A method and device for predicting irradiance and power generation for a photovoltaic fence

Abstract

The application discloses a kind of photovoltaic fence-oriented irradiance and power generation capacity prediction method and device, belong to photovoltaic irradiance calculation and power prediction technical field.The application is aimed at the special optical environment that the back of vertical double-sided photovoltaic module in photovoltaic fence system is influenced by wall reflection, ground reflection and wall-ground cavity multiple reflection, module, wall and ground are respectively dispersed into several sub-blocks, based on three-dimensional finite length view factor calculation the optical transmission relationship between each sub-block and sky view factor, combined with shadow judgment, wall-ground cavity multiple reflection matrix solution and wall non-lambertian reflection correction, realize the accurate calculation of module front, back and equivalent irradiance under photovoltaic fence scene, and further input component electrical model with equivalent irradiance and ambient temperature to carry out power prediction and power generation capacity evaluation.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic irradiance calculation and power prediction technology, specifically to a method and apparatus for predicting irradiance and power generation for photovoltaic walls. Background Technology

[0002] Photovoltaic fences are an extension of building-integrated photovoltaics (BIPV). Unlike conventional rooftop or ground-mounted bifacial solar power plants, photovoltaic fences install vertical bifacial photovoltaic modules in gaps on the back of building walls, fences, or boundary walls, maintaining an air gap between the module back and the wall surface. This creates a unique optical environment where the module, gap, wall, and ground interact. The back-side irradiance of this system is significantly influenced not only by ground reflection but also by wall reflection, multiple wall-to-ground reflections, and the non-Lambertian scattering characteristics of the wall material. Therefore, its irradiance and power generation characteristics cannot be simply modeled using traditional tilted bifacial photovoltaics or general facade photovoltaics.

[0003] Currently, there are significant deficiencies in the industry's irradiance simulation and calculation technology for photovoltaic fence scenarios. In particular, there is a lack of comprehensive modeling capabilities for multiple reflections of the wall-ground cavity, the directional scattering characteristics of the wall surface, and closed-loop prediction from irradiance to power. This makes it difficult to support the design, orientation assessment, and seasonal operation optimization of photovoltaic fence systems.

[0004] The existing irradiance calculation schemes for photovoltaic walls have the following main drawbacks:

[0005] 1. Existing two-dimensional methods are insufficient to accurately describe the true three-dimensional perspective relationship between photovoltaic modules, walls, and the ground within a photovoltaic enclosure;

[0006] 2. Existing methods generally assume that the wall surface is an ideal Lambertian surface, ignoring the directional scattering characteristics of building wall materials, which leads to errors in the calculation of wall surface reflected radiation;

[0007] 3. Existing methods typically use single-time approximations or empirical coefficients to represent multiple reflections between the wall and the ground, which makes it difficult to accurately express the radiation equilibrium state.

[0008] Therefore, a dedicated method for photovoltaic walls is needed that can accurately describe the three-dimensional view relationship and wall-to-ground multiple reflections, correct the non-Lambertian scattering characteristics of the wall surface, and ultimately achieve the engineering prediction of irradiance and power generation. Summary of the Invention

[0009] This invention aims to solve the problems in the existing technology of complex irradiance composition, large deviation in wall reflection calculation, incomplete expression of multiple reflections, and lack of integrated prediction capability from irradiance to power in photovoltaic wall scenarios. It provides a high-precision and highly engineering-available method and device for predicting irradiance and power generation for photovoltaic walls.

[0010] To achieve the above objectives, the present invention provides a method for predicting irradiance and power generation for a photovoltaic wall, wherein the photovoltaic wall includes vertically arranged bifacial photovoltaic modules and a wall and ground located on the front and back sides of the modules, respectively. The method includes the following steps:

[0011] (1) Obtain basic parameters: Obtain the system parameters of the photovoltaic fence and the meteorological data of the installation location; the system parameters include the geometric parameters, optical parameters, azimuth angle, ground clearance, distance between the component and the wall, lateral position relative to the wall, and reflectivity of the wall and the ground; the meteorological data include total horizontal irradiance, horizontal diffuse irradiance, ambient temperature and wind speed;

[0012] (2) Calculate the sun's position: Calculate the sun's zenith angle and azimuth angle based on the latitude and longitude of the installation location and the current time;

[0013] (3) Surface segmentation and perspective factor calculation: The component, the ground and the wall are divided into several sub-blocks respectively, and the mutual perspective factors between each sub-block and the sky perspective factors of each sub-block relative to the sky area and the distant ground are calculated; the mutual perspective factors include ground-wall perspective factor, wall-ground perspective factor, component-wall perspective factor, wall-component perspective factor, component-ground perspective factor and ground-component perspective factor;

[0014] The mutual perspective factor is calculated based on an analytical formula between three-dimensional finite-length rectangular surfaces, specifically:

[0015] The method for calculating the viewing angle factor between two rectangular surface units on a parallel plane is as follows: Equations (1)-(2) give the viewing angle factor FR ki from wall unit k to component unit i.

[0016] (1)

[0017] (2)

[0018] In the formula, the superscript "R" represents the back of the component; x, y, z represent the coordinates on the coordinate axes, the subscripts "1, 2" represent the points closer to and farther from the origin of this segment vertex, and the superscripts "i, j, k" represent the surface segments on the component, the ground, and the wall, respectively; M has no special meaning.

[0019] The method for calculating the viewing angle factor between rectangular surface elements on a vertical plane is as follows: Formulas (3)-(4) give the viewing angle factor FRjk from ground element j to wall element k.

[0020] (3)

[0021] (4)

[0022] In the formula, x, y, z represent the coordinates on the coordinate axes, the subscripts "1, 2" represent the points that are closer to and farther from the origin, the superscripts "i, j, k" represent the surface blocks on the component, the ground, and the wall, respectively; K has no special meaning.

[0023] Furthermore, the reverse viewing factor between the two surfaces can be derived through the reciprocal relationship. Formula (5) gives the calculation of the distinctness of the reverse viewing factor FRkj from wall element k to ground element j, and the reverse viewing factor FRjk from ground element j to wall element k:

[0024] (5).

[0025] The sky view factor is determined as follows: For each component sub-block i, all wall sub-blocks k are traversed, and the center height of each wall sub-block k is compared with that of the component sub-block i. If the center height of the wall sub-block k is higher than that of the component sub-block i, the view factor corresponding to the wall sub-block k is included in the sky view factor of the component sub-block i; otherwise, it is included in the distant ground view factor.

[0026] (4) Shadow judgment and initial irradiance calculation: The shadow occlusion status of each sub-block is determined according to the position of the sun, and the initial irradiance of the front of the component, the initial irradiance of each sub-block on the ground and the initial irradiance of each sub-block on the wall are calculated respectively. The shadow occlusion status is determined by the parallel light projection geometric model: the three-dimensional vertices of the component and the wall are projected onto the plane where the ground is located and the plane where the wall is located along the light propagation direction, respectively, to generate shadow polygons. For any target sub-block, its occlusion ratio is defined as the ratio of the intersection area of ​​the rectangular area of ​​the sub-block and the shadow polygon to the area of ​​the sub-block.

[0027] (5) Calculation of multiple reflections of wall-ground cavity: Construct a set of linear equations for the reflection irradiance of each sub-block of the ground and wall, and solve the final reflection irradiance under the equilibrium state of multiple reflections in matrix form;

[0028] The method for constructing the linear equation system of reflected irradiance for each sub-block of the multi-ground and wall surfaces is as follows:

[0029] The walls, floor, and components form a semi-enclosed cavity; the components have an anti-reflective coating with negligible reflection, while multiple reflections between the walls and floor establish an equilibrium state; analysis from optical and thermal perspectives:

[0030] The initial irradiance GRj,0 of the ground surface patch j is the surface irradiance without any reflection, as shown in equation (6).

[0031] (6)

[0032] In the formula, GHI is the total horizontal irradiance; DHI is the horizontal scattered irradiance; GRj,cir is the irradiance of the annular region received by ground unit j; GRj,iso is the isotropic scattered irradiance received by ground unit j, calculated using the Perez model; SRj is the shadow ratio of ground unit j; FR sky-j is the sky view factor of ground unit j; the sun is considered as an infinitely distant light source, and the shadow calculation adopts the parallel light projection geometry model, which projects the three-dimensional vertices of the components and walls onto the target plane along the light propagation direction to generate shadow polygons, where the wall Z = 0 and the ground Y = 0; the shading ratio SRj is defined as the ratio of the intersection area of ​​the ground patch rectangle and the shadow polygon;

[0033] The initial irradiance of wall unit k is the surface irradiance after any reflection, as shown in equation (7).

[0034] (7)

[0035] In the formula, RR b is the ratio of tilted direct irradiance to horizontal direct irradiance; FR sky-k is the sky view factor of wall element k; GR k,hor is the horizon scattered irradiance received by wall element k.

[0036] The irradiance GRj ultimately reflected outward by ground unit j is as shown in equation (8).

[0037] (8)

[0038] In the formula, ρ gnd FRkj is the reflectivity of the ground; FRkj is the viewing angle factor of wall element k towards ground element j; GRk is the irradiance ultimately reflected outward by wall element k, calculated as shown in equation (9).

[0039] (9);

[0040] Parameter η hapke The non-Lambertian correction factor is defined as the wall surface reflected irradiance. 6. A method for predicting irradiance and power generation for a photovoltaic wall according to claim 5, characterized in that the non-Lambertian correction factor is calculated based on the Hapke model as follows:

[0041] The Hapke model is as follows (10)

[0042] (10)

[0043] Where μ0 and μ are the cosines of the incident angle and the exit angle / observation angle, respectively; g is the phase angle; B(g) is the backscattering function; w is the average single-scattering albedo of the medium; since multi-angle BRDF measurement data of the actual wall coating cannot be obtained, the optical properties of the wall are characterized by effective Hapke parameters; the short-wavelength hemispherical reflectivity of the wall is set as ρ. wall Assuming it is an optically thick, non-transmittent, semi-infinite granular medium with approximately isotropic scattering, the effective single-scattering albedo is estimated using the commonly used Hapke approximation formula as shown in formula (11):

[0044] (11)

[0045] P(g) is the average particle phase function, as shown in formula (12):

[0046] (12)

[0047] Where b = -0.4 and c = 0.25 are the approximate average values ​​of the forward and side-scattering minerals, respectively;

[0048] H(μ0) and H(μ) are multiple scattering functions, as shown in equation (13):

[0049] (13)

[0050] Parameter η hapke The non-Lambertian correction factor, defined as the wall surface reflected irradiance, is calculated as shown in equation (14):

[0051] (14).

[0052] (6) Non-Lambertian Reflection Correction of Wall Surface: Based on the material properties of the wall surface, a non-Lambertian correction factor is applied to the irradiance reflected outward by each sub-block of the wall surface to obtain the corrected wall surface irradiance.

[0053] (7) Calculation of irradiance and equivalent irradiance on the back of the component: Based on the corrected wall reflected irradiance and the ground reflected irradiance obtained in step (5), the irradiance on the back of the component is calculated, and the equivalent irradiance of the component is calculated in combination with the bifacial coefficient of the component.

[0054] (8) Power prediction and power generation assessment: Input the equivalent irradiance and the ambient temperature into the component electrical model to perform power prediction and power generation assessment.

[0055] The component electrical model is a five-parameter single-diode equivalent circuit model, and the power prediction is obtained by numerically solving the IV equation of the model to obtain the maximum power point.

[0056] The present invention also provides an irradiance and power generation prediction device for photovoltaic walls, used to implement the above-mentioned method, comprising:

[0057] Parameter acquisition module: used to acquire the geometric and optical parameters of the component, the wall and the ground, as well as the meteorological data of the installation location;

[0058] Solar position calculation module: used to calculate the solar zenith angle and solar azimuth angle based on the latitude, longitude and time of the installation location;

[0059] Viewpoint factor calculation module: used to divide the components, ground and wall into sub-blocks, and calculate the mutual viewpoint factor between sub-blocks and the sky viewpoint factor of each sub-block;

[0060] Irradiance calculation module: used to calculate the initial irradiance of the front of the component, the shadow shading ratio of each sub-block, the irradiance of the multi-reflection balance state between the wall and the ground, and to apply non-Lambertian correction to the wall reflection;

[0061] Equivalent irradiance and power prediction module: used to calculate the equivalent irradiance based on the irradiance of the front and back of the component, and to perform power prediction and power generation assessment in combination with the ambient temperature;

[0062] Output module: Used to output the equivalent irradiance, power prediction results, and power generation assessment results.

[0063] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.

[0064] The present invention also provides an electronic device, comprising:

[0065] Processor; and

[0066] Memory for storing the executable instructions of the processor;

[0067] The processor is configured to execute the above method by executing the executable instructions.

[0068] Compared with the prior art, the present invention has the following advantages:

[0069] 1. Dedicated Modeling for Photovoltaic Walls: This invention establishes a dedicated model for special scenarios where there are wall reflections and wall-to-ground cavity reflections on the back of the wall, which is more consistent with the actual optical structure of photovoltaic walls than the general facade photovoltaic model.

[0070] 2. Considering complex reflections and directional scattering: This invention not only models multiple reflections between the wall and the ground, but also distinguishes the scattering characteristics of different wall materials by correcting the non-Lambertian reflection of the wall surface, thereby improving the accuracy of back-side irradiance prediction.

[0071] 3. Strong 3D finite length modeling capability: Through 3D finite length view factor and sub-block discretization, it can adapt to finite segment walls, different wall heights and different installation spacings, avoiding the deviation caused by the 2D infinite length assumption.

[0072] 4. High engineering usability: Based on analytical formulas and matrix operations, the computational complexity is controllable, making it easy to embed into design software, energy management platforms, or power plant operation optimization systems.

[0073] 5. A closed loop is formed from irradiance to power: The method not only outputs irradiance, but also further outputs power prediction and power generation prediction, which directly supports the design of photovoltaic fence projects, orientation optimization and the formulation of seasonal operation strategies. Attached Figure Description

[0074] Figure 1 is a schematic diagram showing the spatial relationship and radiation composition of the three radiation receiving surfaces: vertical bifacial photovoltaic module, ground surface, and wall surface in this invention.

[0075] Figure 2 is a schematic diagram of the relevant angle definitions involved in the irradiance calculation in this invention.

[0076] Figure 3 shows the irradiance meteorological data during the test in the experimental verification of this invention.

[0077] Figure 4 shows the scatter plots of calculated and measured front and back irradiance and power during the test; (a) shows the calculated and measured front irradiance data; (b) shows the calculated and measured back irradiance data; and (c) shows the calculated and measured power data.

[0078] Figure 5 shows a comparison of meteorological data, irradiance data, and power data under three typical weather conditions: sunny, cloudy, and overcast. (a), (e), and (i) show irradiance data under the three typical weather conditions; (b), (f), and (j) show ambient temperature and wind speed data under the three typical weather conditions; (c), (g), and (k) show a comparison of irradiance data on the front and back sides under the three typical weather conditions; and (d), (h), and (l) show a comparison of power data under the three typical weather conditions. Detailed Implementation

[0079] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0080] Example:

[0081] A method for predicting irradiance and power generation for photovoltaic (PV) fences, comprising the following steps:

[0082] The input information for the photovoltaic fence system includes the component model, component geometric parameters, component azimuth angle, ground reflectivity, wall reflectivity, distance between the component and the wall, component height above the ground, lateral position of the component relative to the wall, as well as local latitude and longitude and meteorological data (horizontal total irradiance, horizontal diffuse irradiance, ambient temperature, and wind speed).

[0083] The sun's position (zenith angle, azimuth angle) can be calculated based on the installation location and time, or obtained using industry-standard sun position algorithms or tools such as pvlib.

[0084] The viewing angle factor (VF), defined as the proportion of radiation leaving surface 1 that is intercepted by surface 2, is the basis for irradiance calculation. This study employs an analytical formula for the three-dimensional viewing angle factor (Ehlert & Smith, 1993; Gross et al., 1981), with particular emphasis on the calculation of irradiance on the back side. For the front side irradiance, a simplified Perez model or the same calculation method as for the back side is used to achieve higher accuracy.

[0085] Figure 2 defines the coordinate system and variables for calculating the viewpoint factor. The x-axis extends along the width of the wall / component, the y-axis represents the vertical height, and the z-axis represents the depth (the wall is located at z = 0, and the component is located at z = D).

[0086] Mesh sensitivity analysis was performed by progressively refining the wall and ground surfaces from 2×2 to 30×30, while keeping the component discretization fixed at 6×22 (corresponding to the physical cell layout, see Table 1). The spatial distribution of the view factor and sky view factor under each mesh was evaluated. The results show that when the mesh is refined from 2×2 to 5×5, the change in the spatial view factor distribution decreases significantly, and further decreases from 5×5 to 7×7. After the 7×7 mesh, the rate of decrease gradually diminishes. Further refinement from 7×7 to 10×10 produces only limited improvement, while increasing the mesh density beyond 10×10 leads to a significant increase in the number of surface units, resulting in a significant increase in computational cost. Therefore, a 10×10 discretization for the wall and ground surfaces is chosen as a suitable trade-off between the accuracy of the spatial view factor representation and computational efficiency.

[0087] Table 1

[0088]

[0089] The VF between two rectangular surface units on a parallel plane, such as the VF between a wall and a module, is given by equations (1)-(2), where FR ki is the VF from wall unit k to component unit i.

[0090]

[0091]

[0092] In the formula, the superscript "R" represents the back of the component; x, y, z represent the coordinates on the coordinate axes, the subscripts "1, 2" represent the points closer to and farther from the origin of this segment vertex, respectively, and the superscripts "i, j, k" represent the surface segments on the component, the ground, and the wall, respectively; M has no special meaning. See Figure 2 for detailed definitions.

[0093] The calculation method for the viewing factor between rectangular surface units on the vertical plane (such as between the ground and the wall or between components) is similar. Formulas (3)-(4) give the viewing factor FRjk from the ground unit j to the wall unit k. The calculation method for the viewing factor FRji from the ground to the component is similar and will not be described in detail here.

[0094]

[0095]

[0096] In addition to the direct calculations mentioned above, the reverse viewing angle factor between two surfaces can be derived through the reciprocal relationship. Formula (5) gives the reverse viewing angle factor FRkj from wall element k to ground element j, and the reverse viewing angle factor FRjk from ground element j to wall element k. The calculation of the dissimilarity of these factors is also provided.

[0097]

[0098] The above VF calculation applies to a limited ground area. However, distant ground areas also contribute and should be included. The effective rear sky view factor VF value FR sky-i and the distant ground view factor VF value F Rgnd-dis-i are calculated block by block. For each component element i, all wall elements k are traversed and their center heights are compared. If the center height of wall element k exceeds the center height of component element i (ycenter k > ycenteri), the VF term is attributed to the sky view factor; otherwise, it is attributed to the distant ground view factor, as in equations (6)-(7).

[0099]

[0100]

[0101] Irradiance calculation requires local meteorological data, which can be obtained from local meteorological databases, field measurements, or databases such as Meteonorm. A simplified Perez model is used to calculate the frontal irradiance G. F The focus is on the complex calculation of rear-side irradiance G. R As shown in equation (8).

[0102]

[0103] GR b represents the direct irradiance received from the back; GR d represents the diffuse irradiance received from the back; GR r represents the reflected irradiance received from the back; IAM R b represents the direct incident angle correction; IAM R d represents the diffuse incident angle correction.

[0104] While a rough wall surface is often considered a near-ideal Lambertian plane, in reality, at low solar angles, the irradiance in the direction of wall reflection is significantly higher than in other directions. This study introduces bidirectional reflectance to account for this factor. Bidirectional reflectance refers to the ratio of the irradiance received per unit area and unit solid angle to the irradiance per unit area of ​​the light source when the detector observes the surface from another specific direction under parallel light illumination from a specific direction. The core formula for calculating bidirectional reflectance (i.e., the Hapke model) is used. As shown in equation (9).

[0105]

[0106] Where μ0 and μ are the cosines of the incident angle and the exit angle / observation angle, respectively; g is the phase angle (the angle between the incident direction and the observation direction); and B(g) is the backscattering function. Due to the unavailability of multi-angle BRDF measurement data for actual wall coatings, the angle width parameter h cannot be determined. Previous photometric studies, including the work of Mustard and Pieters (Mustard & Pieters, 1989), have shown that the backscattering effect is mainly observed at small phase angles. In VBPV-F systems, the effective wall-to-component geometry is predominantly based on relatively large phase angles, and cases where g < 15° are uncommon. Therefore, B(g) is neglected in the final model.

[0107] w is the average single-scattering albedo of the medium. Since multi-angle BRDF measurement data for actual wall coatings are unavailable, the optical properties of the wall are characterized using an effective Hapke parameter. The short-wavelength hemispherical reflectivity of the wall is set as ρ. wallAssuming an optically thick, non-transmittent, semi-infinite granular medium with approximately isotropic scattering, the effective single-scattering albedo is estimated using the commonly used Hapke approximation (Hapke, 1981) (Equation (10)):

[0108]

[0109] P(g) is the average particle phase function (Equation (11)):

[0110]

[0111] Where b = -0.4 and c = 0.25, these are approximate average values ​​for forward and lateral scattering minerals. Mustard and Pieters showed that the coefficients b describing the degree of forward and backward scattering and c describing the degree of lateral scattering are different for all minerals. To simplify the model, we choose to use the average values ​​of b and c from Mustard and Pieters, which have been adopted by Lemelin (Lemelin et al., 2013).

[0112] H(μ0) and H(μ) are multiple scattering functions (Equation (12)).

[0113]

[0114] Parameter η hapke The non-Lambert correction factor, defined as the wall surface reflectance, is calculated as shown in Equation (13).

[0115]

[0116] The walls, floor, and components form a semi-enclosed cavity. The components have an anti-reflective coating with negligible reflection, while multiple reflections between the walls and floor establish an equilibrium. This requires analysis from both optical and thermal perspectives.

[0117] The initial irradiance GRj,0 of the ground patch j (i.e., the surface irradiance without any reflection) is as shown in equation (14).

[0118]

[0119] In the formula, GHI is the total horizontal irradiance; DHI is the horizontal scattered irradiance; GRj,cir is the irradiance scattered around the sun received by ground unit j; GRj,iso is the isotropic scattered irradiance received by ground unit j, which can be calculated using the Perez model; SRj is the shadow ratio of ground unit j; FR sky-j is the sky view factor of ground unit j; the sun is considered as an infinitely distant light source, and the shadow calculation adopts the parallel light projection geometry model. The three-dimensional vertices of the components and walls are projected onto the target plane (wall Z = 0, ground Y = 0) along the light propagation direction to generate shadow polygons. The patch shading ratio SRj is defined as the ratio of the intersection area of ​​the ground patch rectangle and the shadow polygon.

[0120] The initial irradiance of wall unit k (i.e., the surface irradiance after any reflection) is as shown in equation (15).

[0121]

[0122] In the formula, RR b is the ratio of tilted direct irradiance to horizontal direct irradiance; FR sky-k is the sky view factor of wall element k; GR k,hor is the horizon diffuse irradiance received by wall element k.

[0123] The irradiance GRj ultimately reflected outward by ground unit j is as shown in equation (16).

[0124]

[0125] In the formula, ρ gnd FRkj is the reflectivity of the ground; FRkj is the viewing angle factor of wall element k towards ground element j; GRk is the irradiance ultimately reflected outward by wall element k, calculated as shown in equation (17).

[0126]

[0127] Finally, the detailed formula for calculating the average irradiance on the back side is as follows: .

[0128]

[0129] The photovoltaic module has multiple sub-strings protected by bypass diodes. Within each sub-string, the cells are connected in series, and the current is limited by the darkest cell in that sub-string. This paper takes the minimum irradiance of all cells within a sub-string as its equivalent irradiance. The final equivalent irradiance is the average of all sub-strings.

[0130] Using the equivalent irradiance G as the subsequent power input parameter, the calculation is as follows: .

[0131]

[0132] In the formula, φ is the bifaciality coefficient of the module, obtained from the module manufacturer.

[0133] To balance computational speed and model complexity, the thermal model uses the classic nominal operating cell temperature (NOCT) relation (Davis et al., 2001) as a first-order engineering approximation. The cell temperature is calculated as shown in the formula:

[0134]

[0135] Where T cell and T ambient Here, η represents the cell temperature, and τα represents the temperature of the solar cell; τα is usually unknown, but is typically taken as 0.9. Since the η / τα term is relatively small, the resulting error is not significant.

[0136] The electrical output is calculated using a five-parameter single diode equivalent circuit model, Equation (21).

[0137]

[0138] First, five parameters are extracted based on reference conditions, following a commonly used five-parameter extraction model in the industry. Then, the calculated operating conditions are corrected. The operating point is obtained by numerically solving the implicit IV equation, and the maximum power point is selected from P = VI.

[0139] Experimental verification example:

[0140] The power comparison involved in the experimental verification can be calculated using the industry-standard single diode five-parameter model, and the working conditions can be corrected by combining the equivalent irradiance and ambient temperature.

[0141] Figure 3 shows the meteorological data during the test.

[0142] Figure 4 shows the scatter plots of calculated and measured irradiance and power on the front and back sides during the test period. This comparison illustrates the accuracy of the proposed method in calculating irradiance and power output in the photovoltaic fence scenario. The errors in irradiance and power on the front and back sides of the overall data over multiple days all meet the engineering acceptance requirements, indicating that the model has stable predictive capabilities under various weather conditions and can support the design and operational evaluation of photovoltaic fence systems.

[0143] Figure 5 shows a comprehensive evaluation of the accuracy of the photovoltaic power prediction model. This invention selects a measured dataset with typical meteorological abrupt changes for model validation. The validation results show that the model can follow the measured changes well under sunny, cloudy, and overcast conditions, further demonstrating that this method is suitable for systems like photovoltaic fences with complex back-reflection sources.

[0144] Overall, the daily power prediction results show good fit. In the back-side irradiance prediction, if only the Lambertian assumption is used, the contribution of some wall reflections may be underestimated, especially when the wall material has certain directional scattering characteristics; after introducing the non-Lambertian reflection correction of the wall, the model can more closely approximate the actual optical behavior of the photovoltaic wall.

[0145] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for predicting irradiance and power generation for a photovoltaic (PV) fence, wherein the PV fence comprises vertically arranged bifacial PV modules and a wall and ground respectively located on the front and back sides of the modules, characterized in that, The method includes the following steps: (1) Obtain basic parameters: Obtain the system parameters of the photovoltaic fence and the meteorological data of the installation location; the system parameters include the geometric parameters, optical parameters, azimuth angle, ground clearance, distance between the component and the wall, lateral position relative to the wall, and reflectivity of the wall and the ground; the meteorological data include total horizontal irradiance, horizontal diffuse irradiance, ambient temperature and wind speed; (2) Calculate the sun's position: Calculate the sun's zenith angle and azimuth angle based on the latitude and longitude of the installation location and the current time; (3) Surface segmentation and perspective factor calculation: The component, the ground and the wall are divided into several sub-blocks respectively, and the mutual perspective factors between each sub-block and the sky perspective factors of each sub-block relative to the sky area and the distant ground are calculated; the mutual perspective factors include ground-wall perspective factor, wall-ground perspective factor, component-wall perspective factor, wall-component perspective factor, component-ground perspective factor and ground-component perspective factor; (4) Shadow judgment and initial irradiance calculation: The shadow occlusion status of each sub-block is determined according to the position of the sun, and the initial irradiance of the front of the component, the initial irradiance of each sub-block on the ground and the initial irradiance of each sub-block on the wall are calculated respectively. (5) Calculation of multiple reflections of wall-ground cavity: Construct a set of linear equations for the reflection irradiance of each sub-block of the ground and wall, and solve the final reflection irradiance under the equilibrium state of multiple reflections in matrix form; (6) Non-Lambertian Reflection Correction of Wall Surface: Based on the material properties of the wall surface, a non-Lambertian correction factor is applied to the irradiance reflected outward by each sub-block of the wall surface to obtain the corrected wall surface irradiance. (7) Calculation of irradiance and equivalent irradiance on the back of the component: Based on the corrected wall reflected irradiance and the ground reflected irradiance obtained in step (5), the irradiance on the back of the component is calculated, and the equivalent irradiance of the component is calculated in combination with the bifacial coefficient of the component. (8) Power prediction and power generation assessment: Input the equivalent irradiance and the ambient temperature into the component electrical model to perform power prediction and power generation assessment.

2. The method for predicting irradiance and power generation for a photovoltaic wall according to claim 1, characterized in that, In step (3), the mutual viewing angle factor is calculated based on the analytical formula between three-dimensional finite-length rectangular surfaces, specifically: The method for calculating the viewing angle factor between two rectangular surface units on a parallel plane is as follows: Equations (1)-(2) give the viewing angle factor FR ki from wall unit k to component unit i. （1） （2） In the formula, the superscript "R" represents the back of the component; x, y, z represent the coordinates on the coordinate axes, the subscripts "1, 2" represent the points closer to and farther from the origin of this segment vertex, respectively, and the superscripts "i, j, k" represent the surface segments on the component, the ground, and the wall, respectively; M has no special meaning. The method for calculating the viewing angle factor between rectangular surface elements on a vertical plane is as follows: Formulas (3)-(4) give the viewing angle factor FRjk from ground element j to wall element k. （3） （4） In the formula, x, y, z represent the coordinates on the coordinate axes, the subscripts "1, 2" represent the points that are closer to and farther from the origin, the superscripts "i, j, k" represent the surface blocks on the component, the ground, and the wall, respectively; K has no special meaning. Furthermore, the reverse viewing factor between the two surfaces can be derived through the reciprocal relationship. Formula (5) gives the calculation of the distinctness of the reverse viewing factor FRkj from wall element k to ground element j, and the reverse viewing factor FRj-k from ground element j to wall element k: （5）。 3. The method for predicting irradiance and power generation for a photovoltaic fence according to claim 1, characterized in that, In step (3), the sky view factor is determined in the following way: for each component sub-block i, all wall sub-blocks k are traversed, and the center height of each wall sub-block k is compared with that of the component sub-block i. If the center height of the wall sub-block k is higher than that of the component sub-block i, the view factor corresponding to the wall sub-block k is included in the sky view factor of the component sub-block i; otherwise, it is included in the distant ground view factor.

4. The method for predicting irradiance and power generation for a photovoltaic wall according to claim 1, characterized in that, In step (4), the shadow occlusion state is determined by the parallel light projection geometric model: the three-dimensional vertices of the component and the wall are projected along the light propagation direction onto the plane where the ground is located and the plane where the wall is located, respectively, to generate shadow polygons. For any target sub-block, its occlusion ratio is defined as the ratio of the intersection area of ​​the rectangular area of ​​the sub-block and the shadow polygon to the area of ​​the sub-block.

5. The method for predicting irradiance and power generation for a photovoltaic wall according to claim 1, characterized in that, In step (5), the method for constructing the linear equation system of reflected irradiance for each sub-block of the multi-ground and wall surfaces is as follows: The walls, floor, and components form a semi-enclosed cavity; the components have an anti-reflective coating with negligible reflection, while multiple reflections between the walls and floor establish an equilibrium state; analysis from optical and thermal perspectives: The initial irradiance GRj,0 of the ground patch j, i.e. the surface irradiance without any reflection, is as shown in equation (6): （6） In the formula, GHI is the total horizontal irradiance; DHI is the horizontal scattered irradiance; GRj,cir is the irradiance of the annular region received by ground unit j; GRj,iso is the isotropic scattered irradiance received by ground unit j, calculated using the Perez model; SRj is the shadow ratio of ground unit j; FR sky-j is the sky view factor of ground unit j; the sun is considered as an infinitely distant light source, and the shadow calculation adopts the parallel light projection geometry model, which projects the three-dimensional vertices of the components and walls onto the target plane along the light propagation direction to generate shadow polygons, where the wall Z = 0 and the ground Y = 0; the shading ratio SRj is defined as the ratio of the intersection area of ​​the ground patch rectangle and the shadow polygon; The initial irradiance of wall unit k is the surface irradiance after any reflection, as shown in equation (7): （7） In the formula, RR b is the ratio of tilted direct irradiance to horizontal direct irradiance; FR sky-k is the sky view factor of wall element k; GR k,hor is the horizon scattered irradiance received by wall element k. The irradiance GRj ultimately reflected outward by ground unit j is as shown in equation (8): （8） In the formula, ρ gnd Let be the reflectivity of the ground; FR kj be the viewing angle factor of wall element k towards ground element j; GR k be the irradiance ultimately reflected outward by wall element k, calculated as shown in equation (9): （9）； Parameter η hapke It is defined as the non-Lambertian correction factor for wall reflectance.

6. The method for predicting irradiance and power generation for a photovoltaic wall according to claim 5, characterized in that, The non-Lambertian correction factor is calculated based on the Hapke model as follows: The Hapke model is as follows (10) （10） Where μ0 and μ are the cosines of the incident angle and the exit angle / observation angle, respectively; g is the phase angle; B(g) is the backscattering function; w is the average single-scattering albedo of the medium; since multi-angle BRDF measurement data of the actual wall coating cannot be obtained, the optical properties of the wall are characterized by effective Hapke parameters; the short-wavelength hemispherical reflectivity of the wall is set as ρ. wall Assuming it is an optically thick, non-transmittent, semi-infinite granular medium with approximately isotropic scattering, the effective single-scattering albedo is estimated using the commonly used Hapke approximation formula as shown in formula (11): （11） P(g) is the average particle phase function, as shown in formula (12): （12） Where b = -0.4 and c = 0.25 are the approximate average values ​​of the forward and side-scattering minerals, respectively; H(μ0) and H(μ) are multiple scattering functions, as shown in equation (13): （13） Parameter η hapke The non-Lambertian correction factor, defined as the wall surface reflected irradiance, is calculated as shown in equation (14): （14）。 7. The method for predicting irradiance and power generation for a photovoltaic wall according to claim 1, characterized in that, In step (8), the component electrical model is a five-parameter single diode equivalent circuit model, and the power prediction is obtained by numerically solving the IV equation of the model to obtain the maximum power point.

8. A device for predicting irradiance and power generation for a photovoltaic wall, used to implement the method as described in any one of claims 1 to 7, characterized in that, include: Parameter acquisition module: used to acquire the geometric and optical parameters of the component, the wall and the ground, as well as the meteorological data of the installation location; Solar position calculation module: used to calculate the solar zenith angle and solar azimuth angle based on the latitude, longitude and time of the installation location; Viewpoint factor calculation module: used to divide the components, ground and wall into sub-blocks, and calculate the mutual viewpoint factor between sub-blocks and the sky viewpoint factor of each sub-block; Irradiance calculation module: used to calculate the initial irradiance of the front of the component, the shadow shading ratio of each sub-block, the irradiance of the multi-reflection balance state between the wall and the ground, and to apply non-Lambertian correction to the wall reflection; Equivalent irradiance and power prediction module: used to calculate the equivalent irradiance based on the irradiance of the front and back of the component, and to perform power prediction and power generation assessment in combination with the ambient temperature; Output module: Used to output the equivalent irradiance, power prediction results, and power generation assessment results.

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

10. An electronic device, characterized in that, include: processor; as well as Memory for storing the executable instructions of the processor; The processor is configured to perform the method of any one of claims 1 to 7 by executing the executable instructions.

Images