among components in the same string, reduced risk of mismatch and hot spots, and shorter DC cable paths leading to reduced voltage drop / line loss. This collectively increases annual power generation, power generation per kilowatt of installed capacity, and system performance ratio, while reducing the overall cost per kilowatt-hour. Furthermore, because the preceding buffer distance and compliance setback are used for screening, earthwork / road / electrical modifications are controllable, and the use of thresholds as acceptance criteria avoids ineffective displacement, enabling iterative optimization to improve monotonically and converge quickly.

[0078] Referring to Figure 2, the second aspect of the present invention provides a method for optimizing the layout of photovoltaic modules in complex terrain based on multi-algorithm fusion, including: S1. obtaining the real-time output power generation of the photovoltaic modules, determining the effective state of the photovoltaic module layout based on a genetic algorithm, and if it is determined to be a state that needs optimization, then proceeding to S2.

[0079] S2. Obtain the hourly solar elevation angle and hourly solar azimuth angle of each grid point under the photovoltaic module layout architecture, perform terrain and / or obstacle shading resolution, and obtain the time-terrain shading kernel of each grid point.

[0080] S3. Obtain the hourly arrangement parameters of each grid point under the photovoltaic module layout architecture, analyze the adjacent row projection of the photovoltaic module, obtain the matrix temporal mutual shading kernel of each grid point, and combine the temporal-terrain shading kernel of each grid point with the minimum value coupling algorithm to obtain the temporal-spatial shading kernel of each grid point.

[0081] S4. Obtain hourly meteorological parameters of the photovoltaic module deployment area, and couple them with the time-space shading kernel of each grid point to obtain the effective incident irradiance assessment value of each grid point. Determine the hourly energy effective state of the photovoltaic module based on the particle swarm algorithm.

[0082] S5. Adjust the photovoltaic module layout architecture according to the hourly energy availability status of the photovoltaic modules, update the photovoltaic module layout adaptation architecture, and complete the photovoltaic module layout optimization.

[0083] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion, characterized in that, include: The layout optimization judgment module is used to obtain the real-time output power generation of photovoltaic modules and determine the effective state of photovoltaic module layout based on genetic algorithm. If it is determined to be a state that needs optimization, it enters the time-terrain shading kernel resolution module. The temporal-terrain shading kernel resolution module is used to obtain the hourly solar altitude angle and the hourly solar azimuth angle of each grid point under the photovoltaic module layout architecture, perform terrain and / or obstacle shading resolution, and obtain the temporal-terrain shading kernel of each grid point. The temporal-spatial occlusion kernel parsing module is used to obtain the hourly arrangement parameters of each grid point under the photovoltaic module layout architecture, parse the adjacent row projection of the photovoltaic module, obtain the matrix temporal mutual occlusion kernel of each grid point, and combine the temporal-terrain occlusion kernel of each grid point to obtain the temporal-spatial occlusion kernel of each grid point based on the minimum value coupling algorithm. The hourly energy effective state determination module is used to obtain hourly meteorological parameters of the photovoltaic module deployment area, and couple them with the time-space shading kernel of each grid point to obtain the effective incident irradiance evaluation value of each grid point. The hourly energy effective state of the photovoltaic module is determined based on the particle swarm algorithm. The photovoltaic module deployment optimization module is used to adjust the photovoltaic module deployment architecture according to the hourly energy effective state of the photovoltaic module, update the photovoltaic module deployment adaptation architecture, and complete the photovoltaic module deployment optimization.

2. The photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion according to claim 1, characterized in that: The specific process for determining the effective state of the photovoltaic module arrangement is as follows: The photovoltaic module monitoring unit is equipped with a genetic algorithm to obtain the expected output power based on the cumulative historical output power of the photovoltaic modules; the real-time output power of the photovoltaic modules is compared with the expected output power. If the real-time output power of the photovoltaic modules is greater than or equal to the expected output power, the photovoltaic module monitoring unit determines that the photovoltaic module arrangement is in an effective state; if the real-time output power of the photovoltaic modules is less than the expected output power, a continuous window in which the real-time output power is less than the expected output power is recorded as the output weakening window of the photovoltaic modules, and compared with a predefined output weakening boundary window. A comparison is performed to obtain a weakening window comparison result, which includes a first weakening window comparison result and a second weakening window comparison result. The first weakening window comparison result indicates that the output weakening window of the photovoltaic module is less than or equal to the output weakening boundary window, and the second weakening window comparison result indicates that the output weakening window of the photovoltaic module is greater than the output weakening boundary window. If the weakening window comparison result shows the first weakening window comparison result, the photovoltaic module arrangement state is determined to be in an effective state. If the weakening window comparison result shows the second weakening window comparison result, the photovoltaic module arrangement state is determined to be in an optimization state, and photovoltaic module arrangement optimization is performed.

3. The photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion according to claim 1, characterized in that: The specific analysis process for obtaining the temporal-terrain shading kernel for each grid point is as follows: Based on the hourly solar azimuth angle of each grid point, and using the mapping table of the relationship between the solar azimuth angle and the horizon angle threshold, the hourly solar azimuth angle of each grid point is substituted into the mapping table to obtain the horizon angle threshold for each grid point, which is recorded as the horizon angle adaptation threshold for each grid point; the hourly solar altitude angle of each grid point under the photovoltaic module layout architecture is compared with the horizon angle adaptation threshold for each grid point. If the hourly solar altitude angle of a certain grid point is greater than the horizon angle adaptation threshold for each grid point, the hourly shading kernel of that grid point is recorded as 1; if the hourly solar altitude angle of a certain grid point is less than or equal to the horizon angle adaptation threshold for each grid point, the hourly shading kernel of that grid point is recorded as 0; the hourly shading kernel of each grid point is statistically analyzed and recorded as the temporal-terrain shading kernel for each grid point.

4. The photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion according to claim 3, characterized in that: The temporal-spatial shading kernel for each grid point is obtained through the following analysis process: The hourly arrangement parameters of each grid point under the photovoltaic module layout architecture include the tilt angle of the photovoltaic module plane where each grid point is located, the azimuth angle of the photovoltaic module plane where each grid point is located, the row spacing between two adjacent rows of photovoltaic module arrays, and the equivalent shading height of two adjacent rows of photovoltaic module arrays. Based on the tilt angle of the photovoltaic module plane where each grid point is located, the azimuth angle of the photovoltaic module plane where each grid point is located, the row spacing between two adjacent rows of photovoltaic module arrays, and the equivalent shading height of two adjacent rows of photovoltaic module arrays, a side view profile of the photovoltaic module array is formed. This profile is then compared with the photovoltaic module... In a cross-section perpendicular to the row direction of the photovoltaic module array, the modules in the upstream row of the photovoltaic module array are denoted as the shading body. The upper edge segment (top edge) of the shading body is used as the projection generatrix, and the current hourly solar azimuth angle is used as the projection angle. The azimuth angle corresponding to the center grid point of the shading body is denoted as the azimuth angle of the shading body. The shading body is projected onto the plane of the downstream row of modules, and the shadow area of ​​the shading body on the plane of the downstream row of modules is projected. The shadow area is rasterized onto each grid point of the downstream row of modules. If a grid point falls on the shadow area, the matrix time-series mutual shading kernel of the grid point is recorded as 0, otherwise it is recorded as 1, thus obtaining the matrix time-series mutual shading kernel of each grid point. The temporal mutual occlusion kernel of each grid point and the temporal-terrain occlusion kernel of each grid point are input into the minimum value coupling algorithm. The minimum value coupling algorithm outputs the temporal-spatial occlusion kernel of each grid point.

5. The photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion according to claim 1, characterized in that: The specific process for determining the hourly energy availability of photovoltaic modules is as follows: Irradiance curves are generated for each grid point based on the effective incident irradiance assessment values. The similarity between each grid point's irradiance curve and the irradiance curves of other grid points is calculated, and the standard deviation is processed to obtain the photovoltaic module's similarity standard deviation. Based on the photovoltaic module's similarity standard deviation, it is matched with predefined similarity clustering threshold factors corresponding to each similarity standard deviation interval to obtain a similarity clustering threshold factor. This threshold factor is then coupled with a similarity clustering reference value to obtain a similarity clustering adaptation value, configuring the clustering process for each grid point. Finally, the effective incident irradiance assessment values ​​corresponding to the grid points belonging to the same photovoltaic module are clustered to obtain the photovoltaic module. The effective incident irradiance assessment value of the photovoltaic module is compared with predefined effective incident irradiance assessment value intervals. Based on the particle swarm optimization algorithm, the effective incident irradiance assessment values ​​are screened to determine the interval to which the effective incident irradiance assessment value of the photovoltaic module belongs. The effective incident irradiance assessment value intervals include a first effective incident irradiance assessment value interval, a second effective incident irradiance assessment value interval, a third effective incident irradiance assessment value interval, and a fourth effective incident irradiance assessment value interval. If the effective incident irradiance assessment value of the photovoltaic module belongs to the first effective incident irradiance assessment value interval, then the hourly energy of the photovoltaic module is determined to be in an effective state; otherwise, the hourly energy of the photovoltaic module is determined to be in a sub-healthy state.

6. The photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion according to claim 1, characterized in that: The specific coupling process for obtaining the effective incident irradiance assessment value for each grid point is as follows: the hourly meteorological parameters of the photovoltaic module arrangement area include the hourly solar altitude angle, hourly water vapor content, and hourly cloud cover of each grid point; the hourly water vapor content and hourly cloud cover of each grid point are coupled together to obtain the hourly atmospheric transmittance influence coefficient of each grid point; the hourly solar altitude angle and hourly atmospheric transmittance influence coefficient of each grid point are normalized together with the temporal-spatial shading kernel of each grid point, and the normalization results are coupled together to obtain the effective incident irradiance assessment value of each grid point.

7. The photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion according to claim 5, characterized in that: The process of clustering the effective incident irradiance assessment values ​​corresponding to grid points belonging to the same photovoltaic module to obtain the effective incident irradiance assessment value of the photovoltaic module further includes: a clustering process of configuring an unsupervised clustering algorithm based on similarity clustering adaptation value, and semi-clustering to obtain each effective incident irradiance assessment value cluster of the photovoltaic module; calculating the standard deviation and mean of each effective incident irradiance assessment value cluster of the photovoltaic module respectively, and coupling the standard deviation and mean of the effective incident irradiance assessment value cluster of the photovoltaic module to obtain the intra-group irradiance characteristic variation coefficient of the photovoltaic module; and coupling the intra-group irradiance characteristic variation coefficient of the photovoltaic module with a predefined... The irradiance feature variation delimitation coefficient is compared. If the irradiance feature variation coefficient within a group of photovoltaic modules is less than or equal to the irradiance feature variation delimitation coefficient, then the effective incident irradiance assessment value clusters of photovoltaic modules are fully clustered to obtain the effective incident irradiance assessment value of the photovoltaic modules. If the irradiance feature variation coefficient within a group of photovoltaic modules is greater than the irradiance feature variation delimitation coefficient, then based on the difference between the irradiance feature variation coefficient within a group of photovoltaic modules and the irradiance feature variation delimitation coefficient, a similarity threshold correction factor is matched and coupled with the similarity clustering adaptation value to obtain the similarity clustering correction value. The similarity clustering adaptation value is then updated to reconfigure the clustering process.

8. The photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion according to claim 1, characterized in that: The adjustment process for the photovoltaic module layout architecture based on the hourly energy effectiveness status of the photovoltaic modules is as follows: When the effective incident irradiance assessment value of the photovoltaic module falls within the second, third, and fourth effective incident irradiance assessment value intervals, the hourly energy of the photovoltaic module is determined to be in a sub-healthy state; when the effective incident irradiance assessment value of the photovoltaic module falls within the second effective incident irradiance assessment value interval, the difference between the effective incident irradiance assessment value of the photovoltaic module and the minimum value of the second effective incident irradiance assessment value interval is calculated to obtain the effective incident irradiance assessment value deviation of the photovoltaic module, and the photovoltaic module is then matched accordingly. An elevation angle adjustment factor is coupled with the current elevation angle of the photovoltaic module to obtain the adaptive elevation angle of the photovoltaic module. The photovoltaic module layout parameters are then updated. The adaptive elevation angle of the photovoltaic module satisfies the constraint that it is less than or equal to the defined elevation angle of the photovoltaic module. If there is an adaptive elevation angle of the photovoltaic module that is greater than the defined elevation angle of the photovoltaic module, the layout coordinates of the photovoltaic module are updated. When the effective incident irradiance assessment value of the photovoltaic module belongs to the third effective incident irradiance assessment value range, the layout coordinates of the photovoltaic module are updated. When the effective incident irradiance assessment value of the photovoltaic module belongs to the fourth effective incident irradiance assessment value range, an early warning is required for the current layout architecture of the photovoltaic module.

9. The photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion according to claim 8, characterized in that: The update process for the photovoltaic module layout coordinates is as follows: Obtain the current layout coordinates of the photovoltaic module's center point; calculate the straight-line distances between the current layout coordinates of the photovoltaic module's center point and the layout coordinates of the center points of the surrounding candidate layout regions, denoted as buffer distances; compare each buffer distance with a predefined buffer boundary distance, and filter out several surrounding candidate layout regions whose straight-line distances are less than or equal to the buffer boundary distances, denoted as the effective candidate region set; extract the effective candidate region set as input to the particle swarm optimization-gravity search hybrid algorithm, which outputs the simulated energy efficiency value of the optimal candidate region; extract the photovoltaic modules corresponding to the third effective incident irradiance evaluation value interval. The current energy efficiency value is used to compare the simulated energy efficiency value of the optimal candidate region with the current energy efficiency value of the photovoltaic module to obtain the energy efficiency improvement ratio of the optimal candidate region. This ratio is then compared with the predefined expected energy efficiency improvement ratio. If the energy efficiency improvement ratio of the optimal candidate region is greater than or equal to the expected energy efficiency improvement ratio, the coordinates corresponding to the optimal candidate region are used as the layout coordinates of the photovoltaic module for updating. If the energy efficiency improvement ratio of the optimal candidate region is less than the expected energy efficiency improvement ratio, several peripheral candidate layout regions with a straight-line distance equal to the buffer boundary distance are removed, the set of valid candidate regions is updated, and the optimal candidate region is screened again until the energy efficiency improvement ratio of the optimal candidate region is greater than or equal to the expected energy efficiency improvement ratio.

10. A method for optimizing the layout of photovoltaic modules in complex terrain based on multi-algorithm fusion, applied to the photovoltaic module layout optimization system for complex terrain based on multi-algorithm fusion as described in any one of claims 1-9, characterized in that: include: S1. Obtain the real-time output power of the photovoltaic module, determine the effective state of the photovoltaic module layout based on the genetic algorithm, and if it is determined to be a state that needs optimization, proceed to S2; S2. Obtain the hourly solar altitude angle and the hourly solar azimuth angle of each grid point under the photovoltaic module layout architecture, perform terrain and / or obstacle shading resolution, and obtain the time-terrain shading kernel of each grid point. S3. Obtain the hourly arrangement parameters of each grid point under the photovoltaic module layout architecture, analyze the adjacent row projection of the photovoltaic module, obtain the matrix time-series mutual shading kernel of each grid point, and combine the time-series-terrain shading kernel of each grid point to obtain the time-series-spatial shading kernel of each grid point based on the minimum value coupling algorithm. S4. Obtain hourly meteorological parameters of the photovoltaic module deployment area, and couple them with the time-space shading kernel of each grid point to obtain the effective incident irradiance assessment value of each grid point. Determine the hourly energy effective state of the photovoltaic module based on the particle swarm optimization algorithm. S5. Adjust the photovoltaic module deployment architecture according to the hourly energy effective state of the photovoltaic module, update the photovoltaic module deployment adaptation architecture, and complete the photovoltaic module deployment optimization.

Images