A method and system for precise positioning control of film welding of a photovoltaic module

Abstract

The application discloses a kind of photovoltaic module film welding precision positioning control method and system, specifically relates to welding positioning technical field, for solving the systematic deviation problem of alignment in prior art due to welding and film coating process control independent, lack of unified reference and battery string physical deformation;It is by obtaining battery string image after welding and extracting topographic feature point set, which is matched with theoretical feature point set to calculate the current deformation field, then the local surface fitting residual and equivalent strain energy density are analyzed and identified and filtered out pseudo-deformation area, obtain reliable physical deformation field, then combine battery string area function information to generate importance weight distribution, compensate reconstruction with the weight as target to film material theoretical positioning coordinates, generate target fitting trajectory under the reference coordinate system of film coating process, finally control executive mechanism runs according to trajectory, realize the high-precision alignment of battery string and film material and fit.

Description

Technical Field

[0001] This invention relates to the field of welding positioning technology, and in particular to a method and system for precise positioning control of photovoltaic module coating welding. Background Technology

[0002] The manufacturing of photovoltaic modules mainly involves two key processes: stringing together solar cells to form a cell string, and laminating the cell string with materials such as upper glass and encapsulant film. Existing automated production lines typically develop and control the welding station and the lamination station as independent modules. The welding station uses its own mechanical coordinate system or vision system as a reference to complete the positioning of the solar cells and the laying and welding of the solder strips. Afterwards, the welded cell strings are transferred to the lamination station by a robotic arm or conveyor device. The lamination station uses its independent positioning system to capture images and calculate the position of the cell strings delivered there, and controls the actuators to align and bond them with the coating material. The positioning control accuracy within each station can already reach a high level under static or ideal conditions.

[0003] However, in actual continuous production, the battery string, as a flexible body, may undergo unpredictable microscopic changes in its physical state compared to its initial state when the welding station is completed, during the welding heat process and the transfer of equipment between different workstations. At the same time, the positioning control systems used in the welding and coating processes are independent of each other, lacking a unified spatial reference and state information transmission mechanism. This leads to a systematic deviation between the battery string position information obtained by the stacking station and its actual physical orientation. This deviation directly affects the alignment accuracy of the subsequent coating and welding processes, restricting the yield and performance improvement of photovoltaic module products. Summary of the Invention

[0004] This invention addresses the technical problems existing in the prior art by providing a method and system for precise positioning and control of photovoltaic module coating welding.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0006] This invention provides the following technical solution:

[0007] A method for precise positioning and control of photovoltaic module coating welding includes:

[0008] S1. After the welding process is completed, obtain the image information of the battery string, and extract the current shape feature point set of the corresponding battery string based on the image information;

[0009] S2. Register and compare the current set of morphological feature points with the pre-stored theoretical set of morphological feature points of the corresponding battery string in the reference coordinate system of the welding process, and calculate the current deformation field distribution of the battery string.

[0010] S3. Perform a joint analysis of the local surface fitting residual and the equivalent strain energy density on the current deformation field distribution to identify and isolate pseudo-deformation regions in the current deformation field distribution, and obtain a reliable physical deformation field distribution.

[0011] S4. Based on the reliable physical deformation field distribution and the preset regional functional information of the battery string, generate the corresponding regional importance weight distribution;

[0012] S5. Using the regional importance weight distribution as the optimization objective, the theoretical positioning coordinates of the coating material are compensated and reconstructed to generate the target bonding trajectory of the battery string in the reference coordinate system of the coating process.

[0013] S6. The actuator of the control coating process operates according to the target bonding trajectory to complete the alignment and bonding of the battery string and the coating material.

[0014] Furthermore, after the welding process is completed, image information of the battery string is acquired, and the current morphological feature point set of the corresponding battery string is extracted based on the image information, including:

[0015] Using an image acquisition device positioned above the transmission path between the welding station and the lamination station, when the battery string reaches a predetermined image acquisition position, at least one surface image of the battery string is acquired as image information.

[0016] Surface images are preprocessed to enhance feature contrast;

[0017] Based on the preprocessed image, the pixel coordinates of multiple key feature points on the battery string are identified and located using a feature extraction algorithm.

[0018] The pixel coordinates of the multiple key feature points located are converted to three-dimensional spatial coordinates in the welding process reference coordinate system according to the calibration parameters of the image acquisition device, thereby forming the current shape feature point set.

[0019] Furthermore, the current set of morphological feature points is registered and compared with the pre-stored theoretical set of morphological feature points of the corresponding battery string in the reference coordinate system of the welding process, and the current deformation field distribution of the battery string is calculated, including:

[0020] Based on multiple key feature points that correspond to each other in the current feature point set and the theoretical feature point set, the optimal spatial coordinate transformation is calculated using the rigid body transformation registration algorithm. The spatial coordinate transformation minimizes the overall deviation between the transformed current feature point set and the theoretical feature point set.

[0021] Using the theoretical morphological feature point set as a reference, for the current morphological feature point set after spatial coordinate transformation, calculate the coordinate offset of each point in the welding process reference coordinate system from the corresponding point in the theoretical morphological feature point set.

[0022] By collecting the coordinate offsets of all feature points, a vector field representing the overall deformation magnitude and direction of the battery string is generated, thus obtaining the current deformation field distribution.

[0023] Furthermore, a synergistic analysis of the local surface fitting residual and equivalent strain energy density is performed on the current deformation field distribution to identify and isolate pseudo-deformation regions in the current deformation field distribution, thereby obtaining a reliable physical deformation field distribution, including:

[0024] Based on the current deformation field distribution, reference surfaces are established in multiple local regions and the degree of deviation between each data point and the corresponding reference surface is calculated.

[0025] Regions whose deviation exceeds a preset offset threshold are marked as geometrically abnormal regions;

[0026] Based on the same current deformation field distribution, calculate the energy density generated by deformation at each location, and analyze the distribution of the corresponding energy density on the entire battery string. Mark the area where the energy density value is outside the normal distribution range as a physical anomaly area.

[0027] Both geometrically abnormal regions and physically abnormal regions are identified as pseudo-deformation regions.

[0028] Remove the data corresponding to all pseudo-deformation regions from the current deformation field distribution, and fill the removed regions with data to finally form a reliable physical deformation field distribution.

[0029] Furthermore, establishing reference surfaces in multiple local regions and calculating the deviation of each data point from the corresponding reference surface includes:

[0030] The battery string is divided into multiple continuous local regions based on the cell boundaries and solder strip positions;

[0031] For each local region, a smooth reference surface is fitted using the spatial coordinates of the data points within it;

[0032] Calculate the spatial coordinates of each data point within the corresponding local area and the Euclidean distance between the corresponding projection point on the reference surface. Record the absolute value of the Euclidean distance as the degree of deviation of the corresponding data point.

[0033] Furthermore, based on the reliable physical deformation field distribution and the preset regional functional information of the battery string, a corresponding regional importance weight distribution is generated, including:

[0034] Analyze the distribution of the credible physical deformation field to determine the actual deformation magnitude of each local region on the battery string;

[0035] Call the pre-stored battery string area function information to obtain the graded definition of the degree of influence of different areas on electrical connection performance on the battery string;

[0036] The actual deformation magnitude is jointly judged by classifying the influence degree of the corresponding region in the functional information of the battery string region; based on the judgment result, a regional importance weight distribution containing different weight values ​​is generated for the entire battery string.

[0037] Furthermore, regions with significant actual deformation and high functional impact are assigned the highest priority weight; regions with small actual deformation or low functional impact are assigned a relatively low priority weight.

[0038] Furthermore, using the regional importance weight distribution as the optimization objective, the theoretical positioning coordinates of the coating material are compensated and reconstructed to generate the target bonding trajectory of the battery string in the reference coordinate system of the coating process, including:

[0039] Establish a mapping relationship between the position of each point on the battery string and the position of the corresponding point in the theoretical positioning coordinates of the coating material;

[0040] Based on the reliable physical deformation field distribution, calculate the spatial deviation of each point on the battery string relative to its theoretical position;

[0041] Guided by the regional importance weight distribution, the calculated spatial deviation is corrected. For points with high regional importance weight, their spatial deviation is given a higher priority in the compensation and reconstruction process.

[0042] Based on the corrected spatial deviation, the theoretical positioning coordinates of the coating material are adjusted, and the target spatial coordinates of each point of the battery string in the reference coordinate system of the coating process are calculated.

[0043] The target spatial coordinates of all points are collected to form a complete battery string target fitting trajectory.

[0044] Furthermore, the actuator controlling the coating process operates according to the target bonding trajectory to complete the alignment and bonding of the battery string with the coating material, including:

[0045] The generated target bonding trajectory data is converted into an instruction format that the control system of the lamination process actuator can recognize;

[0046] Based on the target bonding trajectory data, drive the actuator to move to the initial alignment position in the reference coordinate system of the lamination process;

[0047] After the coating material conveying device conveys the coating material to the predetermined station, the control actuator clamps the battery string and moves according to the path and posture defined by the target bonding trajectory data.

[0048] During the movement of the actuator, the actual position of the battery string is compared with the expected position of the target bonding trajectory in real time, and closed-loop motion control is performed based on the comparison results until the battery string and the coating material are completely aligned and bonded.

[0049] On the other hand, the present invention provides a precise positioning and control system for photovoltaic module coating welding, comprising:

[0050] The image acquisition module is used to acquire image information of the battery string after the welding process is completed, and extract the current shape feature point set of the corresponding battery string based on the image information;

[0051] The deformation calculation module is used to register and compare the current set of morphological feature points with the pre-stored theoretical set of morphological feature points of the corresponding battery string in the reference coordinate system of the welding process, and calculate the current deformation field distribution of the battery string.

[0052] The deformation analysis module is used to perform a joint analysis of the local surface fitting residual and the equivalent deformation energy density on the current deformation field distribution, identify and isolate pseudo-deformation regions in the current deformation field distribution, and obtain a reliable physical deformation field distribution.

[0053] The weight generation module is used to generate the corresponding regional importance weight distribution based on the reliable physical deformation field distribution and the preset regional functional information of the battery string.

[0054] The trajectory generation module is used to compensate and reconstruct the theoretical positioning coordinates of the coating material with the regional importance weight distribution as the optimization objective, and generate the target bonding trajectory of the battery string in the reference coordinate system of the coating process.

[0055] The execution control module is used to control the actuator of the coating process to operate according to the target bonding trajectory, so as to complete the alignment and bonding of the battery string and the coating material.

[0056] The beneficial effects of this invention are:

[0057] 1. By establishing a unified spatial reference and state transfer mechanism between the welding and coating processes, the systematic alignment deviation caused by the independence of the processes and the changes in the physical state of the battery string is effectively solved. The shape of the battery string is acquired and the deformation field is calculated in real time after welding. The positioning basis of the lamination station is improved from independent static images to dynamic data that integrates the welding reference and the actual deformation state. This allows the coating positioning control to truly reflect the current physical form of the battery string, thereby avoiding the bonding error caused by the fragmentation of the coordinate system and the uncertainty of the incoming material state at the root.

[0058] 2. By introducing local deformation analysis and physical energy density criterion, it is possible to intelligently distinguish and isolate pseudo-deformations such as measurement noise and local distortion, ensuring the physical authenticity and reliability of deformation field data. Furthermore, by combining the weight distribution generated by the functional importance of battery string regions, the coating coordinates are specifically compensated and reconstructed, so that the final generated target bonding trajectory not only accurately corresponds to the actual deformation, but also prioritizes the alignment accuracy of key electrical connection areas. This achieves a control leap from passive position following to active shape bonding, significantly improving the overall yield of coating welding and the performance of the module. Attached Figure Description

[0059] Figure 1 This is a flowchart of a precise positioning control method for photovoltaic module coating welding according to the present invention;

[0060] Figure 2 This is a schematic diagram of the structure of a photovoltaic module coating welding precision positioning control system according to the present invention. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Example 1: Figure 1 This invention provides a method for precise positioning and control of photovoltaic module coating welding, comprising:

[0063] S1. After the welding process is completed, obtain the image information of the battery string, and extract the current shape feature point set of the corresponding battery string based on the image information;

[0064] S2. Register and compare the current set of morphological feature points with the pre-stored theoretical set of morphological feature points of the corresponding battery string in the reference coordinate system of the welding process, and calculate the current deformation field distribution of the battery string.

[0065] S3. Perform a joint analysis of the local surface fitting residual and the equivalent strain energy density on the current deformation field distribution to identify and isolate pseudo-deformation regions in the current deformation field distribution, and obtain a reliable physical deformation field distribution.

[0066] S4. Based on the reliable physical deformation field distribution and the preset regional functional information of the battery string, generate the corresponding regional importance weight distribution;

[0067] S5. Using the regional importance weight distribution as the optimization objective, the theoretical positioning coordinates of the coating material are compensated and reconstructed to generate the target bonding trajectory of the battery string in the reference coordinate system of the coating process.

[0068] S6. The actuator of the control coating process operates according to the target bonding trajectory to complete the alignment and bonding of the battery string and the coating material.

[0069] S1. After the welding process is completed, obtain the image information of the battery string, and extract the current morphological feature point set of the corresponding battery string based on the image information. The specific implementation includes:

[0070] After the welding process is completed, the battery string is transported by a conveyor to a predetermined image acquisition position between the welding station and the stacking station. An image acquisition device positioned directly above the conveyor path, triggered by a photoelectric sensor, acquires at least one surface image of the battery string as image information upon detecting its arrival. This image acquisition device is specifically an area-array industrial camera, the resolution of which can be selected according to the required detection accuracy; for example, a 2-megapixel or 5-megapixel model can be used. It is equipped with a ring-shaped LED light source to provide a uniform and shadowless lighting environment, ensuring the overall clarity and consistency of the acquired images and avoiding localized overexposure or insufficient illumination.

[0071] The core purpose of preprocessing the acquired surface images is to enhance the contrast between image features, facilitating accurate subsequent identification. Preprocessing first converts the color image to a grayscale image using a weighted averaging method, for example, using the formula: Gray = 0.299 × R + 0.587 × G + 0.114 × B, where R, G, and B represent the pixel values ​​of the red, green, and blue channels, respectively, and Gray is the calculated grayscale value. Subsequently, a histogram equalization algorithm is used to process the grayscale image, expanding the dynamic range of the image by redistributing the grayscale values ​​of the image pixels, thereby enhancing the overall contrast and making the grayscale differences between the battery cells, solder ribbons, and the background more significant. To further suppress random noise interference that may be introduced during image acquisition and transmission, median filtering is applied to the contrast-enhanced image. The size of the filtering window is preset according to the actual resolution of the image and the minimum size of the features to be retained. For example, for a 2-megapixel image (resolution of 1600 pixels × 1200 pixels), a square filtering window with a size of 3 pixels × 3 pixels or 5 pixels × 5 pixels is often used. The window slides on the image pixel by pixel, and the median value of all pixels in the window is sorted and taken as the new gray value of the center pixel.

[0072] Based on the preprocessed image, a feature extraction algorithm is used to identify and locate the pixel coordinates of several predefined key feature points on the battery string. These key feature points are predefined according to the structural characteristics of the battery string, specifically including the four geometric corners of each battery cell and the geometric center point of the intersection area between each solder strip and the main grid line of the battery cell. The feature extraction process first uses the Canny edge detection operator to extract edge contours from the image. This operator obtains continuous edges with a single pixel width by calculating the image gradient and applying non-maximum suppression and double-threshold hysteresis connections. After obtaining the edge image, Hough transform is used to detect straight line segments, thereby accurately separating and locating the rectangular boundaries of the battery cells and the elongated solder strip areas. For the corner features of the battery cells, at the intersection points of the located boundary lines, the Harris corner detection algorithm is used for sub-pixel level precise localization. This algorithm determines the corner position by calculating the corner response function value of each pixel and finding local maxima. For the intersection features of the solder ribbon and the main gate line, the arithmetic mean of the pixel coordinates of all pixels in each connected region of the solder ribbon, determined by edge detection and morphological operations, is calculated and used as the center pixel coordinate of the intersection point. All identified feature points are output in pixel coordinate form, with the origin defined at the top left corner of the image, the horizontal axis positive to the right, and the vertical axis positive downwards.

[0073] The pixel coordinates of multiple key feature points located are transformed into three-dimensional spatial coordinates in the welding process reference coordinate system according to the calibration parameters of the image acquisition device, thus ultimately forming a set of current morphology feature points to characterize the current actual morphology of the battery string. After the image acquisition device is installed and fixed, it must undergo high-precision visual calibration. This process involves taking multiple images of a high-precision checkerboard calibration board with known physical dimensions (e.g., each square has a side length of 2 mm) in different poses, and using classic camera calibration algorithms such as the Zhang Zhengyou calibration method, solving for the intrinsic parameter matrix and external transformation matrix of the image acquisition device.

[0074] The intrinsic parameter matrix describes the optical imaging geometry within the image acquisition device, including focal length parameters, principal point coordinate parameters, and lens distortion coefficient parameters. The focal length parameter includes independent x-axis and y-axis focal length values, representing the equivalent focal length of each pixel on the image sensor at the actual physical scale along the x and y axes, respectively. The principal point coordinate parameters include the x-axis and y-axis coordinates of the principal point in the image coordinate system, representing the coordinates of the intersection of the lens optical axis and the image sensor plane. The lens distortion coefficient parameters quantify image distortion caused by non-ideal lens imaging, including multiple radial and tangential distortion coefficients. The radial distortion coefficient describes the offset distortion of the imaging point along the radial direction, while the tangential distortion coefficient describes the tangential distortion caused by lens manufacturing and installation errors.

[0075] The external transformation matrix describes the spatial relationship between the image acquisition device coordinate system and the welding process reference coordinate system, namely a 3×3 rotation matrix R and a 3×1 translation vector T. The welding process reference coordinate system is a three-dimensional right-handed Cartesian coordinate system, whose XY plane coincides with the working plane of the welding platform, and whose Z-axis is perpendicular to the platform plane and pointing upwards. The origin of the coordinate system is usually defined at a fixed reference corner point of the platform.

[0076] For the pixel coordinates of each key feature point obtained from feature extraction, distortion correction is first performed using the calibrated intrinsic parameter matrix and distortion coefficients to obtain normalized coordinates under the ideal pinhole model. Then, based on the external transformation matrix (i.e., rotation matrix R and translation vector T) obtained from camera calibration, the three-dimensional spatial coordinates (X, Y, Z) of the point in the welding process reference coordinate system are calculated using the coordinate transformation formula. This coordinate transformation process is repeated for all key feature points, and the resulting set of three-dimensional spatial coordinates constitutes the current morphological feature point set. This set completely represents the actual positional distribution of the key feature points on the surface of the battery string in real three-dimensional space after the welding process.

[0077] S2. Register and compare the current set of morphological feature points with the pre-stored theoretical set of morphological feature points of the corresponding battery string in the reference coordinate system of the welding process, and calculate the current deformation field distribution of the battery string. Specific implementation includes:

[0078] Based on the current morphological feature point set obtained in the previous steps, registration and comparison are performed to calculate the current deformation field distribution. The current morphological feature point set contains the actual three-dimensional spatial coordinates of multiple key feature points on the surface of the battery string after the welding process, in the welding process reference coordinate system. The pre-stored theoretical morphological feature point set comes from the ideal three-dimensional computer-aided design model of this battery string model. This model is generated during the design phase, and the theoretical coordinates of all its key feature points are transformed and stored in the same welding process reference coordinate system through coordinate mapping, ensuring the consistency of the two point sets in the coordinate system and providing a basis for subsequent accurate comparison.

[0079] The registration process first uses a rigid body transformation registration algorithm to calculate an optimal spatial coordinate transformation based on multiple corresponding key feature points in two point sets. Here, "corresponding" means that the feature point pairs in the two point sets represent the same physical location on the battery string, such as the coordinates of the top-left corner of the same battery cell or the specific intersection point of a solder strip and the main busbar. Rigid body transformation registration algorithms, such as the iterative nearest-point algorithm, are used to solve for an optimal rotation matrix and translation vector. The core of this algorithm is to minimize an objective function, which is the sum of the squares of the coordinate deviations of all corresponding point pairs between the current feature point set and the theoretical feature point set after the rotation and translation transformation. During calculation, the current feature point set is used as the set to be transformed, and the theoretical feature point set is used as the target reference set. The algorithm iteratively adjusts the estimates of the rotation matrix and translation vector. Each iteration includes two core steps: finding the nearest point pair and solving for the optimal transformation using the least squares method. This continues until the change in the objective function value between two adjacent iterations is less than a pre-set tolerance threshold (e.g., one ten-thousandth of a millimeter), or until the maximum number of iterations is reached (e.g., one hundred). At this point, the algorithm is considered to have converged, obtaining a set of optimal transformation parameters. This calculation process effectively eliminates the overall rigid displacement and rotation that may occur in the battery string during transmission, thus focusing subsequent deformation analysis on the non-rigid, local deformation of the battery string itself.

[0080] After obtaining the optimal spatial coordinate transformation, the theoretical morphological feature point set is used as an invariant reference. For each feature point in the current morphological feature point set, the spatial coordinate transformation is performed using the rotation matrix and translation vector calculated above, transforming it to a new coordinate system that is optimally aligned with the theoretical morphological feature point set. Subsequently, in the welding process reference coordinate system, the coordinate offset between each point in the transformed current morphological feature point set and its strictly corresponding point in the theoretical morphological feature point set is calculated point by point. This coordinate offset is a three-dimensional vector, whose three components represent the offset values ​​in the X-axis, Y-axis, and Z-axis directions of the welding process reference coordinate system, commonly expressed in millimeters. For a specific feature point, the X-axis component of its coordinate offset is equal to the transformed X-coordinate of that point minus the X-coordinate of its theoretical point; the Y-axis and Z-axis components are calculated in the same way. This vector completely quantifies the magnitude and direction of the point's deviation from its ideal position in space.

[0081] Finally, all calculated feature point coordinate offsets are aggregated. Each feature point's coordinate offset is associated with its unique physical location on the battery string. By organizing and arranging all these three-dimensional vector data according to their corresponding two-dimensional spatial locations on the battery string—for example, storing them in a grid data structure corresponding to the two-dimensional layout of the battery string, where each node stores a three-dimensional offset vector—a vector field is generated that characterizes the overall deformation magnitude and direction of the battery string. This vector field is the final current deformation field distribution, accurately revealing the actual deformation of the battery string relative to its ideal theoretical morphology after the welding process, including the area, magnitude, and spatial trend of the deformation. This provides an indispensable and accurate data foundation for subsequent deformation analysis, pseudo-deformation identification, and compensation control.

[0082] S3. Perform a joint analysis of the local surface fitting residual and equivalent strain energy density on the current deformation field distribution to identify and isolate pseudo-deformation regions in the current deformation field distribution, thereby obtaining a reliable physical deformation field distribution. Specific implementation includes:

[0083] A synergistic analysis of the local surface fitting residuals and equivalent strain energy density is performed on the current deformation field distribution obtained in the previous steps to identify and isolate non-physical anomalous deformation regions, ultimately yielding a reliable physical deformation field distribution. The current deformation field distribution is a vector field containing the three-dimensional coordinate offsets of each feature point on the battery string. Its data organization is a two-dimensional grid data corresponding to the spatial layout of the battery string, with each grid node storing a three-dimensional vector.

[0084] First, based on the current deformation field distribution, reference surfaces are established in multiple local regions, and the deviation of each data point from the corresponding reference surface is calculated. The division of local regions is strictly based on the inherent physical structure characteristics of the battery string, specifically the physical boundary of each individual battery cell and the centerline position of each solder strip. The main area of ​​each battery cell is divided into an independent local region, and each solder strip and its adjacent micro-regions are also divided into an independent strip-shaped local region, thus ensuring that the entire surface of the battery string is divided into multiple continuous and seamlessly connected sets of local regions. For each local region, using the spatial coordinates of all data points within that region, i.e., the X, Y, and Z coordinates of all grid nodes in the current deformation field distribution, a smooth reference surface that represents the overall deformation trend of that region is constructed using a mathematical fitting method. For large and flat areas, such as the middle of a battery cell, a two-dimensional plane equation is fitted using the least squares method; for narrow and potentially curved areas, such as solder strips, a two-dimensional surface is fitted using a polynomial. The essence of the fitting process is to find a surface model that minimizes the sum of squared residuals between the theoretical coordinates of all points on the model and the coordinates of the actual data points within the region. After the reference surface is established, for each data point within the local region, the straight-line distance (Euclidean distance) between its three-dimensional spatial coordinates and its perpendicular projection onto the reference surface is calculated, and the absolute value of this Euclidean distance is precisely recorded as the degree of deviation of the data point relative to the reference surface of its region. This degree of deviation quantifies the difference between the measured position of the point and the overall deformation trend of its local region.

[0085] Subsequently, the calculated deviation of all data points is compared and judged against a pre-set offset threshold. This pre-set offset threshold is not arbitrarily chosen, but rather an empirical value derived from historical data analysis of a large number of known normal battery string samples. For example, it could be three times the average deviation of all samples, or the value corresponding to the 95th percentile based on its probability distribution. For any data point, if its calculated deviation exceeds the pre-set offset threshold, the entire local area containing that point is marked as a geometrically abnormal region. Such anomalies typically originate from mismatches during image feature extraction, measurement noise caused by instantaneous vibrations during transmission, or unrealistic localized severe deformation.

[0086] In parallel, based on the same current deformation field distribution, the elastic strain energy density generated by deformation at each grid node is calculated. This calculation simplifies the material model of the battery string to a linear elastic body. For each point in the current deformation field distribution, the components of the strain tensor at that point are first derived from its coordinate offset vector field. Then, combined with the inherent mechanical property parameters of the battery string material, mainly Young's modulus and Poisson's ratio obtained through standard material tensile tests (these parameters are known constants), the equivalent strain energy density at that point is calculated according to elasticity theory. This value characterizes the energy stored per unit volume of material due to elastic deformation. After the calculation, the statistical distribution of energy density values ​​corresponding to all nodes in the entire battery string is analyzed, such as calculating the arithmetic mean and standard deviation of all energy density values. Regions with energy density values ​​far exceeding the levels that normal physical processes can produce are marked as physical anomaly regions. Specifically, points with energy density values ​​exceeding the average plus five times the standard deviation are identified as anomaly points, and their locations are the physical anomaly regions. Such anomalies usually mean that the deformation data at that location seriously violates the basic laws of physics and mechanics, and are very likely caused by severe distortion of the measurement system, data acquisition outliers, or extreme outliers.

[0087] Next, the geometric and physical anomaly regions marked in the two independent analysis steps above are logically merged and jointly identified as pseudo-deformation regions. The merging rule is a logical OR relation, meaning that if a region is identified as anomaly by any criterion, whether it is a geometric or physical anomaly, it is ultimately classified as a pseudo-deformation region that needs to be removed.

[0088] Finally, data cleaning and spatial data reconstruction are performed. All data points identified as pseudo-deformation regions are precisely removed from the current deformation field distribution dataset; that is, the 3D vector data of these location nodes are set to null or missing values, leaving data gaps in the deformation field. These gaps are then filled to restore the spatial continuity of the deformation field, thus forming a complete and reliable physical deformation field distribution. Data filling is performed using spatial interpolation algorithms, such as inverse distance weighted interpolation. This algorithm utilizes the values ​​of all valid data points (i.e., normal data points not marked as pseudo-deformation) within a certain neighborhood around the gap, assigning different weights based on their distance from the point to be interpolated (closer points have higher weights). The filled value is then calculated through a weighted average. Through this series of rigorous operations, a reliable physical deformation field distribution is obtained, which largely eliminates measurement noise and pseudo-deformation signal interference. This distribution more realistically and reliably reflects the actual physical deformation of the battery string caused by the welding process, providing a high-quality data foundation for subsequent calculations of regional importance weight distribution and generation of target fitting trajectories.

[0089] S4. Based on the reliable physical deformation field distribution and the preset regional functional information of the battery string, generate the corresponding regional importance weight distribution. Specific implementation includes:

[0090] Based on the reliable physical deformation field distribution obtained in the previous steps, and combined with the pre-stored functional information of the battery string regions, a corresponding regional importance weight distribution is generated. The reliable physical deformation field distribution is a vector field in the form of a two-dimensional matrix, where each grid node stores a three-dimensional vector, representing the true physical deformation offset of the corresponding point on the battery string after pseudo-deformation removal. This distribution is the direct output result after performing pseudo-deformation identification and data cleaning.

[0091] First, the distribution of the reliable physical deformation field is analyzed to determine the actual deformation magnitude of each local region on the battery string. The definition of a local region here is completely consistent with the regions previously defined during pseudo-deformation identification; that is, a continuous region divided according to the physical boundaries of the battery cells and the positions of the solder ribbons, such as the main body region of each individual battery cell and the strip region containing each solder ribbon. For each local region, all grid nodes within that region are traversed, and the 3D coordinate offset vector of each node is read. The magnitude of this vector, i.e., the Euclidean distance, is calculated. This magnitude represents the single-point deformation amplitude of the node, in millimeters. Subsequently, the arithmetic mean of the magnitude values ​​of all nodes within the local region is calculated, and this average value is used as a quantitative indicator of the overall deformation magnitude of the local region. By processing all local regions one by one and performing this calculation, a data list is finally obtained. Each entry in this list contains an identifier for a local region and its corresponding average deformation magnitude.

[0092] Subsequently, pre-stored functional information for different areas of the battery string is retrieved. This information is structured data pre-defined and stored in a database or configuration file during the product design phase, based on the electrical connection principles of the battery string, vulnerability analysis of the current transmission path, and long-term reliability experimental data. The core content of the functional information for different areas of the battery string is the hierarchical definition of the impact of different physical areas on the overall electrical connection performance and mechanical connection reliability of the module. This hierarchical definition is typically based on the electrical function and mechanical role of the area. For example, the overlapping welding area between the edge of the cell and the interconnecting strip, because it undertakes the main current transmission and mechanical connection functions, is usually defined as having the highest functional impact level; the central area of ​​the cell mainly serves a power generation function and has a relatively low impact on connection reliability, and may be defined as having a medium or low functional impact level. Each local area has a unique identifier and a corresponding functional impact level label in the pre-stored information. This level is usually represented by ordered discrete numerical values, such as 1, 2, 3, 4, 5, representing functional importance from lowest to highest.

[0093] Next, the actual deformation magnitude of each local area is jointly assessed with its corresponding functional impact level to assign a specific priority weight value to each area. This joint assessment process is based on a pre-defined decision rule or mapping relationship grounded in engineering experience. The essence of this rule is to comprehensively consider both the severity of deformation and the criticality of function. Specifically, for a local area, if its calculated actual deformation magnitude exceeds a predetermined deformation threshold (determined based on the statistical distribution of deformation data from historical qualified products, for example, twice the average deformation value of historical data), and its functional impact level is the highest (e.g., level 5), then this area is determined to be the highest priority compensation area and assigned the highest weight value, for example, a weight value of 1.0. If the actual deformation magnitude of an area is very small, below another lower deformation threshold (e.g., the average deformation value of historical data), and its functional impact level is the lowest (e.g., level 1), then this area is determined to be the least important and assigned the lowest weight value, for example, 0.1. For various other combinations of deformation magnitude and functional impact level, an intermediate weight value, such as 0.3, 0.5, or 0.7, is assigned linearly or non-linearly between the highest and lowest values ​​based on their relative importance. This mapping process can be implemented using a lookup table or a simple piecewise function.

[0094] Finally, based on the above judgments and mapping results, a regional importance weight distribution is generated for the entire battery string. This distribution is a two-dimensional scalar field with the exact same grid size and spatial location mapping relationship as the reliable physical deformation field distribution. During generation, each grid node is assigned the priority weight value allocated to its local region. Therefore, all nodes belonging to the same local region will have the same weight value, while nodes in different regions will have different weight values ​​according to the importance of their regions. The final generated regional importance weight distribution is output in the form of a two-dimensional data matrix, where each element is a dimensionless scalar weight value, typically ranging from 0 to 1. A higher value indicates that the location should be given higher priority in subsequent compensation.

[0095] S5. Using the regional importance weight distribution as the optimization objective, the theoretical positioning coordinates of the coating material are compensated and reconstructed to generate the target bonding trajectory of the battery string in the reference coordinate system of the coating process. Specific implementation includes:

[0096] Using the reliable physical deformation field distribution and regional importance weight distribution generated in the previous steps as core input data, the theoretical positioning coordinates of the coating material are compensated and reconstructed, ultimately generating the target bonding trajectory of the battery string in the reference coordinate system of the coating process. The reliable physical deformation field distribution accurately describes the real physical deformation offset vector of each point in the battery string, while the regional importance weight distribution quantifies the priority that different regions should be considered in the compensation process. Both are direct outputs of the previous steps.

[0097] First, a precise mapping relationship is established between the physical positions of each point on the battery string and the corresponding points in the theoretical positioning coordinates of the coating material. The theoretical positioning coordinates of the coating material originate from the ideal three-dimensional computer-aided design model determined during the design phase of this photovoltaic module model. This model defines the three-dimensional coordinates of each point on the coating material that needs to be aligned with a specific position on the battery string under ideal conditions without any deformation. These coordinates are also defined in the reference coordinate system of the coating process. The mapping relationship is established based on the design drawings and assembly relationship of the battery string and the coating material, typically represented by a one-to-one point-to-point mapping table. This table stores the association between the unique identifier of each feature point on the battery string and the unique identifier of its corresponding target point in the theoretical positioning coordinates of the coating material, ensuring that each feature point on the battery string can find its unique corresponding target point in the theoretical positioning coordinates of the coating material. The reference coordinate system for the coating process is a three-dimensional right-handed Cartesian coordinate system based on the plane of the coating equipment's worktable. Its origin is usually defined at a fixed reference angle on the table surface, and the coordinate axis directions are calibrated and fixed during equipment installation and commissioning using precision measuring instruments such as laser trackers.

[0098] Subsequently, based on the reliable physical deformation field distribution, the spatial deviation of each point on the battery string relative to its theoretical position is calculated point by point. The theoretical position of any point on the battery string is the coordinate of its ideal 3D computer-aided design model in the welding process reference coordinate system. For a specific point on the battery string, the 3D coordinate offset vector stored at that point is directly read from the corresponding reliable physical deformation field distribution data. This vector represents the spatial deviation of the point's current actual position relative to its theoretical position in the welding process reference coordinate system. Its three components represent the offsets in the X, Y, and Z directions, respectively, in millimeters. Since the welding process reference coordinate system and the coating process reference coordinate system are two independent coordinate systems, there is a fixed relative pose relationship between them. This relationship is obtained through the spatial calibration of the equipment and is usually described by a 3x3 rotation matrix and a 3x1 translation vector. It is necessary to transform the vector representing the spatial deviation from the welding process reference coordinate system to the coating process reference coordinate system through this known coordinate system transformation to ensure that all subsequent calculations are performed in a unified coating process reference coordinate system. The coordinate system transformation of the vector is achieved by left-multiplying the vector by the rotation matrix.

[0099] Next, guided by the regional importance weight distribution, the calculated original spatial deviation vector is corrected. The regional importance weight distribution is a two-dimensional scalar field with the same grid size as the reliable physical deformation field distribution. Each grid point has a weight coefficient between 0 and 1, with higher values ​​indicating greater importance. The correction process is performed point-by-point: for a point on the battery string, its corresponding weight coefficient is first read from the regional importance weight distribution. Then, each component of the original spatial deviation vector of that point in the coating process reference coordinate system—the X, Y, and Z direction components—is multiplied by this weight coefficient. The technical effect is that for points with high regional importance weights, such as a weight coefficient of 0.9, their calculated spatial deviation vectors are almost entirely adopted in subsequent compensation; for points with low weights, such as a weight coefficient of 0.3, the influence of their spatial deviation vectors is partially weakened. This step is the core of achieving differentiated and precise compensation, ensuring that the compensation behavior focuses more on key areas that have a greater impact on the final performance of the component.

[0100] Then, based on the corrected spatial deviation, the theoretical positioning coordinates of the coating material are adjusted, and the target spatial coordinates of each point in the coating process reference coordinate system are calculated. For each target point in the theoretical positioning coordinates of the coating material, the corresponding point on the battery string is found through the previously established mapping relationship, and the spatial deviation vector of that point after weight correction is obtained. The calculation of the target spatial coordinates is completed by vector addition: the theoretical coordinates of the point on the battery string (which need to be transformed from the welding process reference coordinate system to the coating process reference coordinate system in advance through coordinate system transformation) are added to the corrected spatial deviation vector. Specifically, the X component of the target spatial coordinates is equal to the X component of the theoretical coordinates of the point on the battery string plus the X component of the corrected deviation vector, and the calculation of the Y and Z components is similar. The coordinates obtained by this calculation are the target positions that enable the point on the coating material to achieve optimal fit with the corresponding point on the battery string with the current actual deformation.

[0101] Finally, all calculated target spatial coordinates are collected and organized according to their spatial order and topology on the battery string, forming a complete, continuous, and precise three-dimensional spatial path. This path is the final target bonding trajectory for the battery string. This trajectory is output as a series of ordered three-dimensional coordinate point sequences, clearly defining the sequence of spatial positions that the coating process actuator needs to reach and pass through during operation, thereby precisely guiding the battery string and coating material to complete high-precision alignment and bonding.

[0102] S6. The actuator of the coating process is controlled to operate according to the target bonding trajectory to complete the alignment and bonding of the battery string and the coating material. Specific implementation includes:

[0103] Based on the target bonding trajectory data generated in the previous steps, the actuators of the coating process are controlled to operate along a predetermined path, ultimately achieving precise alignment and bonding of the battery string and the coating material. The target bonding trajectory data is a series of ordered three-dimensional spatial coordinate points, defining the precise path and orientation that the actuators need to guide the battery string through in the reference coordinate system of the coating process. This data is the direct output of the preceding coordinate compensation and reconstruction steps.

[0104] First, the generated target bonding trajectory data is converted into an instruction format that the control system of the laminating process actuator can directly recognize and execute. The actuator typically consists of a multi-axis motion platform, such as linear motion units along the X, Y, and Z axes, and an Rz axis rotating around the Z axis. The instruction format received by its control system depends on the specific controller model and the industrial fieldbus communication protocol used. The conversion process first parses each three-dimensional coordinate point and its associated attitude information in the target bonding trajectory data. Then, based on the motion range, accuracy limits, and dynamic constraints of each motion axis of the actuator, the spatial coordinates and attitude angles of each target point are mapped to the target position instructions for each motion axis. The motion range constraint refers to the maximum physical travel of each axis, and the accuracy limit refers to the minimum movement unit of each axis servo system, such as 1 micrometer. Simultaneously, based on the distance between adjacent points and the preset process cycle requirements, a motion planning algorithm calculates reasonable motion parameters for each path segment. For example, an S-shaped acceleration / deceleration curve planning algorithm is used to calculate the expected velocity, acceleration, and jerk values ​​for each displacement segment, and these parameters are converted into corresponding speed control instructions. Ultimately, all these position and speed commands are encapsulated according to the specific data structures and communication protocols required by the controller. For example, using the EtherCAT bus protocol, the command data is packaged into a data frame format that conforms to the CiA 402 driver specifications, forming a complete sequence of commands that can be directly parsed and executed by the motion controller.

[0105] Next, based on the converted control command sequence, the actuator is driven to move to the initial alignment position in the reference coordinate system of the lamination process. The initial alignment position is the first coordinate point in the target bonding trajectory data, ensuring that the actuator is spatially pre-aligned with the lamination material to be delivered by the lamination material conveying device before the lamination operation begins. During the drive process, the motion controller's task scheduler sends position commands to the servo drives of each axis periodically (e.g., every 1 millisecond). After receiving the commands, the servo drives run a three-loop control algorithm of position loop, speed loop, and current loop through their built-in digital signal processor. The position loop calculates the speed command based on the commanded position and the actual position fed back by the encoder. The speed loop then calculates the torque command based on the speed command and the feedback speed. The current loop finally outputs the corresponding current to control the torque and speed of the servo motor. The motor rotation drives the ball screw pair through the coupling or directly through the electromagnetic effect of the linear motor, transmitting the rotational or linear motion to the end effector of the actuator, ultimately bringing it to a smooth and accurate stop at the initial alignment position. The positioning accuracy is guaranteed by the feedback of the grating ruler, for example, up to ±5 micrometers.

[0106] After the coating material conveying device transports and precisely positions the coating material to the predetermined station via a conveyor belt or servo-driven robotic arm, the end effector of the control mechanism reliably clamps the battery string. The end effector can be a suction cup assembly with a vacuum generator or a precision mechanical clamp driven by a servo motor. The clamping action is triggered by a digital signal from a programmable logic controller (PLC), such as outputting a 24-volt DC signal to control a solenoid valve to open the vacuum circuit or drive the clamp cylinder. The clamping force setting is crucial; its value must be determined through calculation and experimentation to overcome the weight of the battery string and the inertial forces during movement, while also avoiding excessive force that could cause microcracks in the battery cells. For example, a force sensor can be used for calibration to set the clamping force within a safe range, such as between 20 and 50 Newtons. The clamping status is monitored and confirmed in real time by a vacuum pressure sensor or a displacement sensor on the clamp. After clamping is complete, the actuator begins to move along a continuous path and posture defined by the target adhesion trajectory data. During the motion, the motion controller analyzes the next trajectory command in real time and coordinates the motion of each motion axis through inverse kinematics calculation to ensure that the end effector and the battery string it holds can smoothly and accurately reproduce the target fitting trajectory, and the synchronization error between each axis needs to be controlled within the microsecond level.

[0107] Throughout the entire motion of the actuator, high-precision position sensors integrated on the actuator continuously monitor the actual spatial position of the end effector. Sensors used to detect linear displacement can be optical or magnetic scales with a resolution of up to 0.1 micrometers; encoders used to detect angles have a resolution of 24 bits or higher. These sensors feed the position data back to the motion controller in real time via the same fieldbus (such as EtherCAT). In each control cycle, the controller compares the real-time detected position with the desired position of the target trajectory at the current moment, calculating the position deviation in all three directions. This deviation value is input as a feedback signal into the closed-loop motion control algorithm, such as a proportional-integral-derivative (PI-DI) control algorithm. The proportional gain, integral time, and derivative time need to be tuned according to the inertia and load of the actuator to ensure system stability and rapid response. Based on the magnitude and trend of the deviation, the control algorithm calculates the correction control signal in real time and outputs it to the servo drive, dynamically adjusting the torque and speed of each axis motor, thereby suppressing interference, eliminating tracking errors, and ensuring that the actual motion trajectory always closely follows the target trajectory. This real-time comparison and adjustment closed-loop control process continues until the battery string is guided to its final position on the trajectory, achieving complete alignment and bonding with the coating material stationary at the predetermined workstation. The criterion for successful alignment is that the positional deviation remains below the system's allowable alignment accuracy threshold for a sustained period, for example, less than 0.1 mm for 100 milliseconds. Once bonding is complete, the controller issues a command to release the clamp and controls the actuator to return to the standby position, awaiting the next work cycle.

[0108] Example 2: Figure 2 A schematic diagram of a photovoltaic module coating welding precision positioning control system according to the present invention is provided. The photovoltaic module coating welding precision positioning control system includes:

[0109] The image acquisition module is used to acquire image information of the battery string after the welding process is completed, and extract the current shape feature point set of the corresponding battery string based on the image information;

[0110] The deformation calculation module is used to register and compare the current set of morphological feature points with the pre-stored theoretical set of morphological feature points of the corresponding battery string in the reference coordinate system of the welding process, and calculate the current deformation field distribution of the battery string.

[0111] The deformation analysis module is used to perform a joint analysis of the local surface fitting residual and the equivalent deformation energy density on the current deformation field distribution, identify and isolate pseudo-deformation regions in the current deformation field distribution, and obtain a reliable physical deformation field distribution.

[0112] The weight generation module is used to generate the corresponding regional importance weight distribution based on the reliable physical deformation field distribution and the preset regional functional information of the battery string.

[0113] The trajectory generation module is used to compensate and reconstruct the theoretical positioning coordinates of the coating material with the regional importance weight distribution as the optimization objective, and generate the target bonding trajectory of the battery string in the reference coordinate system of the coating process.

[0114] The execution control module is used to control the actuator of the coating process to operate according to the target bonding trajectory, so as to complete the alignment and bonding of the battery string and the coating material.

[0115] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0116] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0117] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission; wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission includes infrared, microwave, etc. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center containing one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0118] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0119] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules 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 apparatuses or modules may be electrical, mechanical, or other forms.

[0120] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0121] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0122] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion 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 to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0123] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0124] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for precise positioning control of film welding of a photovoltaic module, characterized in that, Comprise: S1, after the welding process is completed, the image information of the battery string is acquired, and the current topographic feature point set of the corresponding battery string is extracted based on the image information; S2, the current topographic feature point set is compared with the pre-stored theoretical topographic feature point set of the corresponding battery string in the welding process reference coordinate system, and the current deformation field distribution of the battery string is calculated; S3, the local surface fitting residual and equivalent strain energy density of the current deformation field distribution are analyzed cooperatively, the false deformation area in the current deformation field distribution is identified and isolated, and the reliable physical deformation field distribution is obtained; S4, based on the reliable physical deformation field distribution and the preset regional function information of the battery string, the corresponding regional importance weight distribution is generated; S5, taking the regional importance weight distribution as the optimization target, the theoretical positioning coordinates of the film coating material are compensated and reconstructed, and the target fitting track of the battery string in the film coating process reference coordinate system is generated; S6, the execution mechanism of the film coating process is controlled to operate according to the target fitting track, and the alignment and fitting of the battery string and the film coating material are completed.

2. The method of claim 1, wherein the method further comprises: After the welding process is completed, the image information of the battery string is acquired, and the current topographic feature point set of the corresponding battery string is extracted based on the image information, comprising: Using the image acquisition device arranged above the transmission path between the welding station and the lamination station, when the battery string reaches the predetermined image acquisition position, at least one surface image of the battery string is collected as image information; The surface image is preprocessed to enhance the feature contrast; Based on the preprocessed image, the pixel coordinates of the pre-set multiple key feature points on the battery string are identified and located by a feature extraction algorithm; The pixel coordinates of the located multiple key feature points are converted to three-dimensional space coordinates in the welding process reference coordinate system according to the calibration parameters of the image acquisition device, thereby forming the current topographic feature point set.

3. The method of claim 1, wherein the method further comprises: The current topographic feature point set is compared with the pre-stored theoretical topographic feature point set of the corresponding battery string in the welding process reference coordinate system, and the current deformation field distribution of the battery string is calculated, comprising: Based on the multiple key feature points corresponding to each other in the current topographic feature point set and the theoretical topographic feature point set, the optimal spatial coordinate transformation is calculated by using rigid body transformation registration algorithm, and the spatial coordinate transformation makes the overall deviation between the transformed current topographic feature point set and the theoretical topographic feature point set minimum; Taking the theoretical topographic feature point set as the reference, the coordinate offset of each point in the current topographic feature point set after spatial coordinate transformation from the corresponding point in the theoretical topographic feature point set in the welding process reference coordinate system is calculated; Collecting the coordinate offsets of all feature points, a vector field representing the overall deformation size and direction of the battery string is generated, that is, the current deformation field distribution is obtained.

4. The method of claim 1, wherein the method further comprises: The local surface fitting residual and equivalent strain energy density of the current deformation field distribution are analyzed cooperatively, the false deformation area in the current deformation field distribution is identified and isolated, and the reliable physical deformation field distribution is obtained, comprising: Based on the current deformation field distribution, reference surfaces are established in multiple local areas and the deviation of each data point from the corresponding reference surface is calculated; The area with deviation exceeding the preset deviation threshold is marked as a geometric abnormal area; Based on the same current deformation field distribution, the energy density generated by the deformation at each position is calculated, and the distribution of the corresponding energy density on the entire battery string is analyzed. The area outside the normal distribution range of the energy density value is marked as a physical abnormal area; The geometric abnormal area and the physical abnormal area are jointly determined as a pseudo-deformation area; Remove the data corresponding to all pseudo-deformation areas from the current deformation field distribution, and perform data filling on the removed area to finally form a reliable physical deformation field distribution.

5. The method of claim 4, wherein the method further comprises: The reference surface is established in each local area, and the deviation of each data point from the corresponding reference surface is calculated, including: According to the battery piece boundary and the solder strip position, the battery string is divided into multiple continuous local areas; For each local area, a smooth reference surface is fitted using the spatial coordinates of the data points within it; Calculate the Euclidean distance between the spatial coordinates of each data point in the corresponding local area and the corresponding projection point on the reference surface, and record the absolute value of the Euclidean distance as the deviation of the corresponding data point.

6. The method of claim 1, wherein the method further comprises: Based on the reliable physical deformation field distribution and the preset regional function information of the battery string, the corresponding regional importance weight distribution is generated, including: Analyze the reliable physical deformation field distribution to determine the actual deformation size of each local area on the battery string; Call the pre-stored regional function information of the battery string to obtain the classification definition of the influence degree of different areas on the electrical connection performance of the battery string; Jointly judge the actual deformation size and the influence degree classification of the corresponding area in the battery string regional function information; according to the judgment result, generate a regional importance weight distribution containing different weight values for the battery string as a whole.

7. The method of claim 6, wherein the method further comprises: For areas with large actual deformation and high functional influence, the highest priority weight is given; for areas with small actual deformation or low functional influence, a relatively low priority weight is given.

8. The method of claim 1, wherein the method further comprises: Taking the regional importance weight distribution as the optimization target, the theoretical positioning coordinates of the film material are compensated and reconstructed to generate the target fitting trajectory of the battery string in the film process reference coordinate system, including: Map the position of each point on the battery string to the position of the corresponding point in the theoretical positioning coordinates of the film material; According to the reliable physical deformation field distribution, calculate the spatial deviation of each point on the battery string relative to its theoretical position; Guided by the regional importance weight distribution, correct the calculated spatial deviation. For points with high regional importance weight, their spatial deviation is given a higher priority in the compensation and reconstruction process; Based on the corrected spatial deviation, adjust the theoretical positioning coordinates of the film material to calculate the target spatial coordinates of each point of the battery string in the film process reference coordinate system; Collect the target spatial coordinates of all points to form a complete battery string target fitting trajectory.

9. The method of claim 1, wherein the method further comprises: Control the execution mechanism of the film process to operate according to the target fitting trajectory to complete the alignment and fitting of the battery string and the film material, including: Convert the generated target fitting trajectory data into a command format that can be recognized by the film process execution mechanism control system; According to the target fitting trajectory data, drive the execution mechanism to move to the initial alignment position in the film process reference coordinate system; After the film material conveying device conveys the film material to the predetermined work station, the control execution mechanism clamps the battery string, and moves according to the path and posture defined by the target bonding trajectory data; During the movement of the execution mechanism, the actual position of the battery string is compared with the expected position of the target bonding trajectory in real time, and closed-loop motion control is performed according to the comparison result until the battery string is completely aligned and bonded with the film material.

10. A photovoltaic module laminated soldering precision positioning control system for implementing the photovoltaic module laminated soldering precision positioning control method of any one of claims 1-9, characterized in that, Comprise: An image acquisition module is configured to acquire image information of the battery string after the welding process is completed, and to extract a current topographic feature point set of the corresponding battery string based on the image information; A deformation calculation module is configured to register and compare the current topographic feature point set with a pre-stored theoretical topographic feature point set of the corresponding battery string in a welding process reference coordinate system, and to calculate a current deformation field distribution of the battery string; A deformation analysis module is configured to perform collaborative analysis on the local surface fitting residual and equivalent strain energy density of the current deformation field distribution, identify and isolate the pseudo-deformation region in the current deformation field distribution, and obtain a reliable physical deformation field distribution; A weight generation module is configured to generate a corresponding regional importance weight distribution based on the reliable physical deformation field distribution and the preset regional function information of the battery string; A trajectory generation module is configured to compensate and reconstruct the theoretical positioning coordinates of the film material with the regional importance weight distribution as the optimization target, and to generate a target bonding trajectory of the battery string in a film coating process reference coordinate system; An execution control module is configured to control the execution mechanism of the film coating process to operate according to the target bonding trajectory, and to complete the alignment and bonding of the battery string and the film material.

Images