A flexible support photovoltaic assembly mounting device and method

Abstract

The application discloses a flexible support photovoltaic module installation device and installation method, and belongs to the technical field of solar photovoltaic power generation. The application solves the problems of very tedious operation and low efficiency of the mode of manually installing the flexible support photovoltaic module. The application comprises: a photovoltaic module clamping system, comprising a photovoltaic panel clamping mechanism for connecting the photovoltaic panel and a moving mechanism for moving the photovoltaic panel to the flexible support according to a set angle; a flexible support adjusting mechanism for adjusting the form of the flexible support according to the installation requirement of the photovoltaic panel; and a walking system for moving the photovoltaic module clamping system and the flexible support adjusting mechanism to the installation position of the photovoltaic panel on the flexible support. The application realizes the automatic installation of the photovoltaic panel on the flexible support, greatly reduces the installation efficiency of the photovoltaic panel on the flexible support, and ensures the installation precision of the photovoltaic panel on the flexible support to some extent, so that the photovoltaic panel can be uniformly arranged on the flexible support.

Description

Technical Field

[0001] This invention belongs to the field of solar photovoltaic power generation technology, specifically relating to a flexible bracket photovoltaic module installation device and installation method. Background Technology

[0002] Solar photovoltaic (PV) power generation is a crucial technological means to achieve the "dual carbon" goal (carbon reduction, carbon emission reduction, and carbon sequestration). New large-span flexible support systems composed of prestressed cables are constantly maturing and being innovated, making them suitable for various complex environments such as tidal flats, fishponds, farmland, and deserts. Compared to fixed PV support systems, which suffer from drawbacks such as a large number of supports, high steel consumption, high cost, high requirements for site flatness, poor terrain adaptability, and limited space under the panels, flexible PV support systems offer advantages such as low steel consumption, strong pre-installation capability, high adaptability, large space under the panels, and high land utilization. Currently, the installation of flexible cable systems mainly relies on manual installation. This involves manually lifting the PV panels above the double cables and then manually tightening the locking buckles. Installing one panel requires securing four buckles to the flexible cables. Therefore, manual installation is extremely cumbersome and inefficient.

[0003] In the prior art, CN202511202376.5 discloses a method and equipment for installing photovoltaic panels and their supporting components. It uses a carrier to pull a cable and keep the cable under tension. Then, a robotic arm is used to move the photovoltaic panel to the installation position. After the photovoltaic panel is in place, fasteners are transported to the corresponding installation position of the photovoltaic panel for installation. Then, the carrier moves forward to make room for the installation of the next set of photovoltaic panels. The operation of this device is relatively cumbersome. Summary of the Invention

[0004] To address the problems of cumbersome operation and low efficiency in existing flexible bracket photovoltaic modules, this invention provides a flexible bracket photovoltaic module installation device and method.

[0005] The technical solution adopted in this invention is as follows:

[0006] A flexible support photovoltaic module installation device, comprising:

[0007] A photovoltaic module clamping system includes a photovoltaic panel clamping mechanism for connecting photovoltaic panels and a moving mechanism for moving the photovoltaic panels onto a flexible support at a set angle; a flexible support adjustment mechanism for adjusting the shape of the flexible support according to the installation requirements of the photovoltaic panels; and a walking system for moving the photovoltaic module clamping system and the flexible support adjustment mechanism to the installation position of the photovoltaic panels on the flexible support.

[0008] Preferably, the moving mechanism includes a height adjusting component and an angle adjusting component. The height adjusting component includes an upper support connected to the walking system. The upper support is connected to a lower support via a scissor lift mechanism. The lower support is connected to the angle adjusting component, and the angle adjusting component is connected to the photovoltaic panel clamping mechanism. The scissor lift mechanism includes a lifting cylinder.

[0009] Preferably, the angle adjustment component includes a support leg, which is hinged to one end of the photovoltaic panel clamping mechanism, and the other end of the photovoltaic panel clamping mechanism is hinged to an angle cylinder.

[0010] Preferably, the photovoltaic panel clamping mechanism includes a mounting frame connected to the moving mechanism, and the bottom of the mounting frame is provided with several vacuum suction cups.

[0011] Preferably, the vacuum suction cups are divided into several vacuum suction cup groups, each vacuum suction cup group corresponds to a photovoltaic panel, and the several vacuum suction cup groups are arranged in a row with equal spacing.

[0012] Preferably, the flexible support adjustment mechanism includes four robotic arms mounted on the photovoltaic panel clamping mechanism, with the four robotic arms respectively positioned at the four corners of the front and rear ends of the photovoltaic panel clamping mechanism.

[0013] Preferably, the flexible support adjustment mechanism further includes a visual camera for real-time detection and acquisition of the spatial position of the component cable.

[0014] Preferably, the walking system includes a horizontal moving frame and a vertical moving frame. The horizontal moving frame is arranged perpendicular to the flexible support, and the vertical moving frame is located directly above the flexible support and extends along the length of the flexible support. The vertical moving frame is arranged perpendicular to the horizontal moving support and can move along the horizontal moving frame. A moving trolley is arranged on the vertical moving frame, and the photovoltaic module clamping system is connected to the moving trolley.

[0015] Preferably, the flexible support includes two component cables, the photovoltaic panel includes a photovoltaic panel body, and the bottom surface of the photovoltaic panel body is provided with several cable buckles that cooperate with the component cables.

[0016] A method for installing flexible bracket photovoltaic modules, using a flexible bracket photovoltaic module installation device, includes the following steps:

[0017] S1: Connect the photovoltaic panel through the photovoltaic panel clamping mechanism of the photovoltaic module clamping system;

[0018] S2: The walking system drives the photovoltaic panel and photovoltaic module clamping system and the flexible support adjustment mechanism to move to the installation position of the photovoltaic panel on the flexible support;

[0019] S3: The flexible support adjustment mechanism adjusts the shape of the flexible support according to the installation requirements of the photovoltaic panels;

[0020] S4: The moving mechanism of the photovoltaic module clamping system adjusts the height and angle of the photovoltaic panel according to the shape of the flexible support, so that the photovoltaic panel contacts the flexible support;

[0021] S5: Complete the connection between the photovoltaic panel and the flexible support.

[0022] As a preferred approach, in S3, a finite element model of the double-layer cable-stayed flexible photovoltaic support is established using SAP2000. The cable pretension is applied by selecting the target force iteration method. Based on the geometric nonlinearity of the cable structure, the initial stiffness matrix of the double-layer cable-stayed flexible photovoltaic support is formed on the basis of the initial stiffness of the structure. The cable pulling force at the component cable node is solved iteratively. Then, the relationship curve between the component cable node displacement and the magnitude of the cable pulling force under normal installation conditions is obtained by analyzing the load optimizer. With the minimum node cable pulling force as the objective, the optimal cable pulling force and ideal displacement are obtained by linear interpolation. The adjustment of the flexible support is completed by obtaining the optimal cable pulling force and ideal displacement.

[0023] As a preferred method, the specific process of obtaining the spatial position of the component cable in real time through a vision camera is as follows: the transformation matrix between the camera pixel coordinate system and the world coordinate system is derived through the pinhole imaging model of the vision camera; the camera parameter information is obtained by calibration experiments using the Zhang Zhengyou calibration method, and the stereo correction of the image is completed; then, the image is segmented using the HSV image segmentation algorithm based on the color characteristics of the component cable; the edge contour of the steel strand is obtained through morphological operations and the Canny edge detection algorithm; and non-target areas are filtered out by setting the aspect ratio threshold of the minimum bounding rectangle of the contour, thus achieving accurate identification of the component cable.

[0024] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0025] 1. In this invention, the flexible support adjustment mechanism can adjust the shape of the flexible support to meet the installation requirements of the photovoltaic panel, while the photovoltaic module clamping system can connect the photovoltaic panel and adjust the angle and height of the photovoltaic panel according to the shape of the flexible support. The walking system can move the photovoltaic panel to a suitable position on the flexible support for installation. Finally, through the cooperation of the photovoltaic module clamping system and the flexible support adjustment mechanism, the shape of the flexible support and the shape of the photovoltaic panel are matched, realizing the automatic installation of the photovoltaic panel on the flexible support. This greatly reduces the installation efficiency of the photovoltaic panel on the flexible support and also ensures the installation accuracy of the photovoltaic panel on the flexible support to a certain extent, so that the photovoltaic panel can be evenly arranged on the flexible support.

[0026] 2. This invention can complete the installation of multiple photovoltaic panels simultaneously, further improving the installation efficiency of photovoltaic modules. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the structure from a first angle in one embodiment of the present invention;

[0028] Figure 2 for Figure 1 A magnified view of a section at point A in the middle;

[0029] Figure 3 for Figure 1 A magnified view of a section at point B in the middle;

[0030] Figure 4 This is a schematic diagram of the structure from a second angle in one embodiment of the present invention;

[0031] Figure 5 for Figure 4 A magnified view of a section at point C;

[0032] Figure 6 This is a schematic diagram of the buckle structure in one embodiment of the present invention;

[0033] Figure 7 This is a schematic diagram of the structural dimensions of the 40m span double-layer cable-stayed flexible photovoltaic support system in this invention;

[0034] Figure 8 This is a diagram of the support structure at the end of the flexible support in this invention;

[0035] Figure 9 This is a schematic diagram of the flexible support structure with applied pretension in this invention;

[0036] Figure 10 This is a diagram showing the positional relationship between the vision camera and the robotic arm in this invention;

[0037] Figure 11 This is a schematic diagram showing the naming of four nodes of a photovoltaic panel assembly in this invention.

[0038] Figure 12 This is a diagram showing the deformation of the flexible support after tensioning in this invention.

[0039] Figure 13 The diagram shows the iterative convergence process in this invention, where a is the displacement convergence process diagram and b is the concentrated force convergence process diagram.

[0040] Figure 14 The graphs show the relationship between displacement and concentrated force, and the absolute value of concentrated force, for the second and fourth groups of nodes in this invention. In the graphs, a is the relationship between displacement and concentrated force for the second group of nodes, b is the relationship between displacement and absolute value of concentrated force for the second group of nodes, c is the relationship between displacement and concentrated force for the fourth group of nodes, and d is the relationship between displacement and absolute value of concentrated force for the fourth group of nodes.

[0041] Figure 15 This is a cloud diagram showing the vertical displacement of the flexible support after the installation of the transformer panel in this invention.

[0042] Figure 16 This is a graph showing the relationship between the displacement of the second group of nodes and the absolute value of the concentrated force under different wind pressures in this invention. In the graph, a is node B1, b is node B2, c is node B3, and d is node B4.

[0043] Figure 17 This is a graph showing the relationship between the displacement of the fourth group of nodes and the absolute value of the concentrated force under different wind pressures in this invention. In the graph, a is node D1, b is node D2, c is node D3, and d is node D4.

[0044] Figure 18 This is a graph showing the relationship between the concentrated force at the nodes and the basic wind pressure in this invention, where a represents the first group of nodes and b represents the fifth group of nodes.

[0045] Figure 19 This is a graph showing the relationship between ideal displacement and basic wind pressure in this invention.

[0046] Figure 20 This is a graph showing the relationship between the optimal cable pulling force and the basic wind pressure in this invention, where a represents the second group of nodes and b represents the fourth group of nodes.

[0047] Figure 21 This is a schematic diagram of the pinhole imaging model in this invention;

[0048] Figure 22 This is a diagram showing the relationship between the two coordinate systems of the pixel plane in this invention;

[0049] Figure 23 This is a schematic diagram of the binocular imaging principle in this invention, where a is the imaging model and b is the calculation principle;

[0050] Figure 24 The diagram shows the reprojection error results of the visual camera calibration in this invention, where a represents the calibration error of the left and right infrared cameras, and b represents the calibration error of the color camera.

[0051] Figure 25 This is a diagram showing the positional relationship between the visual camera and the calibration plate in this invention, where a represents the positional relationship between the left and right infrared cameras and the calibration plate, and b represents the positional relationship between the color camera and the calibration plate.

[0052] Figure 26 This is a schematic diagram of the image stereo correction principle in this invention, where a represents the image before stereo correction and b represents the image after stereo correction.

[0053] Figure 27 This is the stereo-corrected image in this invention;

[0054] Figure 28The image shows the threshold segmentation result in this invention, where a is the RGB image of the steel strand and b is the threshold segmentation result of the steel strand.

[0055] Figure 29 This is a diagram showing the morphological operation results in this invention;

[0056] Figure 30 This is a diagram showing the edge contour detection of the steel strand in this invention, where a represents the edge detection of the steel strand and b represents the extraction of the steel strand contour.

[0057] Figure 31 This is a diagram showing the positioning points of the steel strand in this invention;

[0058] Figure 32 This is the identification and positioning detection map in this invention;

[0059] Among them, 1-horizontal moving frame, 2-vertical moving frame, 3-upper support, 4-moving trolley, 5-installation frame, 6-lower support, 7-scissor lift, 8-component cable, 8-1-high-end component cable, 8-2-low-end component cable, 9-robotic arm, 10-vacuum suction cup, 11-lifting cylinder, 12-photovoltaic panel body, 13-support leg, 14-tilt cylinder, 15-cable buckle, 16-vision camera. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0061] First, it needs to be clarified that this embodiment refers to a 40m span double-layer cable-stayed flexible photovoltaic support structure, the dimensions of which are as follows: Figure 7 As shown; in this embodiment, the heights of the flexible support triangular braces from left to right are 1326mm, 2000mm, 2000mm, and 1326mm, arranged symmetrically. The length of each photovoltaic panel is 8m, and the photovoltaic modules are evenly distributed inside with a spacing of 0.25m. The component cable 8 and the stabilizing cable of the flexible support are both made of hot-dip galvanized steel strand with a diameter of 12.7mm. The steel strand is a wire made of multiple high-strength steel wires twisted together. In order to prevent corrosion and extend service life, it is wrapped with an outer sheath of polytetrafluoroethylene.

[0062] Before installing photovoltaic modules, the cables of the flexible support system must be initially tensioned to give them initial stiffness. For example... Figure 8 As shown, this illustrates the connection method at the ends of the double-layer cable-stayed support structure. Component cable 8 and stabilizing cable are anchored to the H-beams of the side beams via anchors. The columns support the overall structure, bearing the main vertical loads such as the self-weight of the photovoltaic modules and external loads. The diagonal tie rods effectively improve the system's wind resistance by transmitting force diagonally, sharing some of the tension and lateral loads of the columns, effectively limiting the lateral displacement at the ends of the support structure, and increasing the system's stability. Figure 9 For a pre-tensioned double-layer cable-stayed flexible photovoltaic support structure, due to the action of the triangular struts and stabilizing cables, the initial curvatures of the module cables (8) and stabilizing cables are opposite, with the module cables exhibiting an upward-convex curve. Before installing the photovoltaic modules, because the pre-tension applied to the module cables (8) and stabilizing cables is typically large, the two module cables (8) can be approximated as straight lines within each group. However, due to the structural characteristics of the double-layer cable-stayed flexible photovoltaic support, the deformation of the end nodes of the module cable strands within the 8m installation range of each group varies in both vertical and horizontal directions. During manual installation of the photovoltaic modules, to ensure coordinated stress and deformation of the flexible photovoltaic support, a symmetrical installation method from both ends towards the middle is typically adopted. After all photovoltaic modules are installed and fixed, the sag-to-span ratio of the module cables (8) is required to be controlled within 1 / 200.

[0063] like Figure 1-6 As shown, a flexible support photovoltaic module installation device includes:

[0064] A photovoltaic module clamping system includes a photovoltaic panel clamping mechanism for connecting photovoltaic panels and a moving mechanism for moving the photovoltaic panels onto a flexible support at a set angle. The photovoltaic panel clamping mechanism is used to connect the photovoltaic panels to be installed. Since the flexible support is mostly tilted to allow the photovoltaic panels to receive sunlight better, the photovoltaic panels need to be adjusted to a tilted state to be installed on the flexible support. The moving mechanism is used to adjust the tilt angle of the photovoltaic panels and make the photovoltaic panels fit against the flexible support to meet the installation requirements.

[0065] The flexible support adjustment mechanism is used to adjust the shape of the flexible support according to the installation requirements of the photovoltaic panel. In this embodiment, the flexible support includes two component cables 8. In other embodiments, the flexible support can also be made of other materials. The photovoltaic panel includes a photovoltaic panel body 12. The bottom surface of the photovoltaic panel body 12 is provided with several cable buckles 15 that cooperate with the component cables 8. Before installation, the photovoltaic module needs to connect its aluminum alloy frame to the component bolts above the cable buckles 15. When installing the photovoltaic module, the positioning arc face of the cable buckle 15 is aligned with the component cable 8, and the cable buckle 15 is fixed to the component cable 8 by tightening the fixing bolts, thereby realizing the connection and fixation between the photovoltaic module and the component cable 8. The main function of the flexible support adjustment mechanism is to adjust (adjust) the shape of the flexible support on the front and rear sides so that the shape of the flexible support below the photovoltaic panel clamping mechanism matches the shape of the photovoltaic panel body 12 on the photovoltaic panel clamping mechanism.

[0066] The walking system is used to move the photovoltaic module clamping system and the flexible support adjustment mechanism to the installation position of the photovoltaic panels on the flexible support. Multiple photovoltaic panels can be installed on one flexible support. The function of the walking system is to drive the photovoltaic module clamping system and the flexible support adjustment mechanism to move to the various installation positions on the flexible support until the installation of all photovoltaic panels is completed.

[0067] In one embodiment, the moving mechanism includes a height adjustment component and an angle adjustment component. The height adjustment component includes an upper support 3 connected to the walking system. The upper support 3 is connected to a lower support 6 via a scissor lift mechanism. The lower support 6 is connected to the angle adjustment component, which is connected to the photovoltaic panel clamping mechanism. The scissor lift mechanism includes a scissor lift component 7 and a lifting cylinder 11. The upper and lower ends of the scissor lift component 7 are hinged to the upper support 3 and the lower support 6, respectively. One end of the lifting cylinder 11 is hinged to the upper support 3, and the other end is hinged to the scissor lift component 7. By extending or retracting the lifting cylinder 11, the scissor lift component 7 can be vertically extended or shortened, thereby changing the distance between the upper support 3 and the lower support 6, realizing the upward or downward movement of the photovoltaic panel body 12, and enabling the picking up and putting down of the photovoltaic panel body 12. In this embodiment, the angle adjustment component includes two support legs 13. Each support leg 13 is located at one end of the same side of the lower support 6. One end of each support leg 13 is fixedly connected to the lower support 6, and the lower end is hinged to the photovoltaic panel clamping mechanism. The other end of the photovoltaic panel clamping mechanism is hinged to a tilting cylinder 14, which is also hinged to the lower support 6. The extension and retraction of the tilting cylinder 14 can rotate one side of the photovoltaic panel clamping mechanism downwards or upwards, thereby tilting the photovoltaic panel upwards or downwards and adjusting its tilt angle. In this embodiment, the photovoltaic panel body 12 can be rotated at angles between 0° and 15°. The preload of the cables during installation is high, and the angle formed by the two cables during photovoltaic panel installation deviates little from the design angle. A tilt sensor is arranged on the photovoltaic panel clamping mechanism, which can control the tilting cylinder 14 to adjust the angle when the angle does not match the cable angle.

[0068] In one embodiment, the photovoltaic panel clamping mechanism includes a mounting frame 5 connected to the moving mechanism, such as... Figure 1The mounting frame 5 shown is a rectangular frame structure. In this embodiment, five photovoltaic panels can be connected to the mounting frame 5 simultaneously. The five photovoltaic panels are arranged in a row with equal spacing. The width of the mounting frame 5 is slightly smaller than the length of the photovoltaic panels. The bottom of the mounting frame 5 is equipped with 20 vacuum suction cups 10. Every four vacuum suction cups 10 form a vacuum suction cup group, and one vacuum suction cup group corresponds to one photovoltaic panel. The photovoltaic panels are connected by the combined suction force of the four vacuum suction cups 10. The 40m span double-layer cable-stayed flexible support structure consists of 25 photovoltaic panel bodies 12 per span, with a total of 100 cable buckles 15. If a single module installation and fixing method is used, the process is cumbersome and affects the installation efficiency. Therefore, the automated photovoltaic module installation equipment adopts a method of installing multiple photovoltaic modules simultaneously. First, the buckles 15 are fixed to the photovoltaic panel body 12. Then, the photovoltaic module clamping system clamps five photovoltaic panel bodies 12 at a time. The photovoltaic module clamping system uses vacuum suction cups 10 to clamp and adsorb the photovoltaic panel bodies 12. Each photovoltaic panel body 12 is adsorbed by four vacuum suction cups 10. In order to meet the photovoltaic panel arrangement sequence of the 40m span double-layer cable-stayed flexible support structure, the spacing between every two photovoltaic panel bodies 12 in the photovoltaic panel clamping mechanism is 0.25m. It should be noted that in order to achieve a precise connection between the vacuum suction cups 10 and the photovoltaic panel body 12, a vision camera 16 can be set on the mounting frame 5. The vision camera 16 captures images of the photovoltaic panel body 12, and the position of the four buckles 15 on the photovoltaic panel body 12 is obtained through image analysis. The relative positional relationship between the photovoltaic panel body 12 and the vacuum suction cups 10 is determined based on the buckles 15, thereby ensuring that the position of the photovoltaic panel body 12 is relatively fixed with the installation position of the robotic arm 9, so that the final adjusted shape of the flexible support meets the installation requirements of the photovoltaic panel body 12.

[0069] In one embodiment, such as Figure 1 As shown, the flexible support adjustment mechanism includes four robotic arms 9 mounted on the photovoltaic panel clamping mechanism. Specifically, the robotic arms 9 are electrically driven articulated robotic arms, which have the characteristics of high control precision, fast response speed and large driving force. They can rotate and bend, and their openable and closable gripping heads can clamp the component cable 8 and drive the component cable 8 to move, thereby realizing cable pulling. The four robotic arms 9 are respectively set at the four corners of the front and rear ends of the photovoltaic panel clamping mechanism, which can realize the gripping of the four control nodes of the high and low end component cables, thereby adjusting the shape of the component cable 8 directly below the photovoltaic module clamping system to meet the installation requirements.

[0070] In this embodiment, the flexible support adjustment mechanism further includes a vision camera 16. In this embodiment, the vision camera 16 is a RealSense D435i camera, which is equipped with a color RGB camera, two infrared cameras, and an infrared dot matrix projector. The RealSense D435i camera has an overall size of 90×25×25mm and an effective detection distance of 0.15~10m. After the photovoltaic module clamping system accurately moves the photovoltaic module to the corresponding installation position, a depth camera is used to detect the spatial position of the module cable 8. Subsequently, a robotic arm 9 is used to adjust the module cable 8 to ensure its alignment meets the requirements for simultaneous installation of multiple photovoltaic modules. It should be noted that the vision camera 16 adopts an "eye-to-hand" installation method (e.g., ...). Figure 10 As shown in the figure, the relative positions of the robotic arm 9 and the vision camera 16 are fixed in order to match the position information of the robotic arm and the gripping point. The vision camera 16 is installed at the four corners of the photovoltaic panel gripping mechanism, and its field of view can cover the component cable 8 in the current installation area. The main function of the vision camera 16 is to guide the robotic arm 9 to grip and accurately detect the line shape of the component cable 8. Through image processing and coordinate transformation of the host computer, the position of the node of the component cable 8 in the world coordinate system can be obtained, realizing the gripping and line shape detection of the steel strand of the component cable 8. When its line shape meets the installation technical requirements of the automated installation equipment, the photovoltaic panel gripping mechanism is lowered by the height adjustment component to complete the positioning and installation of 5 photovoltaic panels, 20 cable buckles 15 and component cable 8.

[0071] In one embodiment, the walking system includes a horizontal moving frame 1 and a vertical moving frame 2 (the horizontal moving frame 1 is the track of a gantry-type traveling mechanism, and the vertical moving frame 2 has wheels on both sides, allowing it to move on the horizontal moving frame 1). The horizontal moving frame 1 is perpendicular to the flexible support, and the vertical moving frame 2 is located directly above the flexible support and extends along the length of the flexible support. The vertical moving frame 2 is perpendicular to the horizontal moving frame and can move along the horizontal moving frame 1. A moving trolley 4 is mounted on the vertical moving frame 2, and the photovoltaic module clamping system is connected to the moving trolley 4. The horizontal moving frame 1 allows the photovoltaic module clamping system to move between the various groups of module cables 8, while the vertical moving frame allows the photovoltaic module clamping system to move to different positions on the module cables 8.

[0072] In one embodiment, a slide rail is installed on each of the two long sides of the mounting bracket 5. A sliding table is mounted on each slide rail, and the table can slide left and right on the rail. The drive mechanism can be a rack and pinion drive. A robotic arm is mounted on the sliding table, with an electric screwdriver and a vision camera 16 at its end. The vision camera 16 is installed in an "eye-on-hand" manner, meaning the relative positions of the electric screwdriver and the vision camera 16 remain unchanged. After the photovoltaic panel is installed and placed in place, the vision camera 16 on the robotic arm identifies and positions the fastening bolts of the cable clips 15. Once positioned, the bolts are tightened using the electric screwdriver. After one cable clip 15 is tightened, the sliding table is moved to control the robotic arm to the next cable clip 15 position for tightening.

[0073] A method for installing flexible bracket photovoltaic modules, using the aforementioned flexible bracket photovoltaic module installation device, includes the following steps:

[0074] S1: The photovoltaic module clamping system and flexible support adjustment mechanism move on the horizontal moving frame 1 and the vertical moving frame 2 to reach the storage position of the photovoltaic panel. Then, the lifting cylinder 11 is controlled to extend, driving the vacuum suction cup 10 to move downward and contact the photovoltaic panel below. Subsequently, the photovoltaic panel is connected by the suction force of the vacuum suction cup 10. It should be noted that the mounting frame 5 and the photovoltaic panel are in a horizontal state at this time. In this embodiment, since the mounting frame 5 can accommodate 5 photovoltaic panels at the same time, the photovoltaic panels can be stacked in five stacks. When stacking, the distance between the stacks is equal to the distance between the two adjacent sets of vacuum suction cups on the mounting frame, so that the photovoltaic module clamping system can connect five photovoltaic panels at the same time, further improving the installation efficiency.

[0075] S2: The photovoltaic module clamping system and the flexible support adjustment mechanism move on the horizontal moving frame 1 and the vertical moving frame 2 to the installation position of the photovoltaic panel on the flexible support;

[0076] S3: The flexible support adjustment mechanism adjusts the shape of the flexible support according to the installation requirements of the photovoltaic panel. After the photovoltaic module clamping system accurately moves the photovoltaic module to the corresponding installation position, the spatial position of the module cable 8 is detected by a depth camera. Then, the robotic arm is used to adjust the cable of the module cable 8 so that its shape meets the shape of multiple photovoltaic modules installed at the same time. Specifically, the clamping head at the end of the robotic arm 9 can clamp the module cable 8 and drive the module cable 8 to move, thereby realizing the cable adjustment and adjusting the shape of the module cable 8 directly below the photovoltaic module clamping system to meet the installation requirements.

[0077] S4: The moving mechanism of the photovoltaic module clamping system adjusts the height and angle of the photovoltaic panel according to the shape of the flexible support, so that the photovoltaic panel contacts the flexible support. Specifically, the extension and retraction of the tilt cylinder 14 can drive one side of the photovoltaic panel clamping mechanism to rotate downward or upward, thereby tilting the photovoltaic panel on the photovoltaic panel clamping mechanism upward or downward, realizing the adjustment of the tilt angle of the photovoltaic panel. Then, the extension of the lifting cylinder 11 can drive the scissor-type lifting component 7 to extend vertically, thereby changing the distance between the upper support 3 and the lower support 6, realizing the downward movement of the photovoltaic panel until the cable buckle 15 under the photovoltaic panel contacts the module cable 8 (the module cable 8 is locked in the groove of the cable buckle 15). Then, the vacuum suction cup 10 stops suction, the extension and retraction of the lifting cylinder 11 shorten, realizing the separation of the photovoltaic panel from the photovoltaic module clamping system. It should be noted that before installing the photovoltaic module, the module cable 8 and the stabilizing cable need to be pre-tensioned, and the initial shape of the module cable 8 is an upward arched state. To ensure the coordination of stress and deformation of the flexible support structure, the third group of positions is installed first. After installation, the adjacent second and fourth groups are installed, and finally the first and fifth groups of photovoltaic modules are installed.

[0078] S5: Tighten the knob of buckle 15 to complete the connection between the photovoltaic panel and the flexible support.

[0079] Because cable structures exhibit nonlinear stress and deformation under concentrated loads, the geometric nonlinearity of the structure must be considered when adjusting the alignment of cable 8. To investigate the required cable-pulling force for alignment adjustment of the flexible photovoltaic support under different installation conditions, the specific process is as follows:

[0080] The materials used in the flexible support are shown in Table 1: Table 1 - Material Properties of Flexible Scaffolds

[0081]

[0082] In this embodiment, a finite element analysis model is established based on a 40m span double-layer cable-stayed flexible photovoltaic support structure, with the photovoltaic modules tilted at 15°. The module cables 8 (including high-end module cable 8-1 and low-end module cable 8-2) and stabilizing cables in the double-layer cable-stayed flexible photovoltaic support are all made of 12.7mm diameter hot-dip galvanized steel strand with a yield strength of 1860MPa. The side beams are welded composite H-beams with cross-sectional dimensions of 400×280×8×12mm. The diagonal tie rods at both ends are 30mm diameter cylindrical tie rods made of Q345 steel. The side columns are made of C30 concrete. The triangular struts are hollow cylindrical tubes with a diameter of 32mm and a wall thickness of 2mm.

[0083] Both the component cable 8 and the stabilizing cable are simulated using frame elements, with a maximum segmentation length of 400mm. Each strut of the triangular support is a two-force member, therefore, end release is required, selecting the release of bending moments at both ends of the primary and secondary axes and torque at either end. When setting boundary conditions, the endpoints of the component cable and the stabilizing cable are connected using weld constraints. When the system detects that the spatial distance between adjacent nodes is less than a preset threshold, a constraint element will be created for these nodes. In this embodiment, the weld tolerance is set to 1.5m, and the degrees of freedom of the constraint are translation and rotation in the x, y, and z directions. Fixed support constraints are added to the columns and tie rods to the ground. The load generated by the photovoltaic module is transferred to the component cable 8 through virtual surfaces.

[0084] As tension members, cables must have a certain initial prestress to bear the self-weight of photovoltaic modules and external loads. In SAP2000, when modeling cable elements using frame elements, there are two methods for applying prestress to cables: the equivalent cable force application method and the target force iteration method.

[0085] (1) Method of applying equivalent cable force

[0086] The methods for applying equivalent cable force include the initial strain method and the cooling method. The basic principle of the initial strain method is based on the theory of elastic deformation of materials. When the cable is in an elastic working state, the strain of the cable can be obtained according to the relationship between stress and strain:

[0087] (1)

[0088] In the formula: E is the elastic modulus; A is the cross-sectional area; F is the equivalent cable force to be applied.

[0089] The essence of the cooling method is to cool the cable. Based on the principle of thermal expansion and contraction, the cable is shortened to apply an equivalent cable force. The amount of cooling of the cable can be calculated as follows:

[0090] (2)

[0091] In the formula: The temperature that was lowered; is the coefficient of thermal expansion of the cable.

[0092] When applying pretension to cables using the initial strain method and the cooling method, the essence is to change the stress-free cable length of the cable to be smaller than the actual modeled length of the cable, thereby generating pretension in the cable.

[0093] (2) Target force iteration method

[0094] For cable-stayed structures, the cable force in each segment is different under real-world conditions, with a significant difference between the cable force at mid-span and the forces at both ends. Therefore, a convergence tolerance for the target force must be set, and the cable force in each segment is limited to:

[0095] (3)

[0096] In the formula: P is the actual cable force value of each cable segment; F is the set target force value. The target force convergence tolerance.

[0097] When adjusting the cable alignment of a double-layer cable-stayed flexible photovoltaic (PV) support module, it is necessary to calculate the difference between the actual cable alignment of the module to be installed and the target alignment. This allows for the calculation of the deflection adjustment amount at each node. Based on the deflection adjustment amount at each node, the required cable-pulling force at each node is calculated to meet the conditions for simultaneous installation of multiple PV modules. Since the magnitude of the concentrated force on the cable and the vertical deformation of each node have a non-linear relationship, this embodiment investigates the magnitude of the cable-pulling force applied to the node based on the influence matrix method.

[0098] Define the modulated vector Define the modulated vector for the cable pulling force applied to the component cable. The displacement of the cross section of interest in the component cable.

[0099] When the magnitude of a certain cable pulling force A unit change occurred, displacement vector And so it changes, causing the amount of change to change. If the magnitude of each cable-pulling force changes, summing these vectors can form an m×n matrix. Matrix of Change This is called the influence matrix:

[0100] (4)

[0101] The above vectors and matrices can be related using the principle of linear superposition, satisfying the following equation:

[0102] (5)

[0103] Expand as

[0104] (6)

[0105] The dead load and initial cable force of the double-layer cable-stayed flexible photovoltaic support are defined as the initial load. The stiffness of the flexible support under the initial load is called the initial stiffness. Based on the initial stiffness of the flexible support structure, an influence matrix for the double-layer cable-stayed flexible photovoltaic support is formed. The specific process is as follows:

[0106] (1) Assume the initial cable forces of the component cable and the stabilizing cable are P1 and P2, and the initial value of the cable pulling force is P1. Calculate the displacement of the component cable key points under the action of initial cable force, initial cable pulling force, and dead load. The key point is to select the location of the connection node between the triangular brace and the component cable;

[0107] (2) Calculate the initial load and the first load. After a unit change in the cable pulling force, the displacement of the key point ;

[0108] (3) After considering the initial stiffness of the double-layer cable-stayed flexible photovoltaic support, the flexible support is basically approximately a linear structure. According to the superposition principle, through The displacement change of the key point is obtained after the i-th unit change in cable pulling force;

[0109] (4) Repeat the above process to obtain the influence matrix. .

[0110] After determining the optimization objective, i.e., determining the expected displacement vector of the key points. Then, by solving the equations Then the cable pulling force vector can be obtained as:

[0111] (7)

[0112] In the formula: The amount that needs to be adjusted for the key points of the cable.

[0113] Since the cables and stabilizing cables of the double-layer cable-supported flexible photovoltaic module are nonlinear structures, iterative approximation of the optimization objective is required. Therefore, the calculation of the cable pulling force is an iterative process, and the specific process is as follows:

[0114] (1) Based on the formation of the influence matrix, the cable pulling force vector Updated to ;

[0115] (2) Substitute the values ​​into the SAP2000 finite element model to calculate the component cable under pretension, dead load, and cable pull force. Displacement vector of key points in cable-stayed structure under combined action ;

[0116] (3) Compare the expected displacement vectors and The error, if If so, proceed to the next step;

[0117] (4) Vector the cable pulling force Updated to ;

[0118] (5) Substitute the components into the finite element model to calculate the cable components under pretension, dead load, and cable pull force. Displacement vector of key points in cable-stayed structure under combined action ;

[0119] (6) Repeat steps (2) to (5) until To obtain the final ideal cable pulling force , where j is the number of iterations.

[0120] The shape of the component cable corresponding to the final cable pulling force is the desired control line shape;

[0121] From equation (7), the following system of linear equations can be established:

[0122] (8)

[0123] In the formula: This is the desired adjustment amount for the cable pulling force.

[0124] In equation (8), the adjustment amount of the key point displacement to be controlled is... The number of elements m is greater than the cable pulling force adjustment amount. The number of elements n makes it a system of contradictory equations, and its generalized solution can be solved. The principle is to find... ,make:

[0125] (9)

[0126] To minimize. According to the principle of extrema, it satisfies smallest must:

[0127] (10)

[0128] or:

[0129] (11)

[0130] The above formula can be transformed into matrix form as follows:

[0131] (12)

[0132] The above equation is a system of linear equations with n unknowns and n equations. When When the rank is full, the system of equations has a unique solution. Solving from the above equation yields... Then, the adjusted keypoint displacement values ​​can be obtained:

[0133] (13)

[0134] In the formula: These are the values ​​before the keypoint displacement adjustment.

[0135] In equation (8), a positive definite diagonal matrix is ​​added. Then the system of equations becomes:

[0136] (14)

[0137] Solving the above equation using the least squares method, we get:

[0138] (15)

[0139] Equation (15) is the weighted least squares solution of equation (8). The weighted matrix consists of elements on its diagonal that are the weighting coefficients of the corresponding equations in equation (8). During parameter optimization, control targets with larger weights will receive higher priority, and their actual output values ​​will approach the preset theoretical values ​​more quickly. Therefore, the shape of the cable after cable pulling can be changed by adjusting the weighting coefficients. When a specific control index is found to exceed the allowable threshold after iterative calculation, the weighting parameter corresponding to that index can be strengthened in a targeted manner.

[0140] like Figure 11 As shown, the naming convention for the four nodes of a group of photovoltaic panel modules is "letter + serial number". For example: A1~A4, where A represents the first group of photovoltaic panels, and 1~4 represent different nodes. The nodes for the second to fifth groups of photovoltaic panels are: B1~B4, C1~C4, D1~D4, E1~E4, where nodes A1, A3, E2, and E4 are connected to the side beams of the flexible support. The initial prestress of the module cable is set to 40000N, and the initial prestress of the stabilizing cable is set to 5000N. Prestress is applied using a target force method. After tensioning, as shown... Figure 12 As shown.

[0141] Due to the large initial prestress applied to the component cables and stabilizing cable strands, the component cables are approximately straight within each group of adjacent nodes. The initial vertical deformation of each node of the component cables is shown in Table 2. The flexible support structure is symmetrical, therefore, the deformation data of the first to third groups are listed. Since nodes A1 and A3 are connected to the side beams, their vertical deformation is relatively small, at 0.2 mm.

[0142] Table 2 - Initial Vertical Deformation of Nodes

[0143]

[0144] (1) Installation of the third group of photovoltaic panels

[0145] Since the two triangular supports in the third group are at the same height of 2m, the vertical displacement difference between the nodes at both ends of the third group's component cables after pretensioning does not exceed 3mm, and they are on the same height plane. Considering the coordination between the stress and deformation of the double-layer cable-supported flexible photovoltaic system, a symmetrical installation method from the middle to both ends is adopted, installing the third group of photovoltaic panels first. The cable alignment of the third group meets the alignment requirements for simultaneous installation of multiple photovoltaic panels, and can be directly installed.

[0146] (2) Installation of the second set of photovoltaic panels

[0147] After the third set of photovoltaic panels was installed, the deformation of each node of the second set of modules is shown in Table 3:

[0148] Table 3 - Vertical Deformation of the Second Group of Nodes

[0149]

[0150] After the third set of photovoltaic panels is installed, a downward uniformly distributed load will be generated on the double-layer cable-stayed flexible support structure, reducing the upward deflection of the module cables. Table 3 shows that at the second set location, the vertical deformation difference between the two ends of the high-end module cable is 9.6 mm, and the difference between the two ends of the low-end module cable is 9.3 mm. Therefore, the alignment of the module cables needs to be adjusted to ensure that the vertical deformation of the four nodes is equal.

[0151] The linear shape of the second group of component cables was optimized using the influence matrix method. The initial optimization target was set at 45mm for the vertical displacement deformation of the four nodes. The unknown quantity is the concentrated force at the four nodes. The preliminary estimate of the concentrated force at nodes B1 and B3 is -10N, and the concentrated force at nodes B2 and B4 is 10N. The influence matrix of the structure is obtained as follows:

[0152] (16)

[0153] After three rounds of iteration The concentrated force of the four nodes is obtained. =(-100.117, 52.935, -102.188, 47.486) T The iterative convergence process is as follows: Figure 13 As shown, from Figure 16 As can be seen from the second iteration, the calculation basically converged.

[0154] The concentrated forces at the nodes were calculated using the load optimizer in SAP2000. =(-99.89, 53.091, -102.021, 47.685)T. By comparing the theoretical calculation results, the maximum error of the concentrated force between the two is no more than 0.5%, thus proving the correctness of the calculation process.

[0155] The required cable pulling force for each node when different displacement deformations are set is shown in Table 4:

[0156] Table 4 - Relationship between nodal displacements and concentrated forces in the second group

[0157]

[0158] Due to the geometric nonlinearity of the double-layer cable-stayed flexible photovoltaic support structure, the concentrated forces acting on nodes B1 to B4 exhibit coupling relationships. For example... Figure 14 In the figure, 'a' represents the curve showing the relationship between the displacement of the second set of nodes and the concentrated force. Due to the tension error between the high-end and low-end component cables, the initial displacements of nodes B1 and B3, and B2 and B4 are not equal. Therefore, when the displacements of these two pairs of nodes are the same, the concentrated forces acting on nodes B1 and B3, and nodes B2 and B4 are not equal, and there is a certain deviation.

[0159] like Figure 14 As shown in b, the graph illustrates the relationship between nodal displacement and the absolute value of concentrated forces. When the nodal displacement is within 0-30 mm, the directions of concentrated forces B1 and B3, and B2 and B4 are all downward, with B1 and B3 being greater than B2 and B4. When the nodal displacement is within 60-100 mm, the directions of concentrated forces B1 and B3, and B2 and B4 are all upward, with B2 and B4 being greater than B1 and B3. When the nodal displacement is 45-50 mm, the curves for nodes B1 and B3 intersect with the curves for nodes B2 and B4. At this point, the peak value of the absolute value of the cable-pulling force acting on the four nodes is minimized, defined as the optimal cable-pulling force, and the corresponding nodal displacement is the ideal displacement. It can be seen that the magnitude of the ideal displacement always lies between the minimum and maximum vertical deformation of the four nodes in Table 3.

[0160] The ideal displacement can be obtained from the relationship curves between the displacement and the absolute value of the concentrated force at nodes B1 and B2 using linear interpolation. The ideal displacement, determined by linear interpolation, is 48 mm. Substituting this node displacement into SAP2000, the corresponding optimal cable pulling force is obtained. =(-74.9, 76.3, -77, 70.9) T .

[0161] (3) Installation of the fourth group of photovoltaic panels

[0162] After the second and third sets of photovoltaic panels were installed, the vertical deformation of each node of the fourth set of components is shown in Table 5.

[0163] Table 5 Vertical Deformation of Nodes in Group 4

[0164]

[0165] After the second and third groups of photovoltaic panels are installed, the double-layer cable-stayed flexible support structure will generate an asymmetrical force, resulting in a significant difference in the fourth group of nodes. Specifically, the difference between the highest and lowest points of the high-end module cables is 25.47 mm, and the difference between the highest and lowest points of the low-end module cables is 25.14 mm. Table 6 shows the magnitude of the cable pulling force required for each node when different displacements are applied to the fourth group of nodes.

[0166] Table 6 - Correspondence between nodal displacements and concentrated forces in the fourth group

[0167]

[0168] Figure 14 As shown in Figure c, when the node displacement is 30~60mm, the concentrated force acting on nodes D1 and D3 is upward, and the concentrated force acting on nodes D2 and D4 is downward. The concentrated force acting on the four nodes can adjust the cable shape of the fourth group of components to meet the requirements of simultaneous installation of multiple photovoltaic panels.

[0169] like Figure 14 As shown in 'd', this is the curve showing the relationship between the fourth set of nodal displacements and the absolute value of the concentrated force. When the nodal displacement is 45~50mm, the curves for nodes D1 and D3 intersect with the curves for nodes D2 and D4. Using linear interpolation, the ideal displacement of 48mm is finally obtained from the curves showing the relationship between the displacement and the absolute value of the concentrated force of nodes D1 and D2. Substituting this nodal displacement into SAP2000, the corresponding optimal cable pulling force is obtained. =(210.3, -205.3, 204.6, -207.2) T .

[0170] (4) Installation of the first group of photovoltaic panels

[0171] After the second, third, and fourth groups of photovoltaic panels were installed, the vertical deformation of each node of the first group of modules is shown in Table 7:

[0172] Table 7 - Vertical Deformation of the First Group of Nodes

[0173]

[0174] Since nodes A1 and A3 are connected to the left-side beam of the flexible support, their vertical deformation is very small. Setting the target displacement of nodes A2 and A4 to 0, the magnitude of the concentrated force applied to nodes A2 and A4 is obtained. =(-120.1, -128.3)T.

[0175] (5) Installation of the fifth group of photovoltaic panels

[0176] After the first set of photovoltaic panels was installed, the vertical deformation of the fifth set of module cable nodes is shown in Table 8:

[0177] Table 8 - Vertical Deformation of Nodes in Group 5

[0178]

[0179] Since nodes E2 and E4 are connected to the right-side beam of the flexible support, their vertical deformation is very small. Setting the target displacements of nodes E1 and E3 to 0, the magnitudes of the concentrated forces applied to nodes E1 and E3 are obtained. =(-184.3, -191.9)T.

[0180] Based on the above analysis, the ideal displacement and optimal cable pulling force of each node of the module cable during the installation of each group of photovoltaic panels are shown in Table 9. Among them, the fourth group has a larger cable pulling force, with the largest cable pulling force at node D1, which is 210.3 N.

[0181] Table 9 - Ideal Displacement and Optimal Cable Pulling Force for Photovoltaic Panel Installation

[0182]

[0183] As the core of the double-layer cable-stayed flexible photovoltaic support system, the deflection and strength of the cables need to be checked after all the photovoltaic panels are installed to meet the installation requirements.

[0184] (1) Cable deflection check

[0185] After all photovoltaic panels were installed, the vertical displacement cloud map of the double-layer cable-stayed flexible photovoltaic support system is shown below. Figure 15 As shown, the maximum deflection of the component cable is 50.5 mm, and the sag-to-span ratio is less than 1 / 200, which meets the deformation requirements of the flexible support structure.

[0186] (2) Cable strength check

[0187] According to the "Technical Specification for Cable Structures", the load-bearing capacity of cables should meet the following formula requirements:

[0188] (17)

[0189] (18)

[0190] In the formula: F is the design value of the tensile strength of the cable; This is the standard value of the ultimate tensile strength of the cable; The resistance factor for the cable is 2.0; This is the design value for the maximum axial tensile force that the cable can withstand; The importance coefficient of the structure is 1.0, as specified in the "Code for Design of Photovoltaic Support Structures".

[0191] According to the finite element simulation results, the maximum axial force of the component cable is 38969N and the maximum axial force of the stabilizing cable is 22064N, both of which are less than the cable bearing capacity of 118KN, thus meeting the strength verification requirements.

[0192] Since the double-layer cable-stayed photovoltaic support system is a flexible structure, it is significantly affected by wind loads. Therefore, it is necessary to investigate the impact of wind loads on the cable-pulling force applied during the adjustment of the module cable alignment. The formula for calculating static wind load is as follows:

[0193] (19)

[0194] In the formula: This is the standard value for wind load; The wind vibration coefficient is taken as 1.5; It is the wind pressure shape coefficient, which is determined when the photovoltaic module is tilted. ≤15° Take 0.8; The wind pressure height variation coefficient is taken as 1.0; This is the basic wind pressure.

[0195] like Figure 16 The figure shows the relationship curves between the absolute values ​​of nodal displacement and concentrated force under different wind pressures. For nodes B1 and B3, when the displacement is between 20mm and 55mm, the concentrated force acts downwards. The larger the nodal displacement, the smaller the concentrated force acting on the node. The larger the basic wind pressure, the larger the concentrated force acting on the node. When the nodal displacement is 60mm, the concentrated force acts upwards. As the basic wind pressure increases, the concentrated force acting on the node decreases. For nodes B2 and B4, when the displacement is between 35mm and 60mm, the concentrated force acts upwards. The larger the nodal displacement, the larger the concentrated force acting on the node. As the basic wind pressure increases, the concentrated force acting on the node also increases. When the nodal displacement is 20mm, the concentrated force acts downwards. As the basic wind pressure increases, the concentrated force acting on the node decreases.

[0196] like Figure 17The figure shows the relationship between the displacement of the cable nodes and the absolute value of the concentrated force under different wind pressures. For nodes D1 and D3, when the displacement is 20mm~70mm, the concentrated force acts upward. As the node displacement increases, the concentrated force acting on the node also increases. As the basic wind pressure increases, the concentrated force acting on the node also increases. For nodes D2 and D4, when the displacement is 10mm~70mm, the concentrated force acts downward. As the node displacement increases, the concentrated force acting on the node decreases. As the basic wind pressure increases, the concentrated force acting on the node increases.

[0197] like Figure 18 The figure shows the relationship between the concentrated force acting on nodes A2 and A4 (first group) and the basic wind pressure (fifth group) when the target displacement is set to 0 mm. The basic wind pressure is 10~50 N / m. 2 At this time, the concentrated force at nodes A4 and E3 is directed downwards. As the basic wind pressure increases, the concentrated force acting on the nodes decreases. The basic wind pressure is 10~30 N / m. 2 At this time, the concentrated force acting on node A2 is downward, and its magnitude decreases as the basic wind pressure increases. The basic wind pressure is 40~50 N / m. 2 At this time, the concentrated force acting on node A2 is upward, and its magnitude increases with the increase of the basic wind pressure. The basic wind pressure is 10~40 N / m. 2 At that time, the concentrated force acting on node E1 is downward, and its magnitude decreases as the basic wind pressure increases. The basic wind pressure is 50 N / m. 2 At that time, the concentrated force acting on node E1 is upward.

[0198] like Figure 19 As shown, the curves represent the relationship between the ideal displacement of the nodes in the second and fourth groups and the basic wind pressure. As the basic wind pressure increases, the ideal displacement decreases downward.

[0199] like Figure 20 The figure shows the relationship between the optimal cable-pulling force and the basic wind pressure for the second and fourth groups of nodes. The optimal cable-pulling force for nodes B1, B3 and D2, D4 is downward, while that for nodes B2, B4 and D1, D3 is upward. The greater the basic wind pressure, the greater the optimal cable-pulling force acting on the node.

[0200] To investigate the required concentrated force at the nodes during component cable alignment adjustment under different installation conditions, this invention establishes a finite element model of a double-layer cable-supported flexible photovoltaic system and determines the method of applying pretension to the cables. Based on the geometric nonlinear characteristics of the cable structure, the influence matrix method is used to iteratively solve for the cable pulling force. The relative error between the solution and the load optimizer calculation results does not exceed 0.5%. Then, the load optimizer simulation yields the relationship curve between the component cable node displacement and the magnitude of the cable pulling force under normal installation conditions. With the goal of minimizing the concentrated force applied to the component cable nodes, the optimal cable pulling force and ideal displacement of the nodes are obtained through linear interpolation. Verification of the cables in the flexible support structure shows that their stress and deformation do not exceed the specified values, meeting the installation requirements.

[0201] Precise positioning of photovoltaic modules is a crucial step in the entire installation process. After obtaining the optimal cable pulling force and ideal displacement of the module cable nodes, the alignment of the flexible support module cables needs to be detected and controlled. This implementation first derives the transformation matrix between the camera pixel coordinate system and the world coordinate system using a pinhole imaging model of the camera. Then, the Zhang Zhengyou calibration method is used to calibrate the camera, obtaining the parameter information of the RGB camera and the left and right infrared cameras, and stereo correction is performed on the image. Next, image segmentation is performed based on the color characteristics of the steel strands in the flexible support. Morphological operations are used to smooth the steel strand images, and Canny edge detection is performed on the processed images. By setting a threshold for the aspect ratio of the minimum bounding rectangle of the image, precise identification and positioning of the steel strands are achieved.

[0202] (1) Basic principles of visual positioning

[0203] Camera image acquisition is an optical imaging process that converts three-dimensional information in the world coordinate system into two-dimensional information in a plane. It involves the transformation between multiple coordinate systems: points in the world coordinate system are first transformed into the camera coordinate system through rigid body transformation. Then, using the principle of optical imaging, points in the camera coordinate system are transformed into the physical coordinate system through projection transformation. Finally, points in the physical coordinate system can be mapped from three-dimensional space to the two-dimensional pixel coordinate system through a second transformation.

[0204] like Figure 21 The diagram shows a pinhole camera model, where D is the target object plane, i.e., the spatial position of the object being measured, and the coordinate system is... C is the world coordinate system, representing the three-dimensional coordinates of an object in the real world; it is generally replaced by the robot arm's base coordinate system. C is the camera plane, and the coordinate system is... Let B be the camera coordinate system, with its Z-axis coinciding with the camera's optical axis; B is the camera's pixel plane. Light reflected from the target object in the world coordinate system is projected onto the pixel plane after passing through the pinhole. The distance between the pixel plane and the pinhole is the focal length f, in mm. This is the image physical coordinate system, with its origin at the intersection of the camera's optical axis and the imaging plane. Let E be the camera's pixel coordinate system; let E be the virtual plane. Assume the coordinates of point P in the world coordinate system are... The coordinates of the imaging point in the pixel coordinate system are .

[0205] Transforming a target point from the world coordinate system to the camera coordinate system requires rotation and translation operations. The rotation transformation matrix is:

[0206] (20)

[0207] Points in the world coordinate system Points in the camera coordinate system The conversion relationship between them is:

[0208] (twenty one)

[0209] In the formula: A point in the world coordinate system; Let R be a point in the camera coordinate system; R represents the rotation matrix; and T represents the translation matrix.

[0210] The transformation formula between the camera coordinate system and the physical coordinate system is derived from the optical imaging principle of the camera:

[0211] (twenty two)

[0212] Transform the above equation into matrix form:

[0213] (twenty three)

[0214] The image pixel coordinate system and the image physical coordinate system are on the same plane, such as... Figure 22 As shown, the u and v axes of the pixel coordinate system are parallel to the x and y axes of the image physical coordinate system, respectively, and the origin of the image physical coordinate system is offset from the origin of the pixel coordinate system. .

[0215] The formula for converting points in the physical coordinate system of an image to points in the pixel coordinate system can be written as:

[0216] (twenty four)

[0217] In the formula: dx represents the pixel scale in the u direction; dy represents the pixel scale in the v direction.

[0218] Equation (24) can be expressed in matrix form as follows:

[0219] (25)

[0220] Ultimately, the transformation formula from world coordinates to pixel coordinates can be derived as follows:

[0221] (26)

[0222] make , , , Then equation (26) can be written as:

[0223] (27)

[0224] In the formula: M1 is the camera intrinsic parameter matrix; M2 is the camera extrinsic parameter matrix.

[0225] Binocular vision systems mimic the mechanism by which the human eye observes objects. By using cameras located at two different positions to observe a target object, the distance from the target point to the camera plane can be calculated based on the principle of triangulation. For example... Figure 23 As shown in a. O l and O r These are the optical centers of the left and right cameras, respectively. P is the target object to be measured in space, and its distance from the center point O of the left and right cameras is [missing information]. l O r The formed plane This is called the polar plane. The lines where this plane intersects the left and right imaging planes are called the epipolar lines, P. l and P r These are the projection points of the target object on the left and right imaging planes, respectively. To narrow down the matching range, P... l and P r The epipolar constraint should be satisfied, that is, the projection point P on the right imaging plane should be satisfied. r It must be at the projection point P relative to the left imaging plane l On the extreme line.

[0226] Figure 23 In the diagram, b represents the distance between the optical centers of the left and right cameras, and x represents the baseline. l x r These are the X-axis coordinates of the left and right camera projection points in the camera coordinate system. Because the left and right camera projection points are located on the same plane, the ordinates of the two points are the same. l and x r Parallax exists Let f be the focal length of the camera, and z be the distance to be measured from the target point P. Based on the principle of triangle similarity, Similar to We can obtain:

[0227] (28)

[0228] because Substituting this into equation (28), we get:

[0229] (29)

[0230] The depth z can be calculated from the above formula after simplification:

[0231] (30)

[0232] The derivation of the formula for transforming from the world coordinate system to the pixel coordinate system shows that obtaining the camera's intrinsic and extrinsic parameters is essential for converting two-dimensional data in the image coordinate system to three-dimensional data in the world coordinate system. The D435i camera has left and right infrared cameras and an RGB camera; the parameters of the same camera can vary due to manufacturing processes or prolonged use. To enable the vision camera to accurately detect the positioning points of the steel strand and thus precisely control its alignment, the D435i camera needs to be calibrated.

[0233] (2) Visual camera calibration method

[0234] In this embodiment, the Zhang Zhengyou calibration method is used to calibrate the vision camera. A 9×12 black and white checkerboard calibration board is selected, with each square measuring 20mm×20mm. The left and right infrared cameras and the RGB camera of the D435i camera are calibrated. Multiple photos of the calibration board are taken by the left and right infrared cameras and the RGB camera at different positions and angles. Finally, the photos are filtered, and 20 pairs of photos of the calibration board at different positions are selected. It is important to note that the infrared dot projector must be turned off when the infrared camera is acquiring images; otherwise, the captured images will exhibit uniformly distributed speckle.

[0235] Twenty pairs of infrared camera images and RGB color images were imported into the CameraCalibration toolbox in MATLAB for camera calibration of the D435i camera. Corner points on the calibration board were detected and extracted, and feature points in the calibration board coordinate system were projected onto the camera coordinate system using the calibration results. The reprojection error was obtained by comparing the pixel difference between the projected points and the real points in the camera coordinate system. Figure 24 As shown. The average error of the left and right infrared cameras is 0.12 pixels, and the average error of the color camera is 0.19 pixels, which is within the allowable error range, therefore the camera parameters are usable. In the same coordinate system, the relationship between the D435i camera and the calibration board is as follows: Figure 25 As shown.

[0236] The calibration results for the D435i infrared camera and RGB color camera can be obtained by exporting the MATLAB calibration results, as shown in Tables 10 and 11.

[0237] Table 10 - Calibration Results of Infrared Cameras for D435i (Right and Left)

[0238]

[0239] Table 11 - D435i Color Camera Parameter Calibration Results

[0240]

[0241] The calibration experiments conducted on the D435i camera in the previous section yielded parameter calibration results for the left and right infrared cameras. It was found that the left and right infrared cameras are not ideal imaging systems, exhibiting rotation matrix R and translation matrix T. Therefore, stereo correction is necessary. The purpose of image correction is to correct imaging errors caused by camera mounting. First, distortion is removed from the image using distortion coefficients; then, epipolar correction is performed to ensure that corresponding pixels of the left and right cameras are on the same horizontal line. The principle is as follows... Figure 26 As shown.

[0242] In this embodiment, the Bouguet stereo correction algorithm is selected to correct the image. This algorithm decomposes the obtained translation and rotation matrices to obtain the minimum reprojection error and maximum overlap area between the two cameras. The image after stereo correction is as follows: Figure 27 As shown, the corresponding pixels in the left and right images are all on the same horizontal straight line, achieving row alignment and ensuring the accuracy of subsequent steel strand identification and positioning.

[0243] (3) Target detection and localization

[0244] Since the prestress applied to the component cables of flexible photovoltaic (PV) brackets before the PV panels are installed is usually quite large, the linearity of the steel strands in the component cables can be detected by measuring the three-dimensional coordinates of the nodes at both ends of the component cables. To accurately detect the spatial pose of the component cables, the RGB image needs to be processed using a target detection algorithm to extract the positioning point information of the steel strands.

[0245] Because the flexible photovoltaic support system uses steel cables wrapped in a blue outer sheath, its color is clearly distinguishable from other components. Therefore, color features are chosen as the basis for image segmentation. Commonly used color space models include RGB and HSV. Camera-acquired information typically uses the RGB color model. The RGB color model mixes color and luminance information across three channels, making it sensitive to changes in lighting conditions and unsuitable for this task. Therefore, the color model needs to be converted from RGB to HSV.

[0246] The RGB color model can be converted to the HSV color model using the following formula:

[0247] (31)

[0248] The formula for calculating hue H is:

[0249] (32)

[0250] The formula for calculating saturation S is:

[0251] (33)

[0252] The formula for calculating brightness V is:

[0253] (34)

[0254] The RGB image captured by the D435i camera was converted to the HSV color space. By defining the HSV range of the blue region of the steel strand in OpenCV: H (100~130), S (50~255), V (50~255), a mask for the blue region was generated, which enabled the separation of the steel strand from the background. The image segmentation result is shown in Figure 28. The blue steel strand, as the target object, is displayed as white, and the background is processed as black. Due to factors such as lighting and surrounding environmental interference, a small portion of the foreground and background image was not completely separated.

[0255] After image segmentation of the target object, noise interference was found in the image edge region. To effectively remove the background region before extracting the outline features of the steel strand, this study combines mathematical morphology algorithms to optimize the incomplete separation of background and foreground.

[0256] Morphological operations involve the interaction of a structuring element with an image to enhance or suppress features of specific shapes. Basic operations include erosion and dilation. Erosion involves sliding a structuring element across the image; the center pixel is set to 1 only when the structuring element completely covers the foreground. Erosion can shrink the boundaries of foreground objects, eliminating small blemishes and separating adhered objects. Dilation involves sliding a structuring element across the image; the center pixel is set to 1 when the area covered by the structuring element contains at least one foreground pixel. It can expand the boundaries of foreground objects, fill holes, and connect broken areas.

[0257] In image processing, morphological opening and closing operations play a significant role as complementary processing methods. Opening operations, through a combination of erosion and dilation, effectively eliminate the interference caused by discrete noise, discontinuous edges, and minute connections in the image. Conversely, closing operations employ a dilation-erosion strategy, which fills gaps and minor breaks within the target region and improves the jagged edges of objects. This combined operation allows for the optimized reconstruction of detailed features while maintaining geometric stability of the image's structural features. This paper applies opening and closing operations to the image segmentation results, as shown below. Figure 29 As shown.

[0258] Image edges are the boundaries between a target object and the background. By detecting image edges and extracting the edges of the target object, unimportant information in the image can be removed, significantly reducing the amount of image data. Edge features essentially originate from the differences in gray-level distribution between adjacent pixels; by capturing abrupt changes in pixel gray-level values, object boundaries can be detected.

[0259] The Canny edge detection algorithm is characterized by its good performance, strong noise resistance, and high accuracy, and is therefore widely used in various detection scenarios. Its core process consists of four steps: (1) Gaussian filtering for noise reduction; (2) Calculating gradient magnitude and direction; (3) Non-maximum suppression; and (4) Double thresholding and edge connection. The Canny edge detection algorithm is applied to the binary image after morphological operations, and the detection results are as follows: Figure 30 As shown in (a). Then, the `findContours` function in OpenCV is used to extract the contour features of the steel strand in the image, and the perimeter information of the steel strand contour is obtained through the `arcLength` function. By setting a perimeter threshold, noise points with small contour perimeters can be removed. The complete steel strand contour can then be obtained as shown in (a). Figure 30 As shown in (b).

[0260] Having obtained the complete edge contour region of the steel strand through the previous processing, to obtain the position information of the steel strand's positioning points, it is necessary to use the minimum bounding rectangle method on the extracted contour to obtain the position of the steel strand, thereby obtaining the pixel coordinates of the target positioning points. Then, through coordinate transformation, the spatial position information of the steel strand can be obtained. For any target object contour, it can be completely enclosed by a rectangle, ensuring that all contour points are located inside or on the boundary of the rectangle. The rectangle with the smallest area is called the minimum bounding rectangle. Since the minimum bounding rectangle of any contour is unique, its length and width can be used as important parameters for target detection.

[0261] In OpenCV, the preprocessed contour information is passed to the `minAreaRect` function. The function first calculates the convex hull of the contour, converts it into a polygon, and then uses a rotational caliper method to traverse each edge of the convex hull as a reference direction to find the rectangle with the smallest area. Finally, the rotation angle, center point coordinates, and dimensions of the smallest bounding rectangle of the contour are obtained. Figure 31 As shown, the green rectangle is the smallest bounding rectangle of the steel strand, the blue straight line is the major axis of the steel strand obtained by connecting the center points of the short side of the smallest bounding rectangle, and the red dots are the positioning points of the steel strand.

[0262] Because flexible photovoltaic (PV) supports apply significant prestress before installing PV panels, after obtaining a set of pixel coordinates for the positioning points of the steel strands, these coordinates can be transformed to obtain the three-dimensional spatial coordinates of the positioning points. This allows for the determination of the steel strands' spatial position information, enabling precise positioning of the component cables. Analysis of the physical characteristics of the steel strands reveals that the lengths of the long and short sides of their minimum bounding rectangle typically differ significantly. Therefore, an aspect ratio threshold can be set to further filter out non-target objects, ultimately achieving accurate identification and positioning of the steel strands.

[0263] After connecting the RealSense D435i camera, the RGB camera acquires color images, and the left and right infrared cameras acquire depth images. Since the color and depth images directly read from the cameras are not aligned, there is no one-to-one correspondence between the pixels of the two images. Therefore, the depth images need to be registered with the color images before reading the coordinates of the positioning points. Then, a target detection algorithm is used to detect the steel strand, obtaining the pixel coordinates and depth information of the positioning points. Substituting the obtained pixel coordinates and depth values ​​into the coordinate transformation formula obtained earlier, the three-dimensional coordinates of the target object's positioning points can be obtained. The detection results are as follows. Figure 32 As shown.

[0264] To verify the positioning effect of the positioning algorithm, the steel strand was placed in different positions for positioning experiments. The measured coordinates and actual coordinates of the steel strand endpoints were detected, and some results are shown in Table 12.

[0265] Table 12 - Results of the Positioning Experiment

[0266]

[0267] The three-dimensional coordinate errors of the steel strand endpoints are within 1.5mm in the X direction, within 1mm in the Y direction, and within 1.5mm in the Z direction. These error ranges meet the positioning accuracy requirements of the steel strand.

[0268] As the main execution unit of the automated photovoltaic module installation equipment, the articulated robotic arm is selected from Estun's ER35B-1810 model as the module cable alignment adjustment device. This robot has 6 degrees of freedom, a maximum load of 35 kg, an arm span of 1813 mm, and a repeatability of ±0.04 mm. It meets the requirements of adjusting the steel strands of the 40 m span double-layer cable-stayed flexible photovoltaic support module cables under a basic wind pressure of 50 N / m. 2 Linear control requirements within the specified range.

[0269] The position information of the steel strand positioning points obtained by the RealSense D435i camera in the previous article is based on the camera coordinate system. When the robotic arm grasps the steel strand positioning points and applies a pulling force to achieve component cable line control, the position points in the robotic arm base coordinate system need to be used. The mapping relationship between the camera coordinate system and the robotic arm base coordinate system can be established by hand-eye calibration to obtain the extrinsic parameter matrix, thereby converting the pixel coordinates of the steel strand positioning points into world coordinates. The position information of the target object relative to the robotic arm base coordinate system can be obtained through equation (27), thereby guiding the robotic arm to perform motion grasping.

[0270] Using a pinhole imaging model of the camera, the transformation matrix between the camera pixel coordinate system and the world coordinate system was derived. A calibration experiment was conducted on the RealSense D435i camera to obtain its parameter information, and stereo correction was performed on the image. Next, based on the color characteristics of the steel strand, the foreground of the image was segmented using the HSV image segmentation algorithm. The edge contour of the steel strand was obtained through morphological operations and the Canny edge detection algorithm. By setting a minimum bounding rectangle aspect ratio threshold, accurate identification of the steel strand was achieved, and the accuracy of the algorithm was verified through positioning experiments. Finally, the selection of a component cable shape adjustment robotic arm was completed to achieve the grasping control of the steel strand nodes.

[0271] The embodiments described above merely illustrate specific implementation methods of this application, and while the descriptions are detailed and specific, they should not be construed as limiting the scope of protection of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the technical solution of this application, and these modifications and improvements all fall within the scope of protection of this application.

Claims

1. A flexible support photovoltaic module installation device, characterized in that: include: A photovoltaic module clamping system includes a photovoltaic panel clamping mechanism for connecting photovoltaic panels and a moving mechanism for moving the photovoltaic panels onto a flexible support at a set angle. The flexible support adjustment mechanism is used to adjust the shape of the flexible support according to the installation requirements of the photovoltaic panels; the walking system is used to move the photovoltaic module clamping system and the flexible support adjustment mechanism to the installation position of the photovoltaic panels on the flexible support.

2. The flexible support photovoltaic module installation device according to claim 1, characterized in that: The moving mechanism includes a height adjustment component and an angle adjustment component. The height adjustment component includes an upper support (3) connected to the walking system. The upper support (3) is connected to a lower support (6) via a scissor lift mechanism. The lower support (6) is connected to the angle adjustment component. The angle adjustment component is connected to the photovoltaic panel clamping mechanism. The scissor lift mechanism includes a lifting cylinder (11). The angle adjustment component includes a support leg (13). The support leg (13) is hinged to one end of the photovoltaic panel clamping mechanism. The other end of the photovoltaic panel clamping mechanism is hinged to an angle cylinder (14).

3. The flexible support photovoltaic module installation device according to claim 1, characterized in that: The photovoltaic panel clamping mechanism includes a mounting frame (5) connected to the moving mechanism. The bottom of the mounting frame (5) is provided with several vacuum suction cups (10). The several vacuum suction cups (10) are divided into several vacuum suction cup groups, each vacuum suction cup group corresponds to a photovoltaic panel, and the several vacuum suction cup groups are arranged in a row with equal spacing.

4. A flexible support photovoltaic module installation device according to any one of claims 1-3, characterized in that: The flexible support adjustment mechanism includes four robotic arms (9) mounted on the photovoltaic panel clamping mechanism, with the four robotic arms (9) respectively mounted at the four corners of the front and rear ends of the photovoltaic panel clamping mechanism; the flexible support adjustment mechanism also includes an information vision camera (16) for real-time detection and acquisition of the spatial position of the component cable.

5. A flexible support photovoltaic module installation device according to any one of claims 1-3, characterized in that: The walking system includes a horizontal moving frame (1) and a vertical moving frame (2). The horizontal moving frame (1) is set perpendicular to the flexible support. The vertical moving frame (2) is located directly above the flexible support and extends along the length of the flexible support. The vertical moving frame (2) is set perpendicular to the horizontal moving support and can move along the horizontal moving frame (1). A moving trolley (4) is set on the vertical moving frame (2). The photovoltaic module clamping system is connected to the moving trolley (4).

6. A flexible support photovoltaic module installation device according to any one of claims 1-3, characterized in that: The flexible support is a double-layer cable structure, which includes two component cables (8). The photovoltaic panel includes a photovoltaic panel body (12). The bottom surface of the photovoltaic panel body (12) is provided with several cable buckles (15) that cooperate with the component cables (8).

7. A method for installing flexible support photovoltaic modules, characterized in that: The installation is carried out using the flexible bracket photovoltaic module installation device according to any one of claims 1-6. Includes the following steps: S1: Connect the photovoltaic panel through the photovoltaic panel clamping mechanism of the photovoltaic module clamping system; S2: The walking system drives the photovoltaic panel and photovoltaic module clamping system and the flexible support adjustment mechanism to move to the installation position of the photovoltaic panel on the flexible support; S3: The flexible support adjustment mechanism adjusts the shape of the flexible support according to the installation requirements of the photovoltaic panels; S4: The moving mechanism of the photovoltaic module clamping system adjusts the height and angle of the photovoltaic panel according to the shape of the flexible support, so that the photovoltaic panel contacts the flexible support; S5: Complete the connection between the photovoltaic panel and the flexible support.

8. The method for installing a flexible support photovoltaic module according to claim 7, characterized in that: In S3, a finite element model of a double-layer cable-stayed flexible photovoltaic support is established using SAP2000. The cable pretension is applied by selecting the target force iteration method. Based on the characteristics of the geometric nonlinearity of the cable structure, the initial stiffness matrix of the double-layer cable-stayed flexible photovoltaic support is formed on the basis of the initial stiffness of the structure. The cable pulling force of the component cable node is solved iteratively. Then, the relationship curve between the displacement of the component cable (8) node and the magnitude of the cable pulling force under normal installation conditions is obtained by analyzing the load optimizer. With the minimum cable pulling force of the node as the target, the optimal cable pulling force and ideal displacement are obtained by linear interpolation. The adjustment of the flexible support is completed by obtaining the optimal cable pulling force and ideal displacement.

9. The method for installing a flexible support photovoltaic module according to claim 7, characterized in that: In S3, the spatial position of the component cable (8) is detected in real time by a vision camera (16), and the flexible support adjustment mechanism is driven to complete the adjustment of the flexible support based on the obtained position of the component cable (8).

10. The method for installing a flexible support photovoltaic module according to claim 9, characterized in that: The specific process of obtaining the spatial position of component cable (8) in real time by using visual camera (16) is as follows: the transformation matrix between camera pixel coordinate system and world coordinate system is derived by using the pinhole imaging model of visual camera (16); the parameter information of camera is obtained by using Zhang Zhengyou calibration method to perform calibration experiment and the stereo correction of image is completed. Then, the image segmentation is completed by using HSV image segmentation algorithm according to the color characteristics of component cable (8). The edge contour of steel strand is obtained by morphological operation and Canny edge detection algorithm. By setting the aspect ratio threshold of the minimum bounding rectangle of the contour, non-target areas are filtered out, and accurate identification of component cable (8) is achieved.

Images