A method and device for bonding a strain gauge to a wind turbine blade

Abstract

This application discloses a method and apparatus for bonding strain gauges to wind turbine blades. The main component is a device for bonding strain gauges to wind turbine blades. Utilizing machine vision technology, the target area on the specimen is precisely located, eliminating subjective errors caused by manual visual inspection and hand-drawing. The adhesive thickness and size are precisely controlled in the area to be coated on the target strain gauge. Simultaneously, an adaptive adhesive application trajectory effectively avoids the problem of inaccurate adhesive application by manual application, improving adhesive uniformity and bonding reliability, thereby enhancing the accuracy of strain test data. The coated strain gauge is then bonded to the target area, resulting in a bonded strain gauge. The method of this application is fully automated, reducing reliance on operator experience and improving bonding efficiency. It is suitable for batch and large-scale strain gauge bonding scenarios, achieving standardized and repeatable operation results.

Description

Technical Field

[0001] This application relates to the field of automated operation technology, and in particular to a method and apparatus for attaching strain gauges to wind turbine blades. Background Technology

[0002] In the structural mechanical performance testing of wind turbine blades, resistance strain gauges are widely used due to their simple principle and convenient operation. However, they need to be precisely aligned with the direction of the strain to be measured and the adhesive layer needs to be uniform and controllable in order to meet the requirements of strain transfer efficiency and measurement data reliability. Therefore, the testing accuracy of resistance strain gauges is highly dependent on the bonding quality.

[0003] Existing technologies employ methods such as using masks for strain gauge positioning windows and auxiliary positioning devices with scribe lines and arrows to repeatedly position strain gauges on multiple blades of the same shape. However, these methods heavily rely on manual operation, and are generally prone to problems due to subjective errors in visual positioning and scribe lines, leading to inaccurate alignment between the strain gauge axis and the direction of the strain being measured. Furthermore, the amount, thickness, and uniformity of adhesive applied manually are difficult to control, easily resulting in poor adhesion, excessively thin or thick adhesive layers, and air bubbles, affecting strain transfer efficiency and introducing creep errors. These problems directly reduce the reliability of measurement data. In addition, manual operation is slow, highly dependent on operator experience, and makes it difficult to achieve high-quality, high-volume, repeatable, standardized operations. Summary of the Invention

[0004] To address the aforementioned issues, this application provides a method and apparatus for bonding strain gauges to wind turbine blades, thereby improving the quality and efficiency of strain gauge bonding operations.

[0005] This application discloses a method for attaching strain gauges to wind turbine blades, wherein the executing entity is a device for attaching strain gauges to wind turbine blades, and the method includes: Locate the target area on the specimen; A glue layer of preset thickness and preset area is applied to the target strain gauge area using an adaptive trajectory to obtain the glue-coated strain gauge. The adhesive-coated strain gauge is then attached to the target area to obtain the attached strain gauge.

[0006] Optionally, the target area on the positioning specimen includes: Feature extraction is performed on the region image to obtain the region image feature points; the region image is a camera image of the area to be pasted on the specimen; Calculate the transformation matrix from the camera coordinate system to the region image coordinate system; the region image coordinate system is obtained by calibrating the region image. Based on the transformation matrix, the regional image feature points, and the model feature points, a coarsely located target region is obtained; the model feature points are from the area to be pasted in the digital model of the specimen. Edge detection and fitting are performed on the coarsely located target region to obtain the target region.

[0007] Optionally, obtaining the coarsely localized target region based on the transformation matrix, the region image feature points, and the model feature points includes: The feature points of the region image are matched with the feature points of the model to obtain feature point matching pairs; Based on the transformation matrix, the reprojection error of the feature point matching pair is minimized to obtain the coarsely localized target region.

[0008] Optionally, the step of performing edge detection and fitting on the coarsely located target region to obtain the target region includes: The region of interest is extracted from the coarsely located target region to obtain the region of interest; Edge detection and Gaussian fitting are performed on the region of interest to obtain the sub-pixel positions of the edges of the region of interest; The target region is located based on the sub-pixel position.

[0009] Optionally, the step of applying an adhesive layer of preset thickness and preset area to the target strain gauge area using an adaptive trajectory to obtain the adhesive-coated strain gauge includes: Position the area to be coated with adhesive; Obtain the adhesive coating parameters of the target strain gauge; the adhesive coating parameters include the size of the target strain gauge, the type of adhesive, and the amount of adhesive. Plan the adhesive application path based on the adhesive application parameters; The adhesive layer is applied according to the described adhesive application path to obtain the adhesive-coated strain gauge.

[0010] Optionally, the step of planning the adhesive application path based on the adhesive application parameters includes: If the size is less than or equal to the first size, determine that the target strain gauge is to be applied with a dot matrix adhesive and generate dot matrix coordinates; When the size is greater than the first size and less than the second size, it is determined that the target strain gauge will be coated with adhesive in a linear reciprocating scanning manner, and a scanning path and speed will be generated; the second size is greater than the first size. If the dimension is greater than or equal to the second dimension, it is determined that the target strain gauge will be contour-filled with adhesive, and a filling path will be generated.

[0011] Optionally, after obtaining the strain gauges after pasting, the method further includes: Feature extraction is performed on the acquired images of the strain gauge after pasting to obtain the feature points after pasting; Calculate the deviation between the model feature points and the pasted feature points; the model feature points are from the area to be pasted in the digital model of the specimen; The uniformity and coverage of the adhesive layer of the strain gauge after pasting are calculated based on the feature points after pasting. Based on the deviation, the uniformity of the adhesive layer, and the coverage of the adhesive layer, the bonding evaluation of the strain gauge after bonding is performed to obtain the evaluation results. The evaluation results will determine whether the strain gauge is qualified after bonding.

[0012] Optionally, after determining that the strain gauge is qualified after pasting, the method further includes: The strain gauges were then cured after being pasted.

[0013] Based on the above-described method for attaching strain gauges to wind turbine blades, this application also discloses an apparatus for attaching strain gauges to wind turbine blades, used to perform the above-described method. The apparatus includes: a control module and an actuator; The control module is used to control the actuator; The actuator includes a camera device, a light source, a strain gauge pickup device, and an adhesive application device.

[0014] Optionally, the control module includes: a positioning unit, an adhesive application unit, and an adhesive bonding unit; The positioning unit is used to locate the target area on the specimen; The adhesive coating unit is used to control the actuator to coat an adhesive layer of preset thickness and preset area on the target strain gauge area with an adaptive trajectory, so as to obtain the adhesive-coated strain gauge. The pasting unit is used to control the actuator to paste the adhesive-coated strain gauge to the target area, thereby obtaining the pasted strain gauge.

[0015] Optionally, the positioning unit includes: An extraction subunit is used to extract features from the region image to obtain region image feature points; the region image is a camera image of the area to be pasted on the specimen; The computational subunit is used to calculate the transformation matrix from the camera coordinate system to the region image coordinate system; the region image coordinate system is obtained by calibrating the region image. The coarse positioning subunit is used to obtain the coarse positioning target region based on the transformation matrix, the region image feature points, and the model feature points; the model feature points are from the area to be pasted in the digital model of the specimen. The fine positioning subunit is used to perform edge detection and fitting on the coarse positioning target region to obtain the target region.

[0016] Optionally, the coarse positioning subunit includes: The matching subunit is used to match the feature points of the region image with the feature points of the model to obtain feature point matching pairs; The minimization sub-unit is used to minimize the reprojection error of the feature point matching pair based on the transformation matrix, thereby obtaining the coarse localization target region.

[0017] Optionally, the fine positioning subunit includes: The region acquisition subunit is used to extract the region of interest from the coarsely located target region to obtain the region of interest. The subpixel acquisition subunit is used to perform edge detection and Gaussian fitting on the region of interest to obtain the subpixel positions of the edges of the region of interest; A positioning subunit is used to locate the target region based on the subpixel position.

[0018] Optionally, the adhesive application unit includes: A positioning subunit for positioning the area to be coated with adhesive; The parameter acquisition subunit is used to acquire the adhesive application parameters of the target strain gauge; the adhesive application parameters include the size of the target strain gauge, the type of adhesive, and the amount of adhesive. The planning subunit is used to plan the adhesive application path based on the adhesive application parameters. A coating subunit is used to coat the adhesive layer according to the coating path to obtain the coated strain gauge.

[0019] Optionally, the planning subunit includes: A lattice-type adhesive application subunit is used to determine the lattice-type adhesive application to the target strain gauge and generate lattice coordinates when the size is less than or equal to a first size. A scanning adhesive application subunit is used to determine whether to perform linear reciprocating scanning adhesive application on the target strain gauge when the second size is larger than the first size and smaller than the second size, and to generate a scanning path and speed; the second size is larger than the first size. A filling-type adhesive application subunit is used to determine contour-filling adhesive application to the target strain gauge and generate a filling path when the size is greater than or equal to the second size.

[0020] Optionally, the device further includes: The feature extraction unit is used to extract features from the acquired image of the strain gauge after pasting to obtain the feature points after pasting. A deviation calculation unit is used to calculate the deviation between the model feature points and the pasted feature points; the model feature points come from the area to be pasted in the digital model of the specimen; Coverage calculation unit, used to calculate the adhesive layer uniformity and adhesive layer coverage of the strain gauge after pasting based on the pasted feature points; An evaluation unit is used to evaluate the bonding of the strain gauge after bonding based on the deviation, the uniformity of the adhesive layer, and the coverage of the adhesive layer, and to obtain the evaluation result. A determining unit is used to determine whether the strain gauge after pasting is qualified based on the evaluation results.

[0021] Optionally, the device further includes: The curing unit is used to cure the bonded strain gauge.

[0022] This application discloses a method and apparatus for bonding strain gauges to wind turbine blades. The main component is a device for bonding strain gauges to wind turbine blades. Utilizing machine vision technology, the target area on the specimen is precisely located, eliminating subjective errors caused by manual visual inspection and hand-drawing. The adhesive thickness and size are precisely controlled in the area to be coated on the target strain gauge. Simultaneously, an adaptive adhesive application trajectory effectively avoids the problem of inaccurate adhesive application by manual application, improving adhesive uniformity and bonding reliability, thereby enhancing the accuracy of strain test data. The coated strain gauge is then bonded to the target area, resulting in a bonded strain gauge. The method of this application is fully automated, reducing reliance on operator experience and improving bonding efficiency. It is suitable for batch and large-scale strain gauge bonding scenarios, achieving standardized and repeatable operation results. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating a method for attaching strain gauges to wind turbine blades, as disclosed in an embodiment of this application. Figure 2 This is a schematic flowchart of another method for attaching strain gauges to wind turbine blades disclosed in an embodiment of this application; Figure 3 This is a schematic diagram of the control module of a wind turbine blade strain gauge bonding device disclosed in an embodiment of this application. Detailed Implementation

[0025] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] Example 1: This application discloses a method for attaching strain gauges to wind turbine blades, wherein the main body for this method is a device for attaching strain gauges to wind turbine blades.

[0027] For details, please refer to Figure 1 The method for attaching strain gauges to wind turbine blades disclosed in this embodiment includes the following steps: Step 101: Locate the target area on the specimen.

[0028] In this embodiment, the specimen (such as a wind turbine blade to be tested) is first fixed on the test bench, and strain gauges are pre-placed in the feeding tray. The specimen has an area on which strain gauges need to be attached for stress-strain testing; this area is referred to as the target area in this embodiment. The vision system (such as a camera) of a wind turbine blade strain gauge attachment device (hereinafter, a robot is used as an example) locates the target area on the specimen and calculates its precise coordinates so that the adhesive-coated strain gauges can be subsequently attached to the target area.

[0029] As a feasible approach, a camera mounted at the end of a robotic arm acquires images of the area to be pasted on a strain gauge, resulting in a region image. At this stage, the region to be pasted has not yet undergone correction and differs from the accurate target area; therefore, it is simply referred to as the region to be pasted. After feature extraction from this region image, the robot obtains the region image feature points. Simultaneously, based on this region image, the robot identifies the surface reference of the wind turbine blade, performs coordinate system calibration, obtains the region image coordinate system, and acquires the surface normal information of the region to be pasted.

[0030] Specifically, Scale-Invariant Feature Transform (SIFT), Oriented Fast and Rotated BRIEF (ORB) and Rotation Robust Independent Basic Features (ORB) algorithms can be used to extract salient feature points of the region to be pasted from the region image. The expression for this can be shown in the following formula: (1) In the formula, P img p is the set of feature points of the region image. i =(u i vi ) represents the coordinates of the i-th region image feature point, and N represents the number of region image feature points.

[0031] In the method of this embodiment, model feature points of the area to be pasted can be obtained from the digital model of the curved surface of the wind turbine blade. The digital model can be a computer-aided design (CAD) model. The expression for these model feature points can be shown in the following formula: (2) In the formula, P cad Let q be the set of feature points of the model. j = (x j y j ) represents the coordinates of the j-th model feature point, and M represents the number of model feature points.

[0032] In the method of this embodiment, a transformation matrix is ​​calculated from the robot camera's camera coordinate system to the region image coordinate system. Based on this transformation matrix, region image feature points, and model feature points, a coarse localization target region is obtained. As one feasible approach, obtaining the coarse localization target region can specifically involve first matching the region image feature points with the model feature points to obtain feature point matching pairs. The expression for this pair can be shown below: (3) In the formula, P is the set of feature point matching pairs, (p k q k ) represents the k-th feature point matching pair, where k is the index of the feature point matching pair and K is the number of feature point matching pairs.

[0033] Subsequently, this feature point matching pair is used as input to the Perspective-n-Point (PnP) algorithm to calculate the transformation matrix T∈R from the camera coordinate system to the region image coordinate system. 4×4 And by minimizing the reprojection error of the feature point matching pairs, a coarsely localized target region is obtained. The formula for minimizing the reprojection error is as follows: (4) In the formula, π(·) represents the camera projection parameter.

[0034] In the method of this embodiment, for the transformation matrix T, the translation compensation amount and the selection compensation amount can be extracted as the positioning deviation of the specimen, which is used to correct the motion path in the subsequent robot motion, and three-dimensional posture compensation is performed based on the surface normal information of the area to be pasted.

[0035] As a feasible approach, after obtaining the coarsely located target area, edge detection and fitting algorithms can be used to precisely locate the area edges in order to achieve sub-pixel accuracy. This improves the edge detection accuracy from pixel level (integer coordinates) to sub-pixel level (decimal coordinates), resulting in higher positioning resolution and allowing the strain gauge placement error to be controlled within a very small range in subsequent steps.

[0036] Specifically, this can involve extracting the Region of Interest (ROI) from the coarsely located target area. The ROI is a region specifically defined within a complete image that requires focused analysis and processing. The edges of the ROI are then detected using the Canny or Sobel operators. In practice, due to factors such as lens optical characteristics, sensor sampling, and lighting, the grayscale change of an ideal step edge is not an ideal "step," but rather exhibits a smooth, approximately Gaussian transition. Therefore, a grayscale profile algorithm using Gaussian function fitting can overcome the discrete sampling limitations of integer pixels, obtaining the sub-pixel position of the edge through mathematical interpolation. The Gaussian function fitting formula is shown below: (5) In the formula, μ is the mean, the sub-pixel level edge position, i.e., the x-coordinate corresponding to the peak of the fitted Gaussian curve. G(x) is the gray value, i.e., the gray value of the image at position x. A is the amplitude, i.e., the gray difference between the two sides of the edge. σ is the standard deviation, used to reflect the blur width of the edge. B is the background brightness, i.e., the background gray value on the dark side (or bright side) of the edge.

[0037] In the method of this embodiment, the sub-pixel positions of the obtained multiple edges are fitted by least squares method to obtain the accurate center coordinates of the target area, thus completing the localization of the target area.

[0038] Step 102: Apply an adhesive layer of preset thickness and preset area to the target strain gauge area using an adaptive trajectory to obtain the adhesive-coated strain gauge.

[0039] In this embodiment, the robot picks up a target strain gauge from a feeding tray. This strain gauge has a pre-built and stored digital model, from which the robot can directly locate the area to be coated and obtain its boundary. Subsequently, the robot receives user-inputted or selected coating parameters for the target strain gauge (including the strain gauge's size, adhesive type, and amount of adhesive). Based on these parameters, a coating path is planned. This path can be adaptively adjusted according to the curvature of the wind turbine blade surface to ensure a uniform distribution of the adhesive layer on the blade's curved surface. Specifically, if the size in the coating parameters is less than or equal to a first size, the target strain gauge is determined to be a micro-strain gauge. A dot-matrix adhesive application is then performed on it, generating dot-matrix coordinates. These coordinates represent the positions where the robot applies adhesive to the area to be coated, resulting in a coated area filled with adhesive dots.

[0040] Accordingly, if the size is larger than the first size but smaller than the second size (which needs to be larger than the first size), the target strain gauge is determined to be a small to medium-sized strain gauge. A linear reciprocating scanning method can be used to apply adhesive to it, and a scanning path and speed can be generated. The linear reciprocating scanning method involves applying adhesive to the area to be coated in a linear back-and-forth motion, so that the area to be coated eventually appears to be covered with adhesive lines.

[0041] Accordingly, when the size is greater than or equal to the second size, the target strain gauge is determined to be a large gauge. Since large gauges require higher adhesive application standards, contour-filling adhesive application can be used to generate a filling path. This path is related to the shape of the area to be coated, ultimately resulting in an area covered with concentric adhesive lines. Taking a rectangular area as an example, the coating path for the first layer of adhesive is obtained by offsetting the outline of the rectangular area towards the center of the rectangle by a preset distance. For example, with the center point of the rectangle as the origin, the coating path for the first layer of adhesive can be (x1, y1) - (x2, y1) - (x2, y2) - (x1, y2) - (x1, y1). Then, the coating path for the second layer of adhesive is obtained by offsetting the coating path of the first layer of adhesive towards the center of the rectangle by a preset distance, which can be (x1+d, y2-d) - (x2-d, y1-d) - (x2-d, y2+d) - (x1+d, y2+d) - (x1+d, y2-d). This process continues until the area to be glued is filled with concentric square glue lines. The spacing between adjacent glue lines can be determined based on the diameter of the robot's dispensing valve nozzle, as shown in the following formula: (6) In the formula, d is the spacing, and k is the overlap coefficient (which can be taken as 0.6~0.8). W nozzle The diameter determines the width of a single adhesive line.

[0042] In the method of this embodiment, a simulated path collision detection and adhesive volume simulation can be performed first. After confirming that there are no problems with the path and adhesive dispensing, an adhesive layer of preset thickness and preset area is applied according to the planned adhesive application path to obtain the strain gauge after adhesive application.

[0043] Step 103: Adhere the adhesive-coated strain gauge to the target area to obtain the adhesive-coated strain gauge.

[0044] In the method of this embodiment, the robot moves the adhesive-coated strain gauge above the target area and precisely attaches the adhesive-coated strain gauge to the surface of the target area at a preset speed and pressure, so that the adhesive-coated strain gauge gradually attaches along the curved surface of the wind turbine blade, and holds the pressure for a preset time to obtain the adhesive-coated strain gauge.

[0045] As a feasible approach, images of the strain gauge after adhesion can be acquired and features extracted to obtain post-adhesion feature points. These feature points reflect the actual position, angle, and adhesive layer edge condition of the strain gauge. The deviation between the model feature points of the wind turbine blade surface model and the post-adhesion feature points is calculated to determine whether the target strain gauge is adhered to the correct position and angle. The formula for calculating the positional deviation is as follows: (7) In the formula, D pos The positional deviation is ΔX. pos ΔY pos This represents the coordinate deviation between the model feature points and the pasted feature points.

[0046] The formula for calculating the angular deviation in this deviation can be shown as follows: (8) In the formula, D ang Δθ is the angular deviation. pos The deviation angle between the model feature points and the pasted feature points.

[0047] As a feasible approach, the uniformity and coverage of the adhesive layer of the strain gauge after bonding can also be calculated based on the feature points after bonding. The formula is as follows: (9) (10) In the formula, U glue For adhesive layer uniformity, The variance of the adhesive layer width. For the average width, C cover For adhesive layer coverage, A glue Let A be the area of ​​the adhesive layer region. strain This represents the area of ​​the region to be coated with adhesive.

[0048] In the method of this embodiment, based on the above-mentioned deviation, adhesive layer uniformity, and adhesive layer coverage rate, the pasted strain gauge can be evaluated after pasting to obtain an evaluation result. This pasting evaluation can be to calculate a comprehensive quality score, and its formula can be as follows: (11) In the formula, S quality is the comprehensive quality score, and e is the natural constant. α1, α2, α3, α4 are preset weight coefficients, and their sum is 1. β1, β2 are preset attenuation coefficients.

[0049] In the method of this embodiment, according to the evaluation result, it can be determined whether the pasted strain gauge is qualified. Taking the comprehensive quality score as an example, it can be set that a comprehensive quality score exceeding the first threshold indicates that the pasted strain gauge is qualified. A comprehensive quality score between the first threshold and the second threshold (which needs to be less than the first threshold) indicates that the pasting deviation of the pasted strain gauge is relatively large and a warning needs to be given. A comprehensive quality score lower than the second threshold indicates that the pasting deviation of the pasted strain gauge is very large and the pasting is unqualified. Among them, the first threshold and the second threshold can be set according to actual needs. For example, the first threshold is set to 0.85 and the second threshold is set to 0.7. Here, the specific values of the first threshold and the second threshold are not limited, as long as the comparison of the values can be achieved.

[0050] As another feasible solution, taking the case of not calculating the comprehensive quality score as an example, the values of the deviation, adhesive layer uniformity, and adhesive layer coverage rate can be judged. For example, if the position deviation is too large, the deviation reason can be correspondingly recorded as positioning error. If the angle deviation is too large, the deviation reason can be correspondingly recorded as pasting alignment error or strain gauge body torsion. If the adhesive layer uniformity and adhesive layer coverage rate are insufficient, the deviation reason can be correspondingly recorded as uneven glue amount, having bubbles, or having contamination. Subsequently, the recorded data is generated into a feedback report, and the system makes a decision on whether it is qualified.

[0051] In the method of this embodiment, as a feasible solution, intelligent recognition can also be performed on defects that are difficult to quantify (such as bubbles, wrinkles, etc.). Specifically, a convolutional neural network can be used to extract the deep features of the image of the pasted strain gauge, and an Autoencoder is used to judge whether there are abnormalities in these deep features. Its formula can be as follows: (12) In the formula, Erecon is the reconstruction error, indicating the degree of difference between the input image and the reconstructed image. I input is the image of the pasted strain gauge, and I recon is the reconstructed image (the image regenerated by the Autoencoder according to the learned "normal mode").

[0052] In this embodiment, after the quality assessment is passed, the robot can move the pasted strain gauge to the curing lamp for curing. If the quality assessment fails, the robot can prompt for re-inspection of slightly defective strain gauges and prompt for repair or disposal of severely defective strain gauges.

[0053] The method described in this embodiment achieves full automation and intelligence throughout the entire process, from specimen reference identification, bonding area positioning, strain gauge pickup, quantitative adhesive application to precise bonding and quality inspection. It utilizes machine vision for sub-pixel-level positioning and real-time deviation compensation, improving bonding accuracy. Furthermore, it reduces reliance on operator experience, significantly improving bonding efficiency and enabling high-quality, high-volume, repeatable standardized operations. Simultaneously, it only requires changing the specimen fixture, adjusting algorithm parameters, and matching the corresponding bonding parameters in the process database to adapt to strain gauges of different sizes and shapes. It also supports bonding operations for planar and curved specimens without significant system modifications, meeting the diverse needs of mechanical performance testing scenarios and demonstrating high adaptability. Quality assessment, alerts, and test data recording are executed simultaneously, forming a closed-loop quality control system that ensures traceability and verifiability of the bonding process and results, promptly mitigating the impact of substandard bonding on subsequent strain tests.

[0054] Example 2: This application discloses another method for attaching strain gauges to wind turbine blades. Please refer to [link / reference]. Figure 2 This embodiment describes the process of strain gauge bonding, using a device for bonding strain gauges to wind turbine blades as the main implementer.

[0055] Step 201: The camera device acquires an image of the area to be pasted on the specimen and performs feature extraction to obtain the feature points of the area image.

[0056] Step 202: The control module acquires the model feature points of the area to be pasted in the digital model of the specimen.

[0057] Step 203: The control module calculates the transformation matrix from the camera coordinate system to the regional image coordinate system of the image.

[0058] Step 204: The control module locates the target region based on the transformation matrix, regional image feature points, and model feature points, using feature point matching, edge detection, and fitting algorithms.

[0059] Step 205: The robotic arm and strain gauge picking device pick up the strain gauges from the feed tray.

[0060] Step 206: The control module locates the area of ​​the strain gauge to be coated with adhesive based on the digital model of the strain gauge and obtains the adhesive coating parameters of the strain gauge.

[0061] Step 207: The control module plans the glue application path in the area to be glued according to the glue application parameters.

[0062] Step 208: The robotic arm and the adhesive application equipment apply the adhesive according to the application path.

[0063] Step 209: The robotic arm and strain gauge picking device attach the adhesive-coated strain gauges to the target area on the specimen.

[0064] Step 210: The camera device acquires images of the strain gauges after they have been pasted, and the control module evaluates their pasting process.

[0065] Step 211: Determine whether the evaluation result is qualified. If yes, proceed to step 212. If no, proceed to step 213.

[0066] Step 212: The robotic arm delivers the pasted strain gauge and the specimen together to the curing lamp for curing.

[0067] Step 213: Based on the evaluation results, either re-examine the strain gauge or scrap it after it has been pasted.

[0068] Based on the method for attaching strain gauges to wind turbine blades disclosed in the above embodiments, this embodiment discloses a device for attaching strain gauges to wind turbine blades, used to perform the above method. The device includes: a control module and an actuator; The control module is used to control the actuator; The actuator includes a camera device, a light source, a strain gauge pickup device, and an adhesive application device.

[0069] In this embodiment, the device can be a robot, specifically a six-axis industrial robot, connected to an industrial control computer via a control module. The robot's robotic arm end effector can be equipped with a vision camera, a ring light source, a vacuum nozzle, and a dispensing valve. A specimen clamp, a strain gauge feed tray, and a curing lamp can be mounted on the worktable.

[0070] In this embodiment, as Figure 3 As shown, the control module may specifically include: a positioning unit 301, an adhesive application unit 302, and an adhesive bonding unit 303; The positioning unit 301 is used to locate the target area on the specimen; The adhesive coating unit 302 is used to coat an adhesive layer of preset thickness and preset area on the area to be coated of the target strain gauge with an adaptive trajectory to obtain the coated strain gauge. The adhesive unit 303 is used to attach the adhesive-coated strain gauge to the target area to obtain the adhesive-coated strain gauge.

[0071] Optionally, the positioning unit 301 includes: An extraction subunit is used to extract features from the region image to obtain region image feature points; the region image is a camera image of the area to be pasted on the specimen; The computational subunit is used to calculate the transformation matrix from the camera coordinate system to the region image coordinate system; the region image coordinate system is obtained by calibrating the region image. The coarse positioning subunit is used to obtain the coarse positioning target region based on the transformation matrix, the region image feature points, and the model feature points; the model feature points are from the area to be pasted in the digital model of the specimen. The fine positioning subunit is used to perform edge detection and fitting on the coarse positioning target region to obtain the target region.

[0072] Optionally, the coarse positioning subunit includes: The matching subunit is used to match the feature points of the region image with the feature points of the model to obtain feature point matching pairs; The minimization sub-unit is used to minimize the reprojection error of the feature point matching pair based on the transformation matrix, thereby obtaining the coarse localization target region.

[0073] Optionally, the fine positioning subunit includes: The region acquisition subunit is used to extract the region of interest from the coarsely located target region to obtain the region of interest. The subpixel acquisition subunit is used to perform edge detection and Gaussian fitting on the region of interest to obtain the subpixel positions of the edges of the region of interest; A positioning subunit is used to locate the target region based on the subpixel position.

[0074] Optionally, the adhesive application unit 302 includes: A positioning subunit for positioning the area to be coated with adhesive; The parameter acquisition subunit is used to acquire the adhesive application parameters of the target strain gauge; the adhesive application parameters include the size of the target strain gauge, the type of adhesive, and the amount of adhesive. The planning subunit is used to plan the adhesive application path based on the adhesive application parameters. A coating subunit is used to coat the adhesive layer according to the coating path to obtain the coated strain gauge.

[0075] Optionally, the planning subunit includes: A lattice-type adhesive application subunit is used to determine the lattice-type adhesive application to the target strain gauge and generate lattice coordinates when the size is less than or equal to a first size. A scanning adhesive application subunit is used to determine whether to perform linear reciprocating scanning adhesive application on the target strain gauge when the second size is larger than the first size and smaller than the second size, and to generate a scanning path and speed; the second size is larger than the first size. A filling-type adhesive application subunit is used to determine contour-filling adhesive application to the target strain gauge and generate a filling path when the size is greater than or equal to the second size.

[0076] Optionally, the device further includes: The feature extraction unit is used to extract features from the acquired image of the strain gauge after pasting to obtain the feature points after pasting. A deviation calculation unit is used to calculate the deviation between the model feature points and the pasted feature points; the model feature points come from the area to be pasted in the digital model of the specimen; Coverage calculation unit, used to calculate the adhesive layer uniformity and adhesive layer coverage of the strain gauge after pasting based on the pasted feature points; An evaluation unit is used to evaluate the bonding of the strain gauge after bonding based on the deviation, the uniformity of the adhesive layer, and the coverage of the adhesive layer, and to obtain the evaluation result. A determining unit is used to determine whether the strain gauge after pasting is qualified based on the evaluation results.

[0077] Optionally, the device further includes: The curing unit is used to cure the bonded strain gauge.

[0078] The embodiments in this specification are described in a progressive manner. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant details can be found in the method section.

[0079] It should also be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0080] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0081] The features described in the embodiments of this specification can be substituted for or combined with each other, so that those skilled in the art can implement or use this application.

[0082] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for bonding strain gauges to wind turbine blades, characterized in that, The executing entity is a device for attaching strain gauges to wind turbine blades, and the method includes: Locate the target area on the specimen; A glue layer of preset thickness and preset area is applied to the target strain gauge area using an adaptive trajectory to obtain the glue-coated strain gauge. The adhesive-coated strain gauge is then attached to the target area to obtain the attached strain gauge.

2. The method according to claim 1, characterized in that, The target area on the positioning specimen includes: Feature extraction is performed on the region image to obtain the region image feature points; the region image is a camera image of the area to be pasted on the specimen; Calculate the transformation matrix from the camera coordinate system to the region image coordinate system; the region image coordinate system is obtained by calibrating the region image. Based on the transformation matrix, the regional image feature points, and the model feature points, a coarsely located target region is obtained; the model feature points are from the area to be pasted in the digital model of the specimen. Edge detection and fitting are performed on the coarsely located target region to obtain the target region.

3. The method according to claim 2, characterized in that, The step of obtaining the coarsely localized target region based on the transformation matrix, the region image feature points, and the model feature points includes: The feature points of the region image are matched with the feature points of the model to obtain feature point matching pairs; Based on the transformation matrix, the reprojection error of the feature point matching pair is minimized to obtain the coarsely localized target region.

4. The method according to claim 2, characterized in that, The step of performing edge detection and fitting on the coarsely located target region to obtain the target region includes: The region of interest is extracted from the coarsely located target region to obtain the region of interest; Edge detection and Gaussian fitting are performed on the region of interest to obtain the sub-pixel positions of the edges of the region of interest; The target region is located based on the sub-pixel position.

5. The method according to claim 1, characterized in that, The process of applying an adhesive layer of preset thickness and preset area to the target strain gauge area using an adaptive trajectory to obtain the coated strain gauge includes: Position the area to be coated with adhesive; Obtain the adhesive coating parameters of the target strain gauge; the adhesive coating parameters include the size of the target strain gauge, the type of adhesive, and the amount of adhesive. Plan the adhesive application path based on the adhesive application parameters; The adhesive layer is applied according to the described adhesive application path to obtain the adhesive-coated strain gauge.

6. The method according to claim 5, characterized in that, The step of planning the adhesive application path based on the adhesive application parameters includes: If the size is less than or equal to the first size, determine that the target strain gauge is to be applied with a dot matrix adhesive and generate dot matrix coordinates; When the size is greater than the first size and less than the second size, it is determined that the target strain gauge will be coated with adhesive in a linear reciprocating scanning manner, and a scanning path and speed will be generated; the second size is greater than the first size. If the dimension is greater than or equal to the second dimension, it is determined that the target strain gauge will be contour-filled with adhesive, and a filling path will be generated.

7. The method according to any one of claims 1-6, characterized in that, After obtaining the strain gauge after pasting, the method further includes: Feature extraction is performed on the acquired images of the strain gauge after pasting to obtain the feature points after pasting; Calculate the deviation between the model feature points and the pasted feature points; the model feature points are from the area to be pasted in the digital model of the specimen; The uniformity and coverage of the adhesive layer of the strain gauge after pasting are calculated based on the feature points after pasting. Based on the deviation, the uniformity of the adhesive layer, and the coverage of the adhesive layer, the bonding evaluation of the strain gauge after bonding is performed to obtain the evaluation results. The evaluation results will determine whether the strain gauge is qualified after bonding.

8. The method according to claim 7, characterized in that, After confirming that the strain gauge is qualified after bonding, the method further includes: The strain gauges were then cured after being pasted.

9. A device for bonding strain gauges to wind turbine blades, characterized in that, The apparatus for performing the method according to any one of claims 1-8, the apparatus comprising: a control module and an actuator; The control module is used to control the actuator; The actuator includes a camera device, a light source, a strain gauge pickup device, and an adhesive application device.

10. The apparatus according to claim 9, characterized in that, The control module includes: a positioning unit, an adhesive application unit, and an adhesive bonding unit; The positioning unit is used to locate the target area on the specimen; The adhesive coating unit is used to control the actuator to coat an adhesive layer of preset thickness and preset area on the target strain gauge area with an adaptive trajectory, so as to obtain the adhesive-coated strain gauge. The pasting unit is used to control the actuator to paste the adhesive-coated strain gauge to the target area, thereby obtaining the pasted strain gauge.

Images