Electrical cabinet panel element high-precision assembly system and method based on digital twinning and visual guidance

Abstract

This invention relates to the fields of digital twin and computer vision technology, and in particular provides a high-precision assembly system and method for electrical cabinet panel components based on digital twin and vision guidance. The system includes: an acquisition and twin mapping subsystem that inputs an array of three-dimensional topographic data of the panel into an initial panel model in digital space, aligning the normal direction of corresponding coordinate points in the initial panel model with the micro-tilt direction of the actual panel surface, forming a dynamic twin panel model with a micron-level geometric deviation field; an assembly coordinate reconstruction subsystem that iteratively converges to obtain the compensated installation coordinates of each component on the actual panel surface, forming a set of corrected assembly instructions corresponding to the sequence of the deviation field; and a real-time comparison and feedback subsystem that superimposes residual deviation values ​​into the calculation process of the corrected assembly instructions for the next component, forming a point-by-point progressive closed-loop correction chain. This invention maintains assembly accuracy while avoiding the risk of secondary deformation caused by traditional fixtures.

Description

Technical Field

[0001] This invention relates to the fields of digital twin and computer vision technology, and in particular to a high-precision assembly system and method for electrical cabinet panel components based on digital twin and vision guidance. Background Technology

[0002] In the field of electrical cabinet manufacturing, the assembly accuracy of components directly determines the overall sealing, electromagnetic shielding, heat dissipation, and long-term reliability of the machine. Traditional assembly relies on rigid benchmarks and mechanical fixtures: first, a theoretical coordinate system is established on a tooling plate; then, tooling pins, pressure plates, or bolts are used to forcibly fix the panel in the theoretical position; finally, the components are installed according to the coordinates given in the two-dimensional drawings. With the increase in power density of electrical cabinets and the rise in customized orders, panel thickness is decreasing and opening density is increasing. Traditional rigid assembly has exposed systemic shortcomings due to the coupling of multiple factors such as surface warping, manufacturing errors, and fixture springback.

[0003] Existing technologies suffer from several problems: Rigid coordinate references: All rely on an ideal panel as the reference, ignoring the micron-level undulations and local tilts of the actual panel caused by stamping, welding, and spraying. This results in gaps or stress concentrations of 0.2-1 mm after assembly. One-way error transmission: Laser scanning or visual inspection is mostly used for spot checks; the results cannot be fed back to the coordinate compensation of subsequent components in real time, leading to error accumulation. Secondary deformation introduction: Mechanical fixtures or vacuum adsorption platforms generate additional bending moments on the thin plate during positioning, introducing new deformations and amplifying the initial error. Broken data link: Digital twins only exist at the macroscopic simulation level, lacking micron-level mapping with the on-site 3D topographic point cloud, thus failing to form an assembly closed loop. Heavy manual intervention: Analysis of laser point clouds or visual images relies on the experience of process engineers, resulting in long parameter tuning cycles and making it difficult to adapt to flexible manufacturing scenarios with small batches and multiple varieties.

[0004] How to generate unique and real-time updated compensated installation coordinates for each component on the surface of an electrical cabinet panel, where there are micron-level manufacturing errors and deformation uncertainties, to achieve a zero-stress, zero-gap fit between the component's bottom surface and the panel's curved surface, and to immediately feed back single-point residual errors to the coordinate correction of subsequent components, thus constructing a closed-loop precision control system that spans the entire process of measurement, modeling, compensation, execution, and re-correction, is a major technical problem that existing technologies continue to address. Summary of the Invention

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] One aspect of the present invention provides a high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance, comprising:

[0007] The data acquisition and twin mapping subsystem is configured to fix the electrical cabinet panel to the air-floating platform, and use a three-axis laser displacement sensor to scan the front of the panel along a preset path to obtain the surface undulation height value and local tilt angle value at each installation coordinate point. After data cleaning, a three-dimensional topography data array of the panel is obtained. The panel three-dimensional topography data array is input into the panel initial model in digital space, so that the normal direction of the corresponding coordinate point in the panel initial model is consistent with the micro-tilt direction of the actual panel surface, forming a dynamic twin panel model with a micron-level geometric deviation field.

[0008] The assembly coordinate reconstruction subsystem is configured to extract the deviation field values ​​in the dynamic twin panel model, apply an offset vector coupled with the surface tilt angle at the theoretical installation center point to each component in the standard installation drawing, and obtain the compensated installation coordinates of each component on the actual surface of the panel through iterative convergence. The compensated installation coordinates form a set of corrected assembly instructions corresponding to the sequence of deviation fields.

[0009] The real-time comparison and feedback subsystem is configured to send the generated set of correction assembly instructions to the six-DOF assembly robot. The pneumatic gripper at the end of the robot grasps the components and positions them one by one according to the compensation coordinates in the instructions. At the same time, the coaxial structured light projector installed at the end of the robot projects a crosshair pattern onto the current assembly point. The high-speed camera captures the image of the relative gap between the bottom edge of the component and the preset reference mark on the panel. The residual deviation value between the actual assembly position and the compensation coordinates is obtained by edge extraction. The residual deviation value is superimposed on the calculation process of the correction assembly instruction for the next component, forming a point-by-point progressive closed-loop correction chain.

[0010] Another aspect of the present invention provides a high-precision assembly method for electrical cabinet panel components based on digital twins and vision guidance, comprising the following steps:

[0011] The electrical cabinet panel is fixed to the air-floating platform. A three-axis laser displacement sensor scans the front of the panel along a preset path to obtain the surface undulation height value and local tilt angle value at each installation coordinate point. After data cleaning, a three-dimensional topography data array of the panel is obtained. The three-dimensional topography data array of the panel is input into the initial model of the panel in digital space so that the normal direction of the corresponding coordinate point in the initial model of the panel is consistent with the micro-tilt direction of the actual panel surface, forming a dynamic twin panel model with a micron-level geometric deviation field.

[0012] Extract the deviation field values ​​from the dynamic twin panel model, and apply an offset vector coupled with the surface tilt angle at the theoretical installation center point to each component in the standard installation drawing; obtain the compensated installation coordinates of each component on the actual surface of the panel through iterative convergence, and form a set of corrected assembly instructions corresponding to the deviation field sequence.

[0013] The generated set of corrective assembly instructions is fed into a six-DOF assembly robot. The pneumatic gripper at the end of the robot grasps the components and positions them one by one according to the compensation coordinates in the instructions. At the same time, a coaxial structured light projector installed at the end of the robot projects a crosshair pattern onto the current assembly point. A high-speed camera captures the image of the relative gap between the bottom edge of the component and the preset reference mark on the panel. The residual deviation value between the actual assembly position and the compensation coordinates is obtained by edge extraction. The residual deviation value is superimposed on the corrective assembly instructions of the next component to form a point-by-point progressive closed-loop correction chain.

[0014] This invention achieves digital modeling of micron-level geometric deviations on actual assembly surfaces through dynamic mapping of a panel's 3D topographic data array acquired by a triaxial laser displacement sensor with a digital twin model. The technical feature of converting theoretical coordinates from standard installation drawings into compensated installation coordinates effectively solves the problem of assembly stress concentration caused by panel manufacturing tolerances. Employing an offset vector calculation method based on surface tilt angle coupling, combined with the spatial pose compensation function of a six-DOF manipulator, standard parts can achieve normal-axis fit assembly on panel surfaces with curved deformations. A closed-loop correction chain formed by coaxial structured light projection and high-speed visual inspection transforms single-point assembly residuals into dynamic compensation amounts during the assembly process. By integrating macroscopic topographic data from laser scanning, microscopic positioning data from structured light projection, and assembly residual data from visual inspection, a 3D precision control system covering the entire assembly process is constructed, enabling the system to simultaneously compensate for initial panel manufacturing deviations and cumulative errors during assembly. The technical feature of dynamically updating the assembly instruction set with the assembly process enables real-time adaptive adjustment of assembly process parameters; it is particularly suitable for large-scale customized production scenarios with batch-specific panel deformation. The non-contact measurement technology of the air-floating platform and triaxial sensors, combined with the online correction function of the robot's motion trajectory, maintains assembly accuracy while avoiding the risk of secondary deformation caused by traditional fixtures. Attached Figure Description

[0015] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0016] Figure 1 This is a block diagram of the high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance provided in Embodiment 1 of the present invention;

[0017] Figure 2 This is a schematic diagram of the high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance provided in Embodiment 1 of the present invention;

[0018] Figure 3 This is a block diagram of the data acquisition and twin mapping subsystem provided in Embodiment 2 of the present invention;

[0019] Figure 4 This is a block diagram of the assembly coordinate reconstruction subsystem provided in Embodiment 4 of the present invention;

[0020] Figure 5 This is a block diagram of the real-time comparison feedback subsystem provided in Embodiment 10 of the present invention;

[0021] Figure 6 This is a flowchart of the high-precision assembly method for electrical cabinet panel components based on digital twin and vision guidance provided in Embodiment 11 of the present invention;

[0022] Figure 7 A block diagram of the electronic device provided by the present invention;

[0023] Figure 8 A block diagram of a computer-readable storage medium provided for this invention. Detailed Implementation

[0024] The technical solutions of the present invention will now be described with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0025] Hereinafter, the terms "first," "second," etc., are used for descriptive convenience only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "multiple" means two or more.

[0026] In this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, a connection can be a fixed mechanical connection, a detachable mechanical connection, or an integral part; or, a connection can be a direct connection or an indirect connection through an intermediate medium. Furthermore, unless otherwise explicitly specified and limited, the term "coupling" should be interpreted broadly. For example, coupling can be a direct electrical connection, such as physical contact and electrical conduction between two components; it can also be understood as an electrical connection between different components in a circuit structure through physical lines capable of transmitting electrical signals, such as copper foil or wires on a printed circuit board (PCB), to transmit electrical signals; or, coupling can be an indirect electrical connection between two components through an intermediate medium; or, coupling can be an electrical connection between two components in a non-contact manner, such as an electrical connection between two components using capacitive coupling to transmit electrical signals.

[0027] In this embodiment of the invention, directional terms such as up, down, left, and right may be defined relative to the orientation of the components shown in the accompanying drawings. It should be understood that these directional terms can be relative concepts, used for relative description and clarification, and can change accordingly depending on the orientation of the components in the accompanying drawings.

[0028] Example 1: As Figure 1 As shown, this embodiment of the invention provides a high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance, comprising:

[0029] The data acquisition and twin mapping subsystem is configured to fix the electrical cabinet panel to the air-floating platform, and use a three-axis laser displacement sensor to scan the front of the panel along a preset path to obtain the surface undulation height value and local tilt angle value at each installation coordinate point. After data cleaning, a three-dimensional topography data array of the panel is obtained. The panel three-dimensional topography data array is input into the panel initial model in digital space, so that the normal direction of the corresponding coordinate point in the panel initial model is consistent with the micro-tilt direction of the actual panel surface, forming a dynamic twin panel model with a micron-level geometric deviation field.

[0030] The assembly coordinate reconstruction subsystem is configured to extract the deviation field values ​​in the dynamic twin panel model, apply an offset vector coupled with the surface tilt angle at the theoretical installation center point to the theoretical installation center point of each component in the standard installation drawing, and obtain the compensated installation coordinates of each component on the actual surface of the panel through iterative convergence. The compensated installation coordinates form a set of corrected assembly instructions corresponding to the order of the deviation fields.

[0031] The real-time comparison and feedback subsystem is configured to send the generated set of correction assembly instructions to the six-DOF assembly robot. The pneumatic gripper at the end of the robot grasps the components and positions them one by one according to the compensation coordinates in the instructions. At the same time, the coaxial structured light projector installed at the end of the robot projects a crosshair pattern onto the current assembly point. The high-speed camera captures the image of the relative gap between the bottom edge of the component and the preset reference mark on the panel. The residual deviation value between the actual assembly position and the compensation coordinates is obtained by edge extraction. The residual deviation value is superimposed on the calculation process of the correction assembly instruction for the next component, forming a point-by-point progressive closed-loop correction chain.

[0032] The preset path refers to the fixed spatial trajectory of the triaxial laser displacement sensor's measuring head relative to the panel surface when scanning the front of the electrical cabinet panel. The trajectory is pre-set based on the panel's geometric boundaries and component layout diagram, using a parallel line reciprocating motion to ensure the sensor head sequentially passes through each installation coordinate point where data needs to be collected. The installation coordinate point refers to the theoretical center point coordinates of each electrical component's installation position on the front of the electrical cabinet panel, determined according to the design drawings. These coordinates are established based on the panel's original design dimensions and serve as a position index for collecting surface undulation height and local tilt angle values ​​during system operation. The undulation height value is the microscopic height difference of the panel surface relative to the reference plane along the normal direction, measured by the triaxial laser displacement sensor at a specific installation coordinate point. The local tilt angle value is the angle between the surface tangent plane at that coordinate point and the reference plane, including both horizontal and vertical tilt angles. These two values ​​together constitute the surface micro-morphological features at that installation coordinate point. The initial panel model in digital space refers to the ideal geometric model of the electrical cabinet panel built in a computer. This model is generated based on the computer-aided design drawings of the panel, and the normal direction of all coordinate points is perpendicular to the ideal reference plane of the panel. The initial model itself does not contain the manufacturing errors and installation deformations of the actual panel and serves as a carrier for receiving the three-dimensional topographic data array. The micron-level geometric deviation field refers to the spatial distribution field formed after mapping the undulation height value and local tilt angle value of each installation coordinate point in the panel's three-dimensional topographic data array to the corresponding coordinate points in the initial panel model in digital space. Each point in this deviation field contains a height offset and tilt values ​​in two orthogonal directions, with a numerical order of microns, representing the total geometric deviation of the actual panel surface from the ideal model. The standard installation drawing refers to the electronic data of the design document for the assembly of electrical cabinet panel components, which includes the type identification of each component, the coordinates of the theoretical installation center point, the bottom surface dimensions of the component, the installation orientation, and the minimum clearance requirements between adjacent components. This drawing serves as the assembly reference under the condition of no actual panel deformation error. After being input into the assembly coordinate reconstruction module, it is coupled with the deviation field of the dynamic twin panel model for calculation.

[0033] In the above embodiments, the principle is referenced in the appendix. Figure 2This embodiment achieves digital modeling of micron-level geometric deviations of the actual assembly surface through dynamic mapping of the panel's three-dimensional topography data array acquired by a three-axis laser displacement sensor with a digital twin model. The technical feature of converting the theoretical coordinates of standard installation drawings into compensated installation coordinates effectively solves the problem of assembly stress concentration caused by panel manufacturing tolerances. Employing an offset vector calculation method based on surface tilt angle coupling, combined with the spatial pose compensation function of a six-degree-of-freedom robot, standard parts can achieve normal-axis fit assembly on panel surfaces with curved deformations. A closed-loop correction chain formed by coaxial structured light projection and high-speed visual inspection transforms single-point assembly residuals into dynamic compensation amounts during the assembly process. By integrating macroscopic topography data from laser scanning, microscopic positioning data from structured light projection, and assembly residual data from visual inspection, a three-dimensional precision control system covering the entire assembly process is constructed, enabling the system to simultaneously compensate for initial panel manufacturing deviations and cumulative errors during assembly. The technical feature of dynamically updating the assembly instruction set with the assembly process enables real-time adaptive adjustment of assembly process parameters; it is particularly suitable for large-scale customized production scenarios with batch-specific panel deformation. The non-contact measurement technology of the air-floating platform and triaxial sensors, combined with the online correction function of the robot's motion trajectory, maintains assembly accuracy while avoiding the risk of secondary deformation caused by traditional fixtures.

[0034] Example 2: Figure 3 As shown, based on Embodiment 1, the acquisition and twin mapping subsystem provided in this embodiment of the invention includes:

[0035] The elevation correction value generation component is configured to decompose the local tilt angle value at each installation coordinate point obtained by the triaxial laser displacement sensor along a preset path into a horizontal tilt component and a vertical tilt component; based on the decomposed horizontal and vertical tilt components, the surface undulation height value at the same installation coordinate point is converted into an elevation correction value along the normal direction of the mesh node in the initial model of the panel that is at the same position as the installation coordinate point; the horizontal tilt component, the vertical tilt component, and the elevation correction value are combined into a shape feature vector, and the shape feature vectors of all installation coordinate points together constitute a shape feature vector set;

[0036] The vector iterative retargeting component is configured to extract the horizontal and vertical tilt components corresponding to each installation coordinate point in the generated topographic feature vector set; the horizontal and vertical tilt components are used as rotation transformation parameters and applied to the unit normal vectors of the mesh nodes in the initial panel model that are at the same position as the installation coordinate points, so that the unit normal vectors of the mesh nodes are deflected around the horizontal and vertical axes in sequence to be consistent with the measured local tilt direction at the installation coordinate points; after completing the unit normal vector deflection operation for all mesh nodes in the initial panel model that are at the same position as the installation coordinate points, a surface fitting intermediate model is obtained in which all mesh node normals have been retargeted and the original mesh topology is preserved;

[0037] The dynamic twin model output component is configured to subtract the original local tilt angle value at the installation coordinate point corresponding to each grid node in the intermediate model of the surface fitting from the redirected normal vector, thus obtaining the residual normal alignment error at the grid node; and algebraically add the residual normal alignment error at each installation coordinate point to the corresponding elevation correction value, thus obtaining the total displacement at the installation coordinate point; using the base grid of the initial panel model as the carrier, the total displacement at each installation coordinate point is applied to the grid nodes in the initial panel model that are at the same position as the installation coordinate point, driving the grid nodes to move along their own redirected normal direction; after all grid nodes have moved, the base grid surface forms a continuous surface that perfectly fits the measured three-dimensional topography data array, outputting a dynamic twin panel model with a micron-level geometric deviation field.

[0038] The original mesh topology refers to the connection relationships and adjacency order between all mesh nodes in the initial panel model. This structure is determined during the discretization of the panel's computer-aided design drawings and includes the number of each mesh node, a list of adjacent nodes connected to each node, and the patch division method of the triangular or quadrilateral units formed by the nodes. During the mesh node normal vector deflection process, only the direction of the unit normal vector of each node is changed; the connection relationships, adjacency order, and patch division shape between nodes remain unchanged. The surface fitting intermediate model refers to the intermediate data model obtained by sequentially deflecting the unit normal vectors of all mesh nodes in the initial panel model that are at the same position as the installation coordinate point around the horizontal and vertical axes to align with the measured local tilt direction at the corresponding installation coordinate point. The normal vectors have been redirected, but the spatial positions of the mesh nodes have not yet moved. Each mesh node in the model holds the redirected normal vector while retaining the connection relationships between each node and its adjacent nodes in the original mesh topology, serving as a transitional form before applying the total displacement to drive the nodes to move along the redirected normal direction.

[0039] In the above embodiments, this embodiment decomposes and converts the local tilt angle value into a tilt correction value along the normal direction of the grid node through the tilt correction value generation component, so that the electrical cabinet panel obtains a micron-level height mapping equivalent to the measured surface undulation for the first time in the digital domain; the vector iteration redirection component uses the decomposed horizontal and vertical tilt components as rotation parameters to sequentially deflect the unit normal vector of the corresponding grid node in the initial model of the panel, and locks the component assembly reference normal and the measured local tilt direction as unique while maintaining the grid topology unchanged; the dynamic twin model output component calculates the residual normal alignment error and algebraically adds it to the tilt correction value to obtain the total displacement, and then drives the node to move along the redirection normal with the basic grid as the carrier, so that the electrical cabinet panel finally generates a dynamic twin model that is completely consistent with the measured three-dimensional topography data array and has a micron-level geometric deviation field; the dynamic twin model constitutes the only legal reference for all subsequent assembly processes in the sense of patent law, and any component position deviation caused by panel warping, stress rebound or processing error can be compared, traced and suppressed in real time within the tolerance zone defined by the reference.

[0040] Example 3: Based on Example 2, the dynamic twin model output component provided in this embodiment of the invention includes:

[0041] The mesh node spatial coordinate binding sub-component is configured to bind the total displacement at each installation coordinate point obtained from the dynamic twin model output component with the current spatial coordinates of the mesh node in the panel initial model corresponding to the installation coordinate point, forming a displacement loading record table indexed by the mesh node number. Each row in the table records the mesh node number, the redirected normal vector component value of the node, and the total displacement value in sequence.

[0042] The point-to-point displacement driving subcomponent is configured to extract the redirected normal vector component value and the total displacement value of each mesh node from the generated displacement loading record table, multiply the total displacement value by the unit direction component of the redirected normal vector to obtain the three-dimensional spatial movement step size of the mesh node in the panel reference coordinate system, and superimpose the movement step size onto the current spatial coordinates of the mesh node so that the mesh node moves along its own redirected normal direction by a distance equal to the total displacement.

[0043] The dynamic twin model output sub-component is configured to regenerate the geometry of each triangle or quadrilateral facet according to the node connection relationships and facet division methods recorded in the original mesh topology, based on the new spatial coordinates of all mesh nodes that have completed the displacement drive. The node coordinates on the common edges between adjacent facets are checked one by one to ensure that there are no node separations. After confirming that there are no node separations, a continuous surface that perfectly matches the measured 3D topography data array is obtained. The continuous surface is then stitched together with the boundary of the non-measured area of ​​the initial panel model to output a dynamic twin panel model with a micron-level geometric deviation field.

[0044] In the above embodiments, this embodiment forms a closed-loop compensation mechanism for the micron-level geometric deviation field between the measured three-dimensional topography data of the electrical cabinet panel components and the initial CAD model by binding sub-components with the spatial coordinates of the mesh nodes, driving sub-components with point-by-point displacement, and allowing the output sub-components of the dynamic twin model to interact sequentially. This mechanism performs the following functions in real time on the assembly site: accurately binding the total displacement measured on site to the initial panel model with the node number as a unique index; using the redirected normal vector as the unique directional reference to ensure that each node only moves along its local normal to the same length as the measured displacement; subsequently reconstructing a continuous surface based on the original mesh topology and performing node consistency verification at common edges to eliminate the risk of node separation; finally, the output dynamic twin panel model is seamlessly spliced ​​with the initial model in non-measurement areas, thereby forming a continuous geometric expression that perfectly matches the measured data on the entire panel surface.

[0045] In this embodiment, during the high-precision assembly of electrical cabinet panel components, a dynamic twin model with zero deviation from the measured morphological data is provided in the form of a single continuous curved surface. This ensures that the micron-level correspondence between the theoretical coordinates and actual coordinates of the assembly reference surface, positioning pin holes, and snap-fit ​​structures is always traceable and reproducible. Through forced coupling of node-level displacement loading and normal redirection, the accumulation of assembly gaps caused by coordinate system transformation, direction errors, or mesh distortion is eliminated, reducing the assembly error of panel components from the traditional millimeter level to the micron level, thus meeting the electrical cabinet's overall sealing, electromagnetic shielding, and structural strength specifications. This dynamic twin model can be directly referenced as the legal geometric benchmark in assembly process documents, quality inspection reports, and subsequent maintenance manuals, achieving data consistency throughout the entire lifecycle of design, manufacturing, and testing.

[0046] Example 4: Figure 4 As shown, based on Embodiment 1, the assembly coordinate reconstruction subsystem provided in this embodiment of the invention includes:

[0047] The offset vector initial value generation component is configured to extract the micron-level geometric deviation field value at the location of each theoretical installation center point in the dynamic twin panel model, and extract the height offset, horizontal tilt component, and vertical tilt component of the location respectively; decompose the height offset along the redirected normal direction of the location to obtain the normal lift; combine the horizontal tilt component and the vertical tilt component to form the tangential plane deflection direction parameter at the location; use the tangential plane deflection direction parameter as the direction reference of the offset vector, use the normal lift as the first part of the magnitude of the offset vector, and use half of the bottom surface outer dimension of the component at the theoretical installation center point multiplied by the sine value of the tangential plane deflection angle as the second part of the magnitude of the offset vector; algebraically add the first part and the second part and multiply by the direction reference to generate the initial value of the offset vector of the theoretical installation center point;

[0048] The compensation coordinate calculation component is configured to apply the initial value of the offset vector of each generated theoretical installation center point to the coordinates of the theoretical installation center point of the component in the standard installation drawing to obtain the component's trial compensation coordinates. Centered on the trial compensation coordinates, a rectangular area with the same dimensions as the bottom surface of the component is extracted on the dynamic twin panel model. The redirected normal vectors and spatial coordinates of all grid nodes within the rectangular area are extracted. The vertical distance between the spatial coordinates of the grid nodes at the four corners of the rectangular area and the theoretical plane of the bottom surface of the component is calculated to obtain four edge gap values. If the maximum value of the four edge gap values ​​exceeds the preset bonding threshold, the tangent plane deflection direction parameter in the offset vector is adjusted according to the direction sign of the maximum value, and the adjusted direction parameter is resubmitted into the offset vector generation process to obtain the updated offset vector. The adjustment and update are repeated until all four edge gap values ​​are less than the preset bonding threshold. The compensation coordinates obtained from the last update are used as the final compensation installation coordinates of the component on the actual surface of the panel.

[0049] The modified assembly instruction set output component is configured to extract the final compensated installation coordinates corresponding to each component, and sequentially retrieve the coordinate values ​​according to the original assembly order of the components in the standard installation drawing. Each final compensated installation coordinate is bound to the height direction offset in the micron-level geometric deviation field at the installation coordinate point in the dynamic twin panel model, forming a single instruction record containing the component identifier, compensated installation coordinate value, normal vector component at the coordinate, and height offset. All instruction records are arranged according to the original assembly order to form a modified assembly instruction set corresponding to the deviation field order.

[0050] In the above embodiments, this embodiment achieves systematic utilization of the micron-level geometric deviation field in the dynamic twin panel model through the sequential action of the offset vector initial value generation component, the compensation coordinate calculation component, and the correction assembly instruction set output component. This enables controllable and predictable offset compensation of the actual installation coordinates of the electrical cabinet panel components relative to the theoretical installation coordinates; and simultaneously eliminates the fitting gap caused by the manufacturing error of the panel's free-form surface in both the normal and tangential dimensions. The offset vector initial value generation component uses the micron-level geometric deviation field value at the theoretical installation center point as the sole data source; through height direction offset decomposition, tilt component combination, and direction reference setting, it generates an initial value of the offset vector with dual attributes of directionality and modulus, so that the compensation calculation is based on the accurate characterization of the actual micro-morphology of the panel, avoiding assembly stress concentration or poor fitting caused by traditional rigid references. Based on the trial compensation coordinates, the compensation coordinate calculation component extracts a local area of ​​the dynamic twin panel model with the bottom surface dimensions of the component as the boundary, and calculates the edge gap values ​​at the four corners of this area. Then, using the maximum edge gap value as a feedback signal, it adjusts the offset vector in a closed loop until all edge gap values ​​are below the preset bonding threshold. This ensures that the final compensated installation coordinates of each component meet the requirements for seamless bonding with the actual panel surface, preventing electrical clearance non-compliance or mechanical fixation failure caused by local warping or residual height differences from processing. The corrected assembly instruction set output component binds the final compensated installation coordinates with the corresponding micron-level geometric deviation field height offset and normal vector components, forming a set of single instruction records arranged according to the original assembly sequence. As a digital tooling data package, the instruction set can directly drive automated assembly equipment or guide manual assembly, achieving traceability and reproducibility of the panel component assembly process, meeting the legal quality standards for assembly accuracy and batch consistency in the electrical cabinet industry.

[0051] In summary, this embodiment has outstanding substantive features compared to existing technologies in the field of high-precision assembly of electrical cabinet panel components. It can systematically and in a closed-loop manner transform panel micro-manufacturing errors into compensable installation coordinate corrections, significantly improving assembly accuracy, reducing rework rates, and ensuring the final product's electrical performance, mechanical strength, and long-term reliability.

[0052] Example 5: Based on Example 4, the compensation coordinate calculation component provided in this embodiment of the invention includes:

[0053] The spatial coordinate extraction sub-component is configured to extract a rectangular area with the same dimensions as the bottom surface of the component from the dynamic twin panel model, centered on the trial compensation coordinates. Then, it locates four grid nodes at the four corner points of the rectangular area from all the grid nodes contained in the rectangular area, and extracts the spatial coordinate values ​​of these four grid nodes in the dynamic twin panel model to form a set of spatial coordinates of the four corner points.

[0054] The vertical distance calculation sub-component is configured to substitute the spatial coordinates of each corner point in the four corner point spatial coordinate sets with the theoretical plane equation of the component bottom surface retrieved from the standard installation drawing, and calculate the vertical distance value from each corner point spatial coordinate to the theoretical plane, thus obtaining four original vertical distance values.

[0055] The gap value sequence generation sub-component is configured to arrange the four original vertical distance values ​​obtained according to the spatial orientation of the four corner points of the rectangular area, corresponding to the four edge positions of the upper left, upper right, lower right and lower left corners, to form four edge gap values; output the four edge gap values ​​as the basis for judging whether the preset fitting threshold has been reached.

[0056] In the above embodiments, the spatial coordinate extraction sub-component extracts a rectangular area with the same dimensions as the bottom surface of the component on the dynamic twin panel model, centered on the trial compensation coordinates. It then precisely locates four grid nodes at the four corner points of the rectangular area from all the grid nodes contained within it, extracting the spatial coordinate values ​​of these four grid nodes in the dynamic twin panel model to form a set of spatial coordinates for the four corner points. This allows the system to dynamically and accurately obtain the actual spatial position information of the four key corner points of the component's bottom surface, providing precise input data for calculating the gap value and ensuring accurate perception of the actual spatial relationship between the component's bottom surface and the dynamic twin panel model during assembly. The vertical distance calculation subcomponent calculates the vertical distance from each corner point in the four corner point spatial coordinate sets to the theoretical plane equation of the component's bottom surface retrieved from the standard installation drawing. This yields four initial vertical distance values. This allows the system to quantitatively assess the deviation of the four key corner points of the component's bottom surface relative to the theoretical installation plane, providing accurate deviation data for generating the gap value sequence. This ensures precise measurement of the vertical distance between the component's bottom surface and the theoretical installation plane during assembly. The gap value sequence generation subcomponent arranges the four initial vertical distance values ​​according to the spatial orientation of the four corner points of the rectangular area, corresponding to the top left, top right, bottom right, and bottom left edges, forming four edge gap values. These four edge gap values ​​are then output as the basis for determining whether a preset fitting threshold has been reached. This allows the system to systematically integrate the vertical distance deviations of the four key corner points into an edge gap value sequence, providing an intuitive and quantitative basis for evaluating the fitting status during assembly, thus ensuring high-precision fitting of electrical cabinet panel components during assembly.

[0057] In summary, this embodiment enables precise perception, quantitative evaluation, and fitting status judgment of the spatial relationship between the bottom surface of the electrical cabinet panel components and the dynamic twin panel model and theoretical mounting plane during the assembly process, ensuring high-precision assembly of the electrical cabinet panel components.

[0058] Example 6: Based on Example 5, the vertical distance calculation sub-component provided in this embodiment of the invention includes:

[0059] The planar parameter pairing module is configured to sequentially extract the spatial coordinates of each corner point from the four corner point spatial coordinate sets obtained from the spatial coordinate extraction sub-component; at the same time, it retrieves the spatial orientation parameter set of the theoretical plane of the component's bottom surface from the standard installation drawing; the spatial orientation parameter set contains the normal pointing data of the theoretical plane and the coordinates of the reference points through which the plane passes; and binds each corner point spatial coordinate with the reference point coordinates and normal pointing data in a one-to-one correspondence to form four sets of coordinate and planar parameter pairs;

[0060] The distance value conversion module is configured to calculate the spatial position deviation vector of the corner point's spatial coordinates relative to the plane reference point, decompose the deviation vector by projection along the direction of the theoretical plane normal, and extract the projection length of the deviation vector in the normal direction; determine the sign of the projection length based on whether the projection direction is in the same or opposite direction to the normal direction, and obtain the directed vertical distance value at the corner point; perform conversion on four sets of coordinate and plane parameter pairs respectively to obtain four directed vertical distance values;

[0061] The distance sequence arrangement output module is configured to obtain four directional vertical distance values, and arrange them in the order of the top left, top right, bottom right, and bottom left corners when locating the four corner points in the spatial coordinate extraction sub-component, forming four original vertical distance value sequences.

[0062] In the above embodiments, the planar parameter matching module of this embodiment extracts the spatial coordinates of each corner point output by the spatial coordinate extraction sub-component in sequence, and binds them one-to-one with the set of theoretical plane spatial orientation parameters of the component bottom surface retrieved by the standard installation drawing. This ensures that the spatial coordinates of each corner point form a clear and unique coordinate-planar parameter pair with the normal pointing data of the theoretical plane and the coordinates of the reference point. This ensures that the spatial reference on which the calculation depends maintains a strict correspondence with the theoretical installation state, avoids coordinate system drift or parameter misalignment caused by the mixing of multiple source data, and provides a traceable and reproducible spatial positioning reference for the assembly of electrical cabinet panel components. The distance value conversion module decomposes the spatial position deviation vector of the corner point's spatial coordinates relative to the plane reference point by projecting it along the normal direction of the theoretical plane. It then assigns positive or negative signs to the projected length based on whether the projection direction is in the same or opposite direction to the normal, generating a directed vertical distance value at the corner point. This allows the system to accurately quantify the vertical deviation and direction between any key corner point on the component's bottom surface and the theoretical mounting plane, fully representing the spatial error components between the actual assembly surface and the theoretical surface in vector form. This provides a direction-sensitive and numerically accurate error measurement basis for fit evaluation. The distance sequence arrangement output module serializes the obtained directed vertical distance values ​​according to the spatial orientation order of the upper left, upper right, lower right, and lower left corners, outputting a uniformly formatted vertical distance value sequence. The system can present the spatial error data in a regular order corresponding one-to-one with the physical edges of the component's bottom surface, facilitating edge-by-edge comparison with preset fit thresholds. This enables accurate judgment of the synchronous fit status of each edge of the electrical cabinet panel component during assembly and closed-loop control of the overall assembly accuracy.

[0063] In summary, this embodiment enables the precise vector-level measurement, direction-sensitive quantitative characterization, and edge-sequential output of the vertical deviation between key corner points on the bottom surface of electrical cabinet panel components and the theoretical mounting plane in high-precision assembly scenarios, providing technical support for assembly fit status evaluation and subsequent compensation correction.

[0064] Example 7: Based on Example 6, the distance value conversion module provided in this embodiment of the invention includes:

[0065] The spatial offset decomposition submodule is configured to compare the spatial coordinates of the corner points and the reference point coordinates of each set of coordinates in the planar parameter pairing module with the coordinates of the reference point. It measures the offset distance of the corner point relative to the reference point in each axis along the three orthogonal axes of the reference coordinate system of the dynamic twin panel model to obtain the original offset component values ​​of the three axes.

[0066] The symbol marking submodule is configured to extract the values ​​of the three original axial offset components corresponding to a set of corner points, identify the quadrant of the corner point relative to the reference point based on the spatial orientation order of the rectangular area where the corner point is located, determine the positive or negative sign to be assigned to each axial offset component according to the quadrant, and combine the original offset component values ​​with the corresponding positive or negative signs to form three signed axial offset components.

[0067] The spatial deviation vector construction submodule is configured to arrange the three signed axial offset components corresponding to a set of corner points according to the axis order of the reference coordinate system, and combine them into a spatial deviation vector with direction and magnitude. The spatial offset decomposition submodule and the symbol marking submodule are executed on the remaining three sets of coordinate and plane parameter pairs respectively. Each completed set generates a corresponding spatial deviation vector, and the four spatial deviation vectors are stored in a temporary buffer for use in the projection decomposition step.

[0068] In the above embodiments, the spatial offset decomposition submodule measures the offset distance of the corner point relative to the reference point along the three orthogonal axes of the reference coordinate system, discretizing the originally unified spatial error into quantifiable axial original offset components. The sign marking submodule further assigns positive and negative signs to each component, so that each component carries directional information simultaneously. The spatial deviation vector construction submodule arranges the signed components in axial order to form a spatial deviation vector with a clear direction and magnitude. This continuous operation chain enables the corner point error of any rectangular area of ​​the electrical cabinet panel to be completely and losslessly mapped to the three-dimensional coordinate system, achieving sub-millimeter-level positioning accuracy of the panel components relative to the reference point, meeting the legal technical specifications for high-precision assembly. The above decomposition, marking, and vector construction process is repeated for each set of coordinate and planar parameter pairs to generate four independent spatial deviation vectors and temporarily store them in a temporary buffer. The parallel processing method ensures that all corner points complete error capture within the same measurement cycle, avoiding cumulative errors caused by time-division measurement, and ensuring the spatial pose consistency of multiple panel components during batch assembly. The four spatial deviation vectors are stored in a structured form in a temporary buffer for use in the projection decomposition step, achieving seamless integration of error data; the caching mechanism forms a complete data link.

[0069] Example 8: Based on Example 7, the spatial offset decomposition submodule provided in this embodiment of the invention includes:

[0070] The axial alignment unit is configured to retrieve a set of coordinate and planar parameter pairs from the planar parameter pairing module, and read the three axial values ​​of the corner point spatial coordinates and the three axial values ​​of the reference point coordinates in the planar parameter pair respectively; the axial values ​​of the corner point and the axial values ​​of the reference point are placed in a one-to-one correspondence according to the X-axis, Y-axis and Z-axis order in the reference coordinate system of the dynamic twin panel model to form two columns of axially aligned value pairs.

[0071] The numerical difference extraction unit is configured to subtract the reference point axial value from the corner axial value in each axial numerical pair to obtain the original offset distance value in the axial direction; the subtraction operation is performed on the three axes of X-axis, Y-axis and Z-axis respectively to obtain three independent original offset distance values ​​in the axial direction.

[0072] The sequential binding unit is configured to bind the three original axial offset distance values ​​in the order of X-axis offset distance, Y-axis offset distance, and Z-axis offset distance to form a three-component offset distance group.

[0073] In the above embodiments, the axial alignment unit sequentially places the X, Y, and Z axial values ​​of the corner points and the reference points one-to-one, ensuring that all coordinate data obtain a unique and exclusive spatial interpretation within the same reference coordinate system. The numerical difference extraction unit only performs directional subtraction, eliminating other computational interference, so that the original offset distance value of each axis forms an inseparable legal binding relationship with the corresponding axis. The sequential binding unit then encapsulates the offset distance into a three-component offset distance group according to the X→Y→Z axis sequence, completing the zero-distortion migration of error data from physical space to the legal data structure. The three-component offset distance group is fixed in a predetermined axis sequence and serves as the input for subsequent compensation algorithms or quality traceability programs. The offset distance group realizes the legally uninterrupted transmission of error data from the measurement module to the correction module, ensuring that the spatial pose of electrical cabinet panel components during batch assembly is always within a defined accuracy range.

[0074] Example 9: Based on Example 8, the numerical difference extraction unit provided in this embodiment of the invention includes:

[0075] The separation and temporary storage sub-unit is configured to extract a pair of values ​​from two columns of axially aligned numerical pairs, write the corner axial values ​​of the pair of values ​​into the first numerical temporary storage, and write the reference point axial values ​​of the pair of values ​​into the second numerical temporary storage, forming a pair of temporary data of the compared value and the reference value.

[0076] The scale displacement measurement subunit is configured to simultaneously input the compared value in the first numerical temporary register and the reference value in the second numerical temporary register into the axial displacement measuring instrument, take the position of the reference value on the axial scale as the zero point, read the unidirectional scale offset of the compared value relative to the zero point along the positive and negative directions of the axial direction, and output a scale offset value with directional attributes as the original offset distance value in the axial direction.

[0077] The sequential filling sub-unit is configured to write the output axial original offset distance values ​​into the corresponding positions of the three-component buffer according to the axis number being processed; the X-axis offset distance is written into the first slot of the buffer, the Y-axis offset distance is written into the second slot of the buffer, and the Z-axis offset distance is written into the third slot of the buffer; after the three axes are processed in sequence, the three values ​​in the three-component buffer are output as a three-component offset distance group.

[0078] In the above embodiments, the separation and temporary storage subunit of this embodiment writes the axial values ​​of the corner point and the axial values ​​of the reference point into independent first and second value temporary storages, forming distinguishable comparison values ​​and reference values; physical isolation ensures that the reference point values ​​do not drift due to computational coupling during subsequent measurement. The scale displacement measurement subunit takes the position of the reference point on the axial scale as the zero point, reads the unidirectional scale offset of the comparison value only along the positive and negative directions of the axial direction, and outputs the original offset distance value with directional attributes; only a unique and exclusive scale offset result is generated; the directional attribute of the scale offset value provides a legal basis for positive and negative compensation for subsequent correction actions, ensuring that the assembly position of the panel components is always within the limited allowable error range. The sequential filling sub-unit writes the scale offset values ​​into the corresponding slots of the three-component buffer according to the axis number, forming a three-component offset distance group arranged sequentially along the X, Y, and Z axes. The encapsulation method is explicitly defined in the claims as a fixed sequence and fixed slot mapping, so that the three-component offset distance group has an unchangeable data structure. The overall output of the buffer can be directly called by the subsequent projection decomposition step, realizing the seamless transmission of error data from the measurement unit to the correction unit, avoiding implementation risks caused by data format incompatibility or human intervention.

[0079] Example 10: As Figure 5 As shown, based on Embodiment 1, the real-time comparison feedback subsystem provided in this embodiment of the invention includes:

[0080] The gripper gripping posture generation module is configured to retrieve a single instruction record from the modified assembly instruction set according to the original assembly sequence, and read the component identifier, compensated installation coordinate value, normal vector component at the coordinate, and height offset from the single instruction record; use the normal vector component as the spatial direction that the pneumatic gripper's central axis should point to; use the compensated installation coordinate value as the position point that the gripper's end needs to reach; and use the height offset as the pre-clamping distance between the bottom surface of the component and the gripper's reference surface after the gripper grips the component. The three parameters are combined into a gripping posture parameter set for the component.

[0081] The spatial pose iterative approach module is configured to send the generated grasping posture parameter set to the motion controller of the six-DOF assembly manipulator. The motion controller generates a set of joint angle adjustment sequences based on the deviation between the current gripper's actual spatial pose and the target pose. The manipulator drives each joint of the manipulator sequentially according to the adjustment sequence, so that the gripper carrying the component moves from the current pose to the target pose. After each joint adjustment step is completed, the distance between the gripper reference surface and the panel surface is measured in real time using a laser ranging array mounted on the manipulator base. The distance value is compared with the height offset. If the difference exceeds the preset approach threshold, the joint angle is adjusted until the difference is consistent with the height offset and the angle between the gripper's central axis direction and the normal vector component is less than 0.01 degrees.

[0082] The positioning completion output module is configured to activate the micro-force holding mode of the pneumatic gripper after completing the spatial pose iteration approach, and move the gripper downward at a constant low speed while monitoring the pressure distribution value output by the thin film pressure sensor array on the gripper fingers; when the pressure distribution value jumps from zero to the preset contact threshold and the difference in readings of the four pressure sensors is less than the allowable range, it is determined that the bottom surface of the component has achieved planar contact with the panel surface at the compensation coordinate, outputs a positioning completion signal, and triggers the coaxial structured light projector to start the cross-shaped light pattern projection at the current assembly point.

[0083] In the above embodiments, this embodiment maps the micron-level geometric deviation closed loop into robot pose control commands; through dual feedback of laser ranging and pressure sensing, it ensures that the bottom surface of the component and the actual curved surface of the panel achieve a gapless and zero-stress fit at the compensation coordinates; and triggers visual re-inspection with a legal signal to form a precision control closed loop that runs through the entire assembly process.

[0084] Example 11: As Figure 6 As shown, based on Examples 1-10, the high-precision assembly method for electrical cabinet panel components based on digital twins and vision guidance provided by this invention includes the following steps:

[0085] Step S100: Fix the electrical cabinet panel to the air-floating platform, and use a three-axis laser displacement sensor to scan the front of the panel along a preset path to obtain the surface undulation height value and local tilt angle value at each installation coordinate point. After data cleaning, a three-dimensional topography data array of the panel is obtained. Input the three-dimensional topography data array of the panel into the initial model of the panel in the digital space, so that the normal direction of the corresponding coordinate point in the initial model of the panel is consistent with the micro-tilt direction of the actual panel surface, forming a dynamic twin panel model with a micron-level geometric deviation field.

[0086] Step S200: Extract the deviation field values ​​from the dynamic twin panel model, and apply an offset vector coupled with the surface tilt angle at the theoretical installation center point to each component in the standard installation drawing; obtain the compensated installation coordinates of each component on the actual surface of the panel through iterative convergence, and form a set of corrected assembly instructions corresponding to the deviation field sequence.

[0087] Step S300: The generated set of correction assembly instructions is fed into a six-DOF assembly robot. The pneumatic gripper at the end of the robot grasps the components and positions them one by one according to the compensation coordinates in the instructions. At the same time, the coaxial structured light projector installed at the end of the robot projects a crosshair pattern onto the current assembly point. The high-speed camera captures the image of the relative gap between the bottom edge of the component and the preset reference mark on the panel. The residual deviation value between the actual assembly position and the compensation coordinates is obtained by edge extraction. The residual deviation value is superimposed on the correction assembly instruction calculation process of the next component to form a point-by-point progressive closed-loop correction chain.

[0088] In the above embodiments, during the assembly of electrical cabinet panel components, the panel surface exhibits micron-level geometric errors, including warping, local tilting, and uneven machining allowances. This leads to problems such as assembly stress, poor contact, sealing failure, and appearance defects in traditional rigid positioning. The solution involves high-fidelity replication of the actual panel's microstructure; real-time geometric compensation of theoretical installation coordinates; and point-by-point dynamic closed-loop correction of assembly deviations. This allows for highly reliable, low-stress, and rework-free assembly of components with sub-millimeter-level overall precision.

[0089] This embodiment eliminates clamping deformation through non-contact air-float positioning and acquires the surface undulation height and local tilt angle values ​​of each installation coordinate point on the panel along a preset scanning path; achieving micron-level acquisition of the panel's three-dimensional topography data array, providing realistic geometric input for the twin model. After denoising and interpolation processing of the acquired data, it is mapped to digital space, making the normal direction of the corresponding point in the initial panel model consistent with the actual surface micro-tilt, forming a dynamic twin panel model with a micron-level geometric deviation field; achieving equivalent reproduction of the real panel's geometric errors in digital space. After extracting the deviation field values, an offset vector coupled with local tilt is applied to the theoretical center point of each component in the standard installation drawing, and the compensated installation coordinates are obtained through iterative convergence; realizing the transformation of the rigid values ​​of the theoretical coordinate system into flexible coordinate values ​​that conform to the real panel surface, eliminating positioning system errors caused by surface tilt. The compensated coordinates are arranged into executable instructions according to the deviation field sequence, ensuring that the robot arm performs assembly according to the actual panel surface trajectory; achieving global adaptation between the assembly path and the real panel geometry, avoiding local warping of components or installation stress concentration. The robotic arm completes component positioning according to compensation instructions; coaxial structured light projects crosshairs at the assembly points, and a high-speed camera acquires real-time images of the gap between the bottom edge of the component and the panel reference mark; thus realizing visualization and quantifiable monitoring of the assembly process; the residual deviation between the actual assembly position and the compensation coordinates is calculated through an edge extraction algorithm, and this deviation is superimposed on the compensation calculation of the next component; thus realizing point-by-point progressive closed-loop correction to ensure that the assembly accuracy of subsequent components does not deteriorate due to the accumulation of previous errors.

[0090] In summary, this embodiment reproduces the actual panel geometry at the micrometer scale, eliminating systematic deviations caused by processing errors; it transforms theoretical coordinates into compensated coordinates that fit the actual curved surface, significantly reducing assembly stress and the risk of sealing failure; through point-by-point closed-loop correction, it achieves assembly accuracy without decreasing as the number of components increases, ultimately reaching sub-millimeter-level overall assembly accuracy and highly reliable electrical connections; the entire process is rework-free, low-stress, and low-damage, meeting the engineering requirements for mass production of high-power, high-density electrical cabinets.

[0091] Specific application scenario: Smart factory production line of a well-known electrical equipment manufacturer

[0092] The factory received an order for a batch of high-density server racks for data centers. The electrical cabinet panels, measuring 1.2 meters x 0.6 meters, were manufactured using sheet metal stamping. Due to uneven stress release in the sheet metal during the batch, approximately 15% of the panels exhibited localized slight warping, with the central area rising 0.1-0.3 millimeters above the corners and a slight surface tilt. Under traditional rigid clamp assembly methods, when installing components such as relays and terminals, uneven panel surfaces often result in gaps between the component's bottom surface and the panel, reaching up to 0.5 millimeters. This causes screws to be unable to be tightened, solder joints to become loose, or adjacent components to interfere with each other, resulting in a rework rate as high as 8%.

[0093] System deployment and application process:

[0094] 1. High-precision twin modeling:

[0095] A worker places a panel to be assembled on an air-floating platform, and the system is activated. A three-axis laser displacement sensor scans the entire panel along a pre-set bow-shaped path, acquiring microscopic undulations and tilt data at the theoretical center of all 126 mounting coordinate points, for each relay or terminal. For example, mounting point 58, located at the center of the panel, was detected to be 0.18 mm higher than the reference plane and tilted 0.3 degrees in the X direction. This data is mapped in real time onto the initial CAD model, generating a dynamic twin panel model with a micrometer-level deviation field—the screen clearly shows a slight red raised cloud pattern in the central area of ​​the panel.

[0096] 2. Intelligent coordinate compensation:

[0097] The assembly coordinate reconstruction subsystem reads the deviation field. For the central protruding mounting point, the system calculates the offset vector: on the one hand, the component mounting coordinates need to be raised 0.18 mm upwards along the normal direction; on the other hand, the bottom surface of the component needs to be slightly deflected by an angle based on a 0.3-degree tilt to fit the actual surface. After iterative calculations and checking the gaps between the four corners of the component's bottom surface and the panel, the finally compensated coordinate instructions are generated: the original theoretical coordinates (X=500.00, Y=300.00) are corrected to (X=500.02, Y=299.98, Z=+0.18), with an attached normal vector; the compensation instructions for all 126 components are arranged in sequence to form a corrected assembly instruction set.

[0098] 3. Visually guided closed-loop assembly:

[0099] A six-DOF robotic arm grasps a relay according to instructions, its pneumatic grippers automatically adjusted to be parallel to a local surface of the panel. The arm moves above the compensation coordinates, and a coaxial structured light projector projects a crosshair pattern onto the panel. During the final 0.1 mm bonding process, a high-speed camera captures real-time images of the component's bottom edge and the reference marks printed on the panel. If the vision system detects a 5-micrometer eastward deviation between the actual position and the compensation coordinates, the system immediately adds this residual deviation value to the instruction calculation for the next terminal. When assembling the next terminal, the robotic arm automatically corrects for this 5-micrometer deviation in advance. Assembly pressure sensors ensure the component is bonded with a constant pressure of 10N, preventing overpressure or improper bonding.

[0100] 4. Actual results:

[0101] All 126 components on the panel were assembled and inspected using an online coordinate measuring machine. The maximum gap between the bottom surface of all components and the panel was less than 30 micrometers. Due to the elimination of assembly stress, no signal interruption failures caused by poor contact were observed during subsequent electrical testing. The rework rate of the batch panels decreased from 8% to below 0.5%, and the overall production line efficiency improved by approximately 15%. The dynamic twin model and correction instruction set for each panel were archived for quality traceability – for example, it's possible to trace back to why the actual installation position of relay number 58 was 0.18 mm off from the theoretical value, corresponding to the scanned data of the panel's center protrusion, providing a complete chain of evidence.

[0102] Figure 7 A block diagram of an exemplary electronic device suitable for implementing embodiments of the present invention is shown.

[0103] Electronic devices may include a central processing unit / microprocessor / main control chip; and a storage medium coupled to the central processing unit / microprocessor / main control chip, wherein computer-executable instructions are stored for performing the steps of various methods of embodiments of the present invention when executed by a processor.

[0104] The central processing unit / microprocessor / main control chip may include, but is not limited to, one or more processors or microprocessors.

[0105] Storage media may include, but are not limited to, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, and computer storage media (such as hard disks, floppy disks, solid-state drives, removable disks, CD-ROMs, DVD-ROMs, Blu-ray discs, etc.).

[0106] In addition, the electronic device may include (but is not limited to) a data bus, an input / output bus / external bus / device bus, a display, and input / output devices (e.g., keyboard, mouse, speaker, etc.).

[0107] The central processing unit / microprocessor / main control chip can communicate with external devices via wired or wireless networks (not shown) through input / output buses / external buses / device buses.

[0108] The storage medium may also store at least one computer-executable instruction for performing the steps of various functions and / or methods in the embodiments described herein when the central processing unit / microprocessor / main control chip is running.

[0109] In one embodiment, the at least one computer-executable instruction may also be compiled into or comprise a software product, wherein one or more computer-executable instructions are executed by a processor to perform the steps of the various functions and / or methods in the embodiments described herein.

[0110] Figure 8 A schematic diagram of a computer-readable storage medium according to an embodiment of the present invention is shown.

[0111] like Figure 8 As shown, instructions, such as computer-readable instructions, are stored on a non-transitory computer-readable storage medium. When the computer-readable instructions are executed by a processor, the various methods described above can be performed. The non-transitory computer-readable storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-transitory non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. For example, the non-transitory computer-readable storage medium can be connected to a computing device such as a computer, and then, when the computing device executes the computer-readable instructions stored on the non-transitory computer-readable storage medium, the various methods described above can be performed.

[0112] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or units may be electrical, mechanical, or other forms.

[0113] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0114] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0115] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for executing all or part of the steps of the methods of the various embodiments of this invention through a computer device (which may be a personal computer, server, or network device, etc.). The aforementioned storage medium includes: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0116] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A high-precision assembly system for electrical cabinet panel components based on digital twin and vision guidance, characterized in that, Include: The acquisition and twin mapping subsystem is used to scan the front of the panel using a triaxial laser displacement sensor to obtain the surface undulation height value and local tilt angle value at each installation coordinate point, forming a dynamic twin panel model with a micron-level geometric deviation field. The assembly coordinate reconstruction subsystem is used to extract the deviation field, apply an offset vector coupled with the surface tilt angle at the theoretical installation center point of each component, obtain the compensated installation coordinates through iterative convergence, and generate a corrected assembly instruction set. The real-time comparison feedback subsystem is used for assembling components by a robot arm according to the compensation coordinates. At the same time, it projects crosshairs through coaxial structured light and acquires images of the gap between the bottom edge of the component and the reference mark, extracts the residual deviation value, and superimposes the residual deviation value into the calculation of the correction command for the next component, forming a point-by-point progressive closed-loop correction chain.

2. The high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance as described in claim 1, characterized in that, The data acquisition and twin mapping subsystem is configured to fix the electrical cabinet panel to the air-floating platform, and use a three-axis laser displacement sensor to scan the front of the panel along a preset path to obtain the surface undulation height value and local tilt angle value at each installation coordinate point. After data cleaning, a three-dimensional topography data array of the panel is obtained. The panel three-dimensional topography data array is input into the panel initial model in digital space, so that the normal direction of the corresponding coordinate point in the panel initial model is consistent with the micro-tilt direction of the actual panel surface, forming a dynamic twin panel model with a micron-level geometric deviation field.

3. The high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance as described in claim 1, characterized in that, The assembly coordinate reconstruction subsystem is configured to extract the deviation field values ​​in the dynamic twin panel model, apply an offset vector coupled with the surface tilt angle at the theoretical installation center point to each component in the standard installation drawing, and obtain the compensated installation coordinates of each component on the actual surface of the panel through iterative convergence. The compensated installation coordinates form a set of corrected assembly instructions corresponding to the sequence of deviation fields.

4. The high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance as described in claim 3, characterized in that, The assembly coordinate reconstruction subsystem includes: The offset vector initial value generation component is configured to extract the micron-level geometric deviation field value at the location of each theoretical installation center point in the dynamic twin panel model, and extract the height offset, horizontal tilt component, and vertical tilt component of the location respectively; decompose the height offset along the redirected normal direction of the location to obtain the normal lift; combine the horizontal tilt component and the vertical tilt component to form the tangential plane deflection direction parameter at the location; use the tangential plane deflection direction parameter as the direction reference of the offset vector, use the normal lift as the first part of the magnitude of the offset vector, and use half of the bottom surface outer dimension of the component at the theoretical installation center point multiplied by the sine value of the tangential plane deflection angle as the second part of the magnitude of the offset vector; algebraically add the first part and the second part and multiply by the direction reference to generate the initial value of the offset vector of the theoretical installation center point; The compensation coordinate calculation component is configured to apply the initial value of the offset vector of each generated theoretical installation center point to the coordinates of the theoretical installation center point of the component in the standard installation drawing to obtain the component's trial compensation coordinates; with the trial compensation coordinates as the center, a rectangular area with the same dimensions as the bottom surface of the component is extracted on the dynamic twin panel model, and the redirected normal vectors and spatial coordinates of all grid nodes within the rectangular area are extracted; the vertical distance between the spatial coordinates of the grid nodes at the four corner points of the rectangular area and the theoretical plane of the bottom surface of the component is calculated to obtain four edge gap values; the adjustment and update are repeated until all four edge gap values ​​are less than the preset bonding threshold, and the compensation coordinates obtained from the last update are used as the final compensation installation coordinates of the component on the actual surface of the panel; The modified assembly instruction set output component is configured to extract the final compensated installation coordinates corresponding to each component, and sequentially retrieve the coordinate values ​​according to the original assembly order of the components in the standard installation drawing. Each final compensated installation coordinate is bound to the height direction offset in the micron-level geometric deviation field at the installation coordinate point in the dynamic twin panel model, forming a single instruction record containing the component identifier, compensated installation coordinate value, normal vector component at the coordinate, and height offset. All instruction records are arranged according to the original assembly order to form a modified assembly instruction set corresponding to the deviation field order.

5. The high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance as described in claim 4, characterized in that, The compensation coordinate calculation component includes: The spatial coordinate extraction sub-component is configured to extract a rectangular area with the same dimensions as the bottom surface of the component from the dynamic twin panel model, centered on the trial compensation coordinates. Then, it locates four grid nodes at the four corner points of the rectangular area from all the grid nodes contained in the rectangular area, and extracts the spatial coordinate values ​​of these four grid nodes in the dynamic twin panel model to form a set of spatial coordinates of the four corner points. The vertical distance calculation sub-component is configured to substitute the spatial coordinates of each corner point in the four corner point spatial coordinate sets with the theoretical plane equation of the component bottom surface retrieved from the standard installation drawing, and calculate the vertical distance value from each corner point spatial coordinate to the theoretical plane, thus obtaining four original vertical distance values. The gap value sequence generation sub-component is configured to arrange the four original vertical distance values ​​obtained according to the spatial orientation of the four corner points of the rectangular area, corresponding to the four edge positions of the upper left, upper right, lower right and lower left corners, to form four edge gap values; output the four edge gap values ​​as the basis for judging whether the preset fitting threshold has been reached.

6. The high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance as described in claim 5, characterized in that, The vertical distance calculation subcomponent includes: The planar parameter pairing module is configured to sequentially extract the spatial coordinates of each corner point from the four corner point spatial coordinate sets obtained from the spatial coordinate extraction sub-component; at the same time, it retrieves the spatial orientation parameter set of the theoretical plane of the component's bottom surface from the standard installation drawing; the spatial orientation parameter set contains the normal pointing data of the theoretical plane and the coordinates of the reference points through which the plane passes; and binds each corner point spatial coordinate with the reference point coordinates and normal pointing data in a one-to-one correspondence to form four sets of coordinate and planar parameter pairs; The distance value conversion module is configured to calculate the spatial position deviation vector of the corner point's spatial coordinates relative to the plane reference point, decompose the deviation vector by projection along the direction of the theoretical plane normal, and extract the projection length of the deviation vector in the normal direction; determine the sign of the projection length based on whether the projection direction is in the same or opposite direction to the normal direction, and obtain the directed vertical distance value at the corner point; perform conversion on four sets of coordinate and plane parameter pairs respectively to obtain four directed vertical distance values; The distance sequence arrangement output module is configured to obtain four directional vertical distance values, and arrange them in the order of the top left, top right, bottom right, and bottom left corners when locating the four corner points in the spatial coordinate extraction sub-component, forming four original vertical distance value sequences.

7. The high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance as described in claim 6, characterized in that, Distance value conversion module, including: The spatial offset decomposition submodule is configured to compare the spatial coordinates of the corner points and the reference point coordinates of each set of coordinates in the planar parameter pairing module with the coordinates of the reference point. It measures the offset distance of the corner point relative to the reference point in each axis along the three orthogonal axes of the reference coordinate system of the dynamic twin panel model to obtain the original offset component values ​​of the three axes. The symbol marking submodule is configured to extract the values ​​of the three original axial offset components corresponding to a set of corner points, identify the quadrant of the corner point relative to the reference point based on the spatial orientation order of the rectangular area where the corner point is located, determine the positive or negative sign to be assigned to each axial offset component according to the quadrant, and combine the original offset component values ​​with the corresponding positive or negative signs to form three signed axial offset components. The spatial deviation vector construction submodule is configured to arrange the three signed axial offset components corresponding to a set of corner points according to the axis order of the reference coordinate system, and combine them into a spatial deviation vector with direction and magnitude. The spatial offset decomposition submodule and the symbol marking submodule are executed on the remaining three sets of coordinate and plane parameter pairs respectively. Each completed set generates a corresponding spatial deviation vector, and the four spatial deviation vectors are stored in a temporary buffer for use in the projection decomposition step.

8. The high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance as described in claim 7, characterized in that, The spatial offset decomposition submodule includes: The axial alignment unit is configured to retrieve a set of coordinate and planar parameter pairs from the planar parameter pairing module, and read the three axial values ​​of the corner point spatial coordinates and the three axial values ​​of the reference point coordinates in the planar parameter pair respectively; the axial values ​​of the corner point and the axial values ​​of the reference point are placed in a one-to-one correspondence according to the X-axis, Y-axis and Z-axis order in the reference coordinate system of the dynamic twin panel model to form two columns of axially aligned value pairs. The numerical difference extraction unit is configured to subtract the reference point axial value from the corner axial value in each axial numerical pair to obtain the original offset distance value in the axial direction; the subtraction operation is performed on the three axes of X-axis, Y-axis and Z-axis respectively to obtain three independent original offset distance values ​​in the axial direction. The sequential binding unit is configured to bind the three original axial offset distance values ​​in the order of X-axis offset distance, Y-axis offset distance, and Z-axis offset distance to form a three-component offset distance group.

9. The high-precision assembly system for electrical cabinet panel components based on digital twin and vision guidance as described in claim 1, characterized in that, The real-time comparison and feedback subsystem is configured to send the generated set of correction assembly instructions to the six-DOF assembly robot. The pneumatic gripper at the end of the robot grasps the components and positions them one by one according to the compensation coordinates in the instructions. At the same time, the coaxial structured light projector installed at the end of the robot projects a crosshair pattern onto the current assembly point. The high-speed camera captures the image of the relative gap between the bottom edge of the component and the preset reference mark on the panel. The residual deviation value between the actual assembly position and the compensation coordinates is obtained by edge extraction. The residual deviation value is superimposed on the calculation process of the correction assembly instruction for the next component, forming a point-by-point progressive closed-loop correction chain.

10. A high-precision assembly method for electrical cabinet panel components based on digital twins and vision guidance, used to implement the high-precision assembly system for electrical cabinet panel components based on digital twins and vision guidance as described in any one of claims 1 to 9, characterized in that, include: The electrical cabinet panel is fixed to the air-floating platform. A three-axis laser displacement sensor scans the front of the panel along a preset path to obtain the surface undulation height value and local tilt angle value at each installation coordinate point. After data cleaning, a three-dimensional topography data array of the panel is obtained. The three-dimensional topography data array of the panel is input into the initial model of the panel in digital space so that the normal direction of the corresponding coordinate point in the initial model of the panel is consistent with the micro-tilt direction of the actual panel surface, forming a dynamic twin panel model with a micron-level geometric deviation field. Extract the deviation field values ​​from the dynamic twin panel model, and apply an offset vector coupled with the surface tilt angle at the theoretical installation center point to each component in the standard installation drawing; obtain the compensated installation coordinates of each component on the actual surface of the panel through iterative convergence, and form a set of corrected assembly instructions corresponding to the deviation field sequence. The generated set of corrective assembly instructions is fed into a six-DOF assembly robot. The pneumatic gripper at the end of the robot grasps the components and positions them one by one according to the compensation coordinates in the instructions. At the same time, a coaxial structured light projector installed at the end of the robot projects a crosshair pattern onto the current assembly point. A high-speed camera captures the image of the relative gap between the bottom edge of the component and the preset reference mark on the panel. The residual deviation value between the actual assembly position and the compensation coordinates is obtained by edge extraction. The residual deviation value is superimposed on the corrective assembly instructions of the next component to form a point-by-point progressive closed-loop correction chain.

Images