A self-adaptive laser cutting control method and system for a special-shaped curved optical protection film

Abstract

This application relates to an adaptive laser cutting control method and system for irregular curved surface optical protective films, comprising: acquiring three-dimensional contour data of the irregular curved surface workpiece and establishing a curved surface coordinate system; extracting several discrete sampling points in the curved surface coordinate system and calculating the curvature change; classifying the discrete sampling points and their surrounding areas into region types; segmenting the cutting path of the optical protective film; during the cutting process, performing contour re-inspection on the edges of the completed optical protective film and comparing deviations; when the deviation comparison result exceeds the preset allowable tolerance range, performing local corrections and continuing cutting until the entire cutting path is completed; in summary, this application achieves precise matching between the cutting path and the irregular curved surface shape, stable control of the cutting depth, and process control of cutting quality, and has the effects of enhancing path adaptability, improving parameter adjustment accuracy, and real-time quality correction.

Description

Technical Field

[0001] This application relates to the technical field of laser cutting of optical protective films, and in particular to an adaptive laser cutting control method and system for irregular curved surface optical protective films. Background Technology

[0002] With the rapid development of consumer electronics, automotive displays and wearable devices, the application of irregular curved surface workpieces is becoming increasingly widespread. The surface of such workpieces usually needs to be covered with optical protective film to meet functional requirements such as scratch resistance, anti-reflection or privacy protection. After the film is covered, the edges of the optical protective film need to be precisely cut so that its outline matches the shape of the irregular curved surface workpiece.

[0003] Currently, laser cutting technology is relatively mature for cutting films on planar or simple regular curved surfaces, and the cutting operation is usually completed by using a preset fixed path and constant laser parameters. However, for irregular curved surfaces with complex curvature changes, the above method has obvious shortcomings: on the one hand, the curvature difference between different regions of the irregular curved surface is significant. If the cutting is advanced along a uniform path at a fixed feed rate, the cutting path is prone to deviate or the film edge lifts in areas with drastic curvature changes. On the other hand, the thickness distribution of the optical protective film is not uniform at different positions, and the laser incident angle changes with the normal direction of the curved surface. If the laser power is kept constant, the cutting depth is prone to be insufficient or overcut in areas with large normal angles, which will damage the surface of the workpiece.

[0004] In addition, existing cutting methods generally lack real-time contour inspection of the cut area. Cutting deviations can only be detected after all cutting is completed, and cannot be corrected in time during the cutting process, resulting in a high scrap rate, which is not conducive to quality control in actual production. Summary of the Invention

[0005] To address the aforementioned shortcomings, this application provides an adaptive laser cutting control method and system for irregular curved surface optical protective films.

[0006] The above-mentioned objective of this application is achieved through the following technical solution: An adaptive laser cutting control method for irregularly shaped curved surface optical protective films, applied to irregularly shaped curved surface workpieces and optical protective films attached to irregularly shaped curved surface workpieces, includes the following steps: Obtain the three-dimensional contour data of the irregular curved surface workpiece, and establish the surface coordinate system corresponding to the irregular curved surface workpiece based on the three-dimensional contour data; In the surface coordinate system, several discrete sampling points are extracted along the surface of the irregular curved workpiece, and the curvature change between adjacent discrete sampling points is calculated. Based on a preset curvature threshold, the discrete sampling points associated with curvature changes and their surrounding areas are divided into regions to determine high curvature transition zones and gently extending regions. Based on the distribution of the high curvature transition zone and the smooth extension zone, the cutting path of the optical protective film is segmented and planned. The segmented planning includes: generating cutting sub-paths point by point in the high curvature transition zone with the normal vector direction of discrete sampling points as the reference; generating continuous cutting sub-paths in the smooth extension zone by spline fitting; and connecting the cutting sub-paths in a preset manner to form the cutting path. Along the travel direction of the cutting path, the local thickness value of the optical protective film at the laser focus and the tilt angle of the surface normal corresponding to the current position are collected in real time. The laser power and feed rate required for the current cutting position are determined by combining the local thickness value and the tilt angle of the surface normal. During the cutting process based on the cutting path, the edge of the completed optical protective film is back-checked at a preset sampling interval, and the actual cutting contour obtained by the contour back-check is compared with the theoretical contour at the corresponding position of the cutting path. When the deviation comparison result exceeds the preset allowable tolerance range, the remaining unexecuted part of the cutting path is locally corrected and the cutting continues until the entire cutting path is completed.

[0007] The second objective of this invention is achieved through the following technical solution: An adaptive laser cutting control system for irregularly shaped curved optical protective films includes: The coordinate system establishment module is used to acquire the three-dimensional contour data of the irregular curved surface workpiece and establish the surface coordinate system corresponding to the irregular curved surface workpiece based on the three-dimensional contour data. The change calculation module is used to extract several discrete sampling points along the surface of an irregular curved workpiece in a curved coordinate system, and to calculate the curvature change between adjacent discrete sampling points. The region determination module is used to mark discrete sampling points whose curvature changes exceed a preset curvature threshold and their neighborhood as high curvature transition regions, and to mark discrete sampling points whose curvature changes do not exceed the preset curvature threshold and their neighborhood as smooth extension regions. The segmented planning module is used to segment the cutting path of the optical protective film based on the distribution of the high curvature transition zone and the smooth extension zone. The segmented planning includes: generating cutting sub-paths point by point in the high curvature transition zone with the normal vector direction of discrete sampling points as the reference; generating continuous cutting sub-paths in the smooth extension zone by spline fitting; and connecting the cutting sub-paths in a preset way to form the cutting path. The parameter determination module is used to collect the local thickness value of the optical protective film at the laser focus and the surface normal tilt angle corresponding to the current position in real time along the travel direction of the cutting path. The laser power and feed rate required for the current cutting position are determined by combining the local thickness value and the surface normal tilt angle. The deviation comparison module is used to perform contour back inspection on the edge of the completed optical protective film at a preset sampling interval during the cutting process based on the cutting path, and to compare the actual cutting contour obtained by contour back inspection with the theoretical contour at the corresponding position of the cutting path. The local correction module is used to locally correct the remaining unexecuted part of the cutting path when the deviation comparison result exceeds the preset allowable tolerance range, and then continue to execute the cutting until the entire cutting path is completed.

[0008] This application also relates to a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described adaptive laser cutting control method for irregular curved surface optical protective films.

[0009] In summary, this application achieves precise matching between the cutting path and the irregular curved surface shape, stable control of the cutting depth, and process control of the cutting quality by differentiating the curved surface region based on the curvature change, planning the cutting path segment by segment, adaptively adjusting the laser power and feed rate in real time according to the local thickness value and the tilt angle of the curved surface normal, and continuously performing contour back inspection and deviation closed-loop correction during the cutting process. It has the effects of enhancing path adaptability, improving parameter adjustment accuracy, and real-time quality correction. Attached Figure Description

[0010] Figure 1 This is a flowchart of an embodiment of an adaptive laser cutting control method for irregular curved surface optical protective film according to this application; Figure 2 This is a flowchart of step S20 in an embodiment of the adaptive laser cutting control method for irregular curved surface optical protective film of this application; Figure 3 This is a flowchart of step S22 in an embodiment of the adaptive laser cutting control method for irregular curved surface optical protective film of this application. Detailed Implementation

[0011] The following is in conjunction with the appendix Figures 1-3 This application will be described in further detail.

[0012] In the traditional cutting process of optical protective films for irregularly shaped curved workpieces, the use of a uniform path and constant parameters in areas with drastic curvature changes easily leads to cutting path deviation and film edge lifting. Furthermore, uneven thickness distribution of the optical protective film and changes in the laser incident angle with the surface normal direction can cause insufficient cutting depth or overcutting. The lack of a real-time contour inspection mechanism for the completed cutting area prevents timely correction of cutting deviations, thus affecting cutting accuracy and product yield. In particular, the lack of effective modeling of the dynamic characteristics of curvature changes, the coupling effect of local thickness values, and the surface normal inclination angle prevents adaptive adjustment of cutting parameters. Moreover, the failure to differentiate the cutting path planning based on regional curvature characteristics reduces cutting stability in high-curvature transition areas and decreases cutting efficiency in gently extending areas, significantly impacting the geometric consistency of the optical protective film edges.

[0013] Specifically, in the manufacturing scenario of automotive curved displays, the edges of the display exhibit complex curvature variations, especially with significant abrupt curvature changes in the R-corner transition area. When cutting along a preset path at a fixed feed rate, the edges of the optical protective film are observed to lift up at locations with drastic curvature changes. Furthermore, the thickness of the optical protective film varies at different locations on the display, and the laser incident angle changes with the normal direction of the curved surface. If the laser power remains constant, insufficient cutting depth or damage to the display surface is detected in areas with large normal tilt angles. In addition, since there is no real-time contour detection step during the cutting process, deviations generated during the process can only be detected after the cutting is completed, resulting in the need for rework or scrapping of the entire batch of products.

[0014] If the above problems are not solved, the failure rate of optical protective film cutting will continue to rise, increasing the waste of raw materials, affecting the continuous operation of the production line, and making it difficult to meet the strict requirements of high-precision display equipment for edge cutting quality.

[0015] In one embodiment, such as Figure 1 As shown, this application discloses an adaptive laser cutting control method for irregularly shaped curved surface optical protective films, applied to irregularly shaped curved surface workpieces and optical protective films attached to irregularly shaped curved surface workpieces, specifically including the following steps: S10: Obtain the three-dimensional contour data of the irregular curved surface workpiece, and establish the surface coordinate system corresponding to the irregular curved surface workpiece based on the three-dimensional contour data; In this embodiment, an irregular curved surface workpiece refers to an object with a complex, irregular three-dimensional geometric shape, whose surface typically contains various curvature variations, such as a mobile phone screen, automotive glass, or the casing of a wearable device; an optical protective film refers to an optical thin film material attached to the surface of the irregular curved surface workpiece to provide functions such as scratch resistance, anti-reflection, and privacy protection. Before cutting, the film material is usually in sheet form, and after cutting, its edge contour needs to precisely match the shape of the workpiece; three-dimensional contour data refers to the geometric information of the irregular curved surface workpiece obtained through measurement, scanning, or design modeling, usually represented in the form of point clouds, meshes, or CAD models, used to describe the precise shape of the workpiece surface; a surface coordinate system refers to a local or global coordinate system established on the surface of the irregular curved surface workpiece, used to accurately describe the position, normal vector, and curvature of any point on the workpiece surface.

[0016] Specifically, three-dimensional contour data can be obtained in various ways, such as scanning the workpiece with a 3D scanner to generate point cloud data; or directly importing the CAD design model of the workpiece; the surface coordinate system can be established by aligning the workpiece model with a preset reference coordinate system, or by constructing a local coordinate system closely related to the workpiece surface based on the geometric features of the workpiece (such as symmetry planes and reference points).

[0017] S20: In the surface coordinate system, extract several discrete sampling points along the surface of the irregular curved workpiece, and calculate the curvature change between adjacent discrete sampling points; In this embodiment, discrete sampling points refer to a finite number of points selected on the surface of an irregular curved workpiece according to certain rules or spacing, used to approximately represent the overall surface features of the workpiece; curvature change refers to the difference in curvature between adjacent discrete sampling points, used to reflect the changing trend of the local curvature of the surface.

[0018] Specifically, discrete sampling points can be sampled at uniform intervals, for example, selecting points on the workpiece surface according to a fixed grid spacing; or, an adaptive sampling method based on curvature changes can be used, that is, increasing the sampling density in areas with large curvature changes and decreasing the sampling density in areas with gentle curvature; the amount of curvature change between adjacent discrete sampling points can be calculated using numerical differentiation, for example, by calculating the difference in curvature at adjacent points to approximate it.

[0019] S30: Based on a preset curvature threshold, the discrete sampling points associated with curvature changes and their surrounding areas are divided into regions to determine the high curvature transition zone and the smooth extension zone. In this embodiment, the high curvature transition zone refers to the area on the surface of the irregular curved workpiece where the curvature changes drastically or the degree of curvature is large, such as acute angles, rounded corners or complex transition surfaces; the smooth extension zone refers to the area on the surface of the irregular curved workpiece where the curvature changes gently and is relatively flat, such as a large area of ​​plane or a gentle arc surface.

[0020] Specifically, the region type division can adopt a threshold judgment method. For example, regions with curvature changes greater than a certain preset threshold are marked as high curvature transition regions, and regions with curvature changes less than the preset threshold are marked as smooth extension regions. The determination of the domain range can adopt a circular or square region with a fixed radius, taking into account the region around the sampling point.

[0021] S40: Based on the distribution of the high curvature transition zone and the smooth extension zone, the cutting path of the optical protective film is segmented and planned. The segmented planning includes: generating a cutting sub-path point by point in the high curvature transition zone with the normal vector direction of the discrete sampling points as the reference; generating a continuous cutting sub-path in the smooth extension zone by spline fitting; and connecting the cutting sub-paths in a preset manner to form a cutting path. In this embodiment, the cutting sub-path refers to the local cutting trajectory generated for a specific region in the segmented planning of the cutting path; spline fitting is a mathematical method that uses a piecewise polynomial function to approximate or interpolate a set of discrete data points to generate a smooth curve or surface.

[0022] Specifically, in the high curvature transition region, due to the drastic changes in the surface, a point-by-point generation method can be adopted. For example, the normal vector direction of each sampling point can be used as the attitude reference of the laser cutting head, and the straight line segment between adjacent sampling points can be used as the cutting sub-path. In the gently extending region, due to the gentle changes in the surface, a smoother curve fitting method can be adopted, such as generating continuous cutting sub-paths through polynomial fitting or Bézier curve fitting. The connection of the cutting sub-paths can be a straight line connection or an arc connection.

[0023] S50: Along the travel direction of the cutting path, the local thickness value of the optical protective film at the laser focus and the tilt angle of the surface normal corresponding to the current position are collected in real time. The laser power and feed rate required for the current cutting position are determined by combining the local thickness value and the tilt angle of the surface normal. In this embodiment, laser power refers to the energy intensity output by the laser beam during laser cutting. Its magnitude directly affects the cutting depth and cutting quality. Specifically, laser power increases with the increase of local thickness and decreases with the increase of the surface normal angle. Feed rate refers to the speed at which the laser cutting head moves along the cutting path. Its magnitude affects the cutting efficiency and cutting effect. Specifically, feed rate decreases with the increase of the surface normal angle. The surface normal angle is the angle between the incident direction of the laser beam and the normal direction of the irregular curved workpiece surface. It affects the absorption and penetration effect of laser energy in the film material.

[0024] Specifically, the local thickness value can be measured in real time using an online thickness sensor; the surface normal angle can be calculated based on the surface geometry information at the current cutting position; the laser power and feed rate can be determined using a pre-established parameter lookup table, where the corresponding power and rate can be directly queried based on the collected thickness and angle values.

[0025] S60: During the cutting process based on the cutting path, the edge of the completed optical protective film is back-checked with a preset sampling interval, and the actual cutting contour obtained by the contour back-check is compared with the theoretical contour at the corresponding position of the cutting path. In this embodiment, contour back inspection refers to measuring and inspecting the edge contour of the cut optical protective film during or after the cutting process to obtain actual cutting contour data; deviation comparison refers to comparing the actual cutting contour obtained from contour back inspection with the preset theoretical cutting contour to quantify the difference between the two.

[0026] Specifically, contour re-inspection can be performed using a machine vision system, which captures images of the cut edges with a camera and performs image processing to extract the actual contour, while the theoretical contour can be obtained directly from the cutting path data; deviation comparison can be performed using point-to-point distance calculation, such as calculating the Euclidean distance between points on the actual contour and corresponding points on the theoretical contour.

[0027] S70: When the result of the deviation comparison exceeds the preset allowable tolerance range, the remaining unexecuted part in the cutting path is locally corrected and the cutting continues until the entire cutting path is completed.

[0028] In this embodiment, the allowable tolerance range refers to the maximum deviation between the actual cutting profile and the theoretical profile that is allowed during the cutting process. If the deviation exceeds this range, the cutting quality is considered unqualified.

[0029] Specifically, local corrections can be achieved by adjusting the path offset, such as shifting the remaining cutting path by a correction amount; or, the remaining path can be locally replanned, such as recalculating a small segment of the cutting path in areas with large deviations.

[0030] For example, suppose there is an irregularly shaped curved workpiece with an optical protective film attached to its surface, and its edges need to be precisely cut. At the same time, the irregularly shaped curved workpiece has a complex shape, including sharply curved edge areas and relatively flat surface areas.

[0031] First, the irregular curved surface workpiece is scanned using a 3D scanning device to obtain its high-precision 3D contour point cloud data. Based on the 3D contour point cloud data, a curved surface coordinate system that closely fits the workpiece surface is established to provide a unified reference for all subsequent geometric calculations.

[0032] Furthermore, in the established surface coordinate system, a series of discrete sampling points are extracted along the workpiece surface at a certain initial interval. At the same time, for each sampling point and its adjacent points, its curvature is calculated through local surface fitting, and the curvature change between adjacent sampling points is further calculated. For example, at the corner of the workpiece, the curvature change will increase significantly, while in flat areas, the curvature change will be relatively small.

[0033] Furthermore, the calculated curvature change is compared with a preset curvature threshold. All sampling points with curvature changes exceeding the curvature threshold and their surrounding areas are classified as high curvature transition zones, which typically correspond to complex geometric features such as sharp bends and acute angles on the workpiece. Meanwhile, the remaining sampling points with smaller curvature changes and their surrounding areas are classified as gentle extension zones, which correspond to flat or gently curved parts of the workpiece.

[0034] Furthermore, based on the distribution of the high-curvature transition zone and the gently extending zone, the cutting path of the optical protective film is segmented and planned. Specifically, in the high-curvature transition zone, due to the drastic surface changes and high cutting precision requirements, short cutting sub-paths can be generated point by point using the normal vector direction of each discrete sampling point as a reference, ensuring that the laser cutting head is always perpendicular to the film surface and avoiding cutting angle deviation. In the gently extending zone, due to the gentle surface changes, spline fitting technology can be used to use the discrete sampling points in this area as control points to generate smooth and continuous cutting sub-paths, improving cutting efficiency. Finally, the cutting sub-paths generated in the high-curvature transition zone and the gently extending zone, as well as the continuous cutting sub-paths, are connected sequentially to form a complete cutting path.

[0035] Furthermore, as the laser cutting head travels along the cutting path, the local thickness value of the optical protective film at the laser focal point and the inclination angle of the surface normal corresponding to the current cutting position are collected in real time. For example, at the edge of the workpiece, the film material may become thinner due to stretching, and the laser incident angle may also change. The real-time collected data is used as input, and a pre-established two-dimensional parameter mapping table is used for querying and interpolation calculation to dynamically determine the laser power and feed rate required for the current cutting position. For example, when the local thickness value increases, the laser power will be increased accordingly to ensure cutting penetration. When the inclination angle of the surface normal increases, the effective depth of laser energy in the film material will decrease, so the laser power will be reduced, and the feed rate will also be reduced to avoid overcutting or incomplete cutting. While cutting continues, the edge of the optical protective film that has been cut is back-checked at a preset sampling interval. For example, a high-resolution vision sensor is used to take pictures of the cut film edge and extract the actual cutting contour. The actual cutting contour is then accurately compared with the theoretical contour at the corresponding position on the cutting path, and the deviation between the two is calculated.

[0036] At this point, if the deviation comparison result exceeds the preset allowable tolerance range, for example, if the actual cutting profile is found to be shifted inward at a sharp bend, the remaining part of the cutting path that has not yet been executed is immediately corrected locally. The local correction can be to fine-tune the subsequent path, shifting it outward by a compensation amount to correct the previous deviation. Finally, after the local correction is completed, the laser cutting process continues until the entire cutting path is completed.

[0037] Through the above technical solution, this embodiment introduces the analysis of the geometric features of irregular curved surface workpieces, which can realize adaptive segmented planning of the cutting path, thereby effectively solving the problems of path deviation and film edge lifting that are prone to occur in complex curved surface areas by traditional fixed path cutting. Compared with the method of using constant laser parameters in the prior art, the method of this embodiment can collect the local thickness value and surface normal tilt angle of the optical protective film in real time, and dynamically adjust the laser power and feed rate accordingly, thereby avoiding insufficient cutting depth or overcutting in areas with large normal tilt angles or uneven film thickness, thus protecting the workpiece surface and improving cutting quality. In addition, the method of this embodiment introduces a real-time contour back inspection and deviation correction link in the cutting process, so that cutting deviations can be detected and corrected in time, which can reduce scrap rate and improve production efficiency and quality control level.

[0038] In one embodiment, such as Figure 2 As shown, step S20 includes: S21: In the surface coordinate system, based on the preset initial sampling interval, extract several initial discrete sampling points along the surface of the irregular surface workpiece, and extract several adjacent initial discrete sampling points within a preset neighborhood range with each initial discrete sampling point as the center. Fit the adjacent initial discrete sampling points with local surface patches to obtain the local surface patches corresponding to each initial discrete sampling point. In this embodiment, the initial sampling interval can be a fixed spatial distance, such as a uniform interval along a predefined scanning path, or it can be set according to the overall size and accuracy requirements of the workpiece; the preset neighborhood range can be a spherical region with a fixed radius, or it can be a region containing a fixed number of nearest neighbor points; the local surface patch fitting can adopt a polynomial fitting method, such as a quadratic or cubic polynomial, to fit the sampling points in the neighborhood by the least squares method to obtain the mathematical expression of the local surface; or, the moving least squares method can be adopted to construct the local surface approximation by weighted averaging.

[0039] S22: Calculate the first principal curvature and the second principal curvature of the local surface patch corresponding to each initial discrete sampling point. Determine the average of the first principal curvature and the second principal curvature as the representative curvature of the initial discrete sampling point. Determine the absolute value of the difference between the representative curvatures of two adjacent initial discrete sampling points as the initial curvature change between the two adjacent initial discrete sampling points. In this embodiment, the first principal curvature and the second principal curvature refer to the geometric quantities describing the maximum and minimum degree of curvature of the surface at a certain point, respectively. Their average value, i.e., the mean curvature, can comprehensively reflect the overall curvature characteristics of the local surface at that point. For a fitted local surface patch, such as an explicit surface represented by z=f(x,y), the principal curvature can be calculated by calculating its first and second partial derivatives and using formulas in differential geometry, such as the Weingarten mapping or directly from the shape operator. The absolute difference in curvature between two adjacent initial discrete sampling points directly quantifies the degree of change in the curvature of the surface between the two adjacent initial discrete sampling points.

[0040] S23: Interpolate and encrypt the interval between adjacent initial discrete sampling points where the initial curvature change exceeds the preset encryption trigger threshold, supplement several encrypted discrete sampling points, and re-execute local surface patch fitting and representative curvature calculation based on the encrypted discrete sampling points and the initial discrete sampling points to obtain the set of discrete sampling points and the corresponding curvature change.

[0041] In this embodiment, when the initial curvature change between adjacent initial discrete sampling points exceeds a preset encryption trigger threshold, it indicates that the surface geometry of the corresponding region is complex and requires denser sampling for accurate description. At this time, new encrypted discrete sampling points can be generated between adjacent initial discrete sampling points through interpolation methods, such as linear interpolation, cubic spline interpolation, or more complex curve interpolation algorithms. The encryption trigger threshold can be set empirically based on the required accuracy and computational resources. After the interpolation encryption is completed, all initial and encrypted discrete sampling points are merged to form a more comprehensive set of sampling points. Local surface patch fitting and representative curvature calculation are performed again on each point in the new set of sampling points to ensure that the curvature information of all points is obtained based on their latest and most accurate local geometric information, thereby obtaining the final set of discrete sampling points and their corresponding curvature changes.

[0042] Specifically, the solution of this application achieves accurate capture of the geometric features of the irregular curved surface workpiece by performing multi-stage adaptive sampling and curvature analysis on the surface of the workpiece. Specifically, firstly, in the surface coordinate system, several initial discrete sampling points are extracted along the surface of the irregular curved surface workpiece based on a preset initial sampling interval. In order to understand the local surface morphology around each initial sampling point, several adjacent initial discrete sampling points within a preset neighborhood are extracted with each initial discrete sampling point as the center, and local surface patches are fitted to the adjacent initial discrete sampling points to obtain the local surface patches corresponding to each initial discrete sampling point. Further... To quantify the curvature of each local surface patch, the first principal curvature and the second principal curvature of the local surface patch corresponding to each initial discrete sampling point are calculated. These represent the maximum and minimum curvature of the surface in different directions, respectively. Their average value is used as the representative curvature of the initial discrete sampling point, which can comprehensively reflect the local surface characteristics of that point. By calculating the absolute value of the difference between the representative curvatures of two adjacent initial discrete sampling points, the initial curvature change between adjacent points can be obtained, directly reflecting the geometric complexity of the surface in the corresponding region. To address the potential issue of insufficient density in the initial sampling, this embodiment further introduces an adaptive encryption mechanism. This involves interpolating and encrypting the interval between adjacent initial discrete sampling points where the initial curvature change exceeds a preset encryption trigger threshold, supplementing several encrypted discrete sampling points. This indicates that in regions with drastic curvature changes, the sampling density will automatically increase, thereby more precisely depicting the geometric details of key regions. Finally, based on the encrypted discrete sampling points and the initial discrete sampling points, the local surface patch fitting and representative curvature calculation are re-executed to obtain the final set of discrete sampling points and the corresponding curvature change.

[0043] Through the above technical solution, this application can overcome the problems of traditional fixed sampling density methods, which may lead to the loss of curvature information due to insufficient sampling or computational redundancy due to oversampling when processing irregular curved surface workpieces. By introducing an adaptive encryption mechanism based on curvature change, the solution of this embodiment can increase the sampling density in areas with drastic curvature changes and maintain a lower sampling density in areas with gentle curvature, thereby improving the efficiency of data processing while ensuring the accuracy of capturing surface geometric features. Through the above-mentioned curvature information acquisition, a basis can be provided for the accurate division of high curvature transition areas and gentle extension areas, thereby enabling the cutting path planning of optical protective films to more accurately adapt to the complex geometry of irregular curved surface workpieces, effectively avoiding cutting deviations caused by inaccurate curvature information, and improving cutting quality and efficiency.

[0044] In one embodiment, such as Figure 3 As shown, step S22 includes: S221: Obtain the design surface model of the irregular curved surface workpiece, determine the theoretical surface normal vector at the location of each initial discrete sampling point based on the design surface model, compare the theoretical surface normal vector with the measured normal vector, and calculate the normal deviation angle. In this embodiment, obtaining the design surface model of the irregular curved workpiece refers to obtaining the ideal geometric shape data of the irregular curved workpiece through computer-aided design software or other modeling tools. The design surface model can be a mathematical surface equation, a NURBS surface defined by a set of control points, or a polygonal mesh model, and its function is to provide the theoretical geometric reference of the workpiece. Determining the theoretical surface normal vector at the location of each initial discrete sampling point based on the design surface model refers to calculating the unit vector orthogonal to the surface at each initial discrete sampling point according to the mathematical definition of the design surface model. This can be achieved by modifying the surface equation... The normal vector can be obtained by performing partial derivative calculations, or by querying the normal vector information of the corresponding point in the CAD model. The theoretical surface normal vector refers to the surface direction of the point in an ideal state. The normal deviation angle is calculated by comparing the theoretical surface normal vector with the measured normal vector. This means obtaining the actual surface normal vector at the initial discrete sampling point through measuring equipment (such as a 3D scanner, laser rangefinder, etc.), and then calculating the angle between the actual surface normal vector and the corresponding theoretical surface normal vector. This angle difference is the normal deviation angle, which reflects the local directional difference between the actual workpiece surface and the design model.

[0045] S222: When the normal deviation angle exceeds the preset normal deviation allowable threshold, the coordinates of the initial discrete sampling points are projected and corrected according to the theoretical surface normal vector. The corrected coordinates of the initial discrete sampling points are then substituted into the local surface patch for fitting to obtain the corrected local surface patch. In this embodiment, the preset allowable threshold for normal deviation refers to the maximum allowable deviation angle when comparing normal deviation angles. This allowable threshold can be set according to factors such as the workpiece's machining accuracy requirements, the accuracy of the measuring equipment, and material properties. It is used to determine whether the initial discrete sampling points need coordinate correction. Projecting the coordinates of the initial discrete sampling points based on the theoretical surface normal vector means that when the normal deviation angle exceeds the allowable threshold, the initial discrete sampling points with deviations are adjusted along the direction of their theoretical surface normal vector to make their coordinates closer to the corresponding positions on the design surface model. This correction can be achieved using orthogonal projection, minimum distance projection, etc., aiming to eliminate the positional error of the sampling points. Substituting the corrected initial discrete sampling point coordinates into the local surface patch fitting results in the corrected local surface patch. After correcting the coordinates of the initial discrete sampling points, the corrected points are used to refit the local surface patch. By using more accurate sampling point data, a more precise or more consistent local surface patch with the actual geometric characteristics of the irregular curved surface workpiece can be obtained. The local surface patch refers to a small-range surface segment obtained by fitting a local surface with a certain initial discrete sampling point as the center and using several adjacent sampling points within its preset neighborhood. It is usually expressed in mathematical forms such as quadratic surface, bicubic surface, or Jet surface. The corrected local surface patch refers to the local surface patch obtained by substituting the corrected sampling point coordinates into the local surface patch fitting calculation after the coordinates of the initial discrete sampling points have been projected and corrected. Compared with the local surface patch before correction, its geometric shape is closer to the design surface shape of the workpiece, which can effectively reduce the interference of measurement error on the principal curvature calculation results.

[0046] S223: Calculate the first principal curvature and the second principal curvature of the corrected local surface patch respectively, determine the average value of the first principal curvature and the second principal curvature as the representative curvature of the initial discrete sampling point, and determine the absolute value of the difference between the representative curvatures of two adjacent initial discrete sampling points as the initial curvature change between the two adjacent initial discrete sampling points.

[0047] In this embodiment, after obtaining the corrected local surface patch, the principal curvature of the local surface patch at each sampling point is calculated according to the principle of surface differential geometry, and its mean is further calculated as the representative curvature. By comparing the representative curvature of adjacent sampling points, the degree of curvature change is quantified to ensure the accuracy of curvature calculation and provide geometric information for region type division and cutting path planning. The first principal curvature and the second principal curvature refer to the normal curvature corresponding to two mutually orthogonal principal directions at a point on the surface, respectively called the first principal curvature and the second principal curvature. The principal curvature is the core geometric quantity describing the degree of local bending of the surface. The principal curvature corresponds to the direction of maximum curvature, and the second principal curvature corresponds to the direction of minimum curvature. Both can be obtained by solving the eigenvalues ​​of the shape operator using the first and second basic forms of the local surface patch. The representative curvature is a comprehensive curvature index obtained by taking the arithmetic mean of the first and second principal curvatures of the local surface patch corresponding to a certain initial discrete sampling point. It reflects the average curvature of the surface at that sampling point in all directions as a single scalar. The initial curvature change is the absolute value of the difference between the representative curvatures corresponding to two adjacent initial discrete sampling points. It quantitatively describes the degree of curvature jump of the surface between adjacent sampling points.

[0048] For example, as a specific implementation, assume that the design surface model of the irregular curved workpiece is stored in a CAD file in the form of a NURBS surface. After obtaining the design surface model, its theoretical surface normal vector can be calculated at each initial discrete sampling point using the mathematical expression of the NURBS surface. For example, for a parameterized NURBS surface S(u, v), its normal vector can be calculated using S... u ×S v Obtain, where S u and S vIt is the partial derivative of S with respect to u and v; simultaneously, a high-precision 3D scanner can be used to scan irregularly shaped curved workpieces to obtain point cloud data containing the coordinates of each sampling point and its corresponding measured normal vector. The measured normal vector obtained from the scan is then multiplied by the theoretical normal vector calculated from the CAD model, and the angle between the two, i.e., the normal deviation angle, is calculated based on the vector magnitude. For example, if the normal deviation angle of a sampling point exceeds the preset allowable threshold of 0.5 degrees, the coordinates of that sampling point need to be corrected. The correction process can be achieved by projecting the sampling point along the direction of its theoretical surface normal vector onto the design surface, or... The corrected coordinates are obtained by finding the point on the design surface closest to the original sampling point through iterative optimization. For example, Newton's iteration method or gradient descent method can be used to find the point on the design surface with the smallest distance from the original sampling point. The corrected initial discrete sampling point coordinates are then used for fitting local surface patches. For example, a quadratic surface patch can be fitted using the least squares method. Based on this corrected quadratic surface patch, the first principal curvature and the second principal curvature can be calculated using its Hessian matrix, and their average value can be taken as the representative curvature of the sampling point. Finally, the change in initial curvature is obtained by calculating the absolute difference between the representative curvatures of adjacent sampling points.

[0049] Through the above technical solution, this application can effectively solve the problem of inaccurate curvature calculation caused by the coordinate deviation of the initial discrete sampling points. Specifically, by introducing a designed surface model as a correction benchmark and comparing the normal deviation and correcting the projection of the initial discrete sampling points, the sampling point data used for fitting the local surface patch can be more accurate, thereby obtaining the corrected local surface patch. The accuracy and reliability of the representative curvature and the initial curvature change calculated based on the corrected surface patch are significantly improved, making the division of region types more accurate, the cutting path planning can more accurately fit the actual geometric features of the irregular curved surface workpiece, and provide a more reliable geometric basis for the adaptive adjustment of laser power and feed rate. In summary, the solution of this embodiment can improve the laser cutting accuracy and quality of irregular curved surface optical protective films, and has the effect of reducing cutting defects and improving production efficiency.

[0050] In one embodiment, step S30 includes: S31: Traverse all discrete sampling points, set adjacent discrete sampling point pairs whose curvature change exceeds the preset curvature threshold as high curvature associated point pairs, and take each discrete sampling point in each high curvature associated point pair as the center, expand outward to the preset first neighborhood radius, and set the covered area as the high curvature transition zone. In this embodiment, step S31 aims to identify and fully cover areas with drastic curvature changes on irregular curved surfaces. This can be achieved in several ways: One approach is to first calculate the curvature change between adjacent discrete sampling points and compare it with a preset curvature threshold. If the curvature change exceeds the threshold, the pair of adjacent discrete sampling points is marked as a high curvature associated point pair. Subsequently, a preset first neighborhood radius is extended outward from each discrete sampling point in the high curvature associated point pair, such as a fixed distance or a distance dynamically adjusted based on local curvature. All areas covered by this neighborhood are defined as high curvature transition zones. Another approach is to use a method based on local curvature gradient analysis to identify a sequence of points where the rate of change of curvature gradient continuously exceeds a certain threshold. The identified sequence of points and its surrounding preset range are defined as high curvature associated regions, and then the high curvature transition zone is formed by expanding the first neighborhood radius.

[0051] S32: Traverse all discrete sampling points, set all discrete sampling points outside the high curvature transition area as smooth correlation points, and expand the preset second neighborhood radius outward with each smooth correlation point as the center, and set the coverage area that does not overlap with the high curvature transition area as the smooth extension area. In this embodiment, step S32 aims to identify and cover regions with gentle curvature changes, providing a stable foundation for continuous cutting path planning. This can be achieved in several ways: One approach is to first exclude all discrete sampling points that have been classified as high curvature transition zones from the total sampling point set, and the remaining unassigned discrete sampling points are considered as gentle correlation points. Subsequently, a preset second neighborhood radius is extended outward from the gentle correlation points. During the extension process, it is necessary to ensure that the formed gentle extension area does not overlap with the already determined high curvature transition zones to maintain the clarity and mutual exclusivity of the region division. Another approach is to first perform a preliminary curvature analysis on all discrete sampling points, and initially mark points with curvature changes below a certain low threshold as potential gentle points. After excluding the already determined high curvature transition zones, the neighborhood of the potential gentle points is extended, ensuring that the extended area does not overlap with the high curvature transition zones, ultimately forming a gentle extension area.

[0052] S33: When there is an uncovered gap region between the high curvature transition region and the smooth extension region, calculate the distance ratio between each location point in the gap region and the nearest boundary of the high curvature transition region and the boundary of the smooth extension region, and determine the region type of the gap region based on the distance ratio.

[0053] Furthermore, the region type of the gap region can be determined based on the distance ratio. If the distance between a certain point in the gap region and the boundary of the high curvature transition region is less than the distance between the point and the boundary of the smooth extension region, then the point is merged into the high curvature transition region; otherwise, it is merged into the smooth extension region.

[0054] In this embodiment, step S33 aims to address the issue of regions that may not belong to either the high curvature transition zone or the smooth extension zone during the region division process, ensuring that the entire surface of the irregular curved workpiece is clearly assigned to a certain region type. Specifically, for any unassigned point within the gap region, the Euclidean distance from that point to the nearest high curvature transition zone boundary and the Euclidean distance to the nearest smooth extension zone boundary can be calculated, and the assignment of the point can be determined by calculating the ratio of these two distances. If the distance from the point to the high curvature transition zone boundary is less than the distance to the smooth extension zone boundary, i.e., the distance ratio is less than 1, then the point is considered to be more inclined towards the high curvature transition zone and is assigned to the high curvature transition zone. Conversely, if the distance from the point to the smooth extension zone boundary is closer, then it can be assigned to the smooth extension zone. Another implementation method is to use an influence-based judgment model, assigning different influence weights to the high curvature transition zone and the smooth extension zone, and determining the final assignment based on the magnitude of the influence of the two regions on the points within the gap region, thereby realizing the region type division of the gap region.

[0055] For example, as a specific implementation, suppose the irregularly shaped curved workpiece is the edge region of a car windshield, which has a complex curved surface shape. First, by calculating the curvature change of discrete sampling points extracted from the glass surface, regions where the glass edge bends sharply, such as the connection between the A-pillar and the glass, can be identified. The curvature change in these regions will significantly exceed a preset curvature threshold. At this time, these point pairs are marked as high curvature associated point pairs, and a preset first neighborhood radius, such as 5 mm, is expanded outward from them to form a high curvature transition zone. Further, in flat or gently curved areas of the glass, such as the middle of the glass, the discrete sampling points in these areas will not belong to the high curvature transition zone, and they are set as gentle associated points. Centered on a point, a pre-defined second neighborhood radius, such as 10 mm, is extended outwards to form a gently sloping extension area. However, it is important to ensure that the gently sloping extension area does not overlap with the already determined high curvature transition area. If there is an uncovered narrow gap area between the high curvature transition area and the gently sloping extension area, such as at the boundary where the glass edge curves to a flat area, the ratio of the distance from each point in this gap area to the boundary of the high curvature transition area to the boundary of the gently sloping extension area is calculated. For example, if a point is 3 mm from the boundary of the high curvature transition area and 7 mm from the boundary of the gently sloping extension area, the distance ratio is less than 1, and the point can be merged into the high curvature transition area. Conversely, if the point is 7 mm from the boundary of the high curvature transition area and 3 mm from the boundary of the gently sloping extension area, the point can be merged into the gently sloping extension area.

[0056] Through the above technical solution, this application can overcome the ambiguity and gap area problems that may occur at the region boundary when using a single threshold division. Specifically, by introducing a neighborhood expansion mechanism, it can ensure full coverage of the high curvature transition area and the gently extending area, avoiding the omission of key areas. Especially when there are uncovered gap areas between the high curvature transition area and the gently extending area, the blind spots of region division can be effectively eliminated by merging the calculated distance ratio. This allows the entire surface of the irregular curved workpiece to be completely and clearly divided into different types of regions, providing clear region boundary information for the segmented planning of the cutting path. This improves the accuracy and continuity of the cutting path planning, thereby enhancing the cutting quality and efficiency of the optical protective film on complex curved surfaces.

[0057] In one embodiment, step S40 includes: S41: Sequentially extract the normal vector direction of each discrete sampling point in the high curvature transition zone, determine the attitude angle of the laser cutting head based on the normal vector direction of each discrete sampling point, and determine the feed direction of the corresponding cutting sub-path segment by connecting the adjacent discrete sampling points. Generate the cutting sub-path point by point according to the arrangement order of the discrete sampling points. In this embodiment, the extraction of the normal vector direction can be based on pre-acquired 3D contour data of the irregular curved surface workpiece. This can be achieved by fitting a local surface to discrete sampling points, for example, using the least squares method to fit a local quadratic surface, and then calculating the normal vector of the fitted surface at that point. Alternatively, if the 3D contour data is in point cloud form, the neighboring points around the discrete sampling points can be found directly using the K-nearest neighbor algorithm or radius search algorithm, and then principal component analysis or plane fitting can be used to estimate the normal vector at that point. The attitude angle of the laser cutting head typically refers to the angle between its optical axis and the normal to the workpiece surface. Using the normal vector direction as a reference, this means that the optical axis of the laser cutting head should be aligned with the normal vector direction or at a preset specific angle to achieve vertical cutting or specific... Angular cutting can be achieved through a multi-axis linkage control system, such as a five-axis or six-axis robot, which converts the calculated normal vector direction into the posture command of the robot's end effector. The direction of the line connecting adjacent discrete sampling points, i.e., the vector direction from the previous sampling point to the next sampling point, is used as the feed direction of the cutting sub-path segment, which can ensure that the cutting path smoothly advances along the sequence of sampling points. Point-by-point generation of cutting sub-paths means that in the high curvature transition region, the cutting path is not a continuous curve generated all at once, but consists of a series of short straight line segments or tiny curve segments defined by adjacent sampling points. Each point is associated with an independent posture angle and feed direction command, enabling more precise control in areas with drastic curvature changes.

[0058] S42: Extract the coordinates of each discrete sampling point in the gently extending region, use the coordinates as control points to perform spline curve fitting, obtain the spline fitting curve, and use the spline fitting curve as the cutting trajectory to generate a continuous cutting sub-path. In this embodiment, spline curves refer to a mathematical tool commonly used in computer graphics and CAD / CAM fields. It can generate smooth curves passing through a series of control points. Common spline curves include B-splines, NURBS curves, or cubic spline curves. By using discrete sampling points in a gently extending region as control points, a smooth curve can be fitted, which can reflect the geometric characteristics of the gentle region. Once the spline fitting curve is determined, it can be used as the cutting trajectory, indicating that the laser cutting head will move continuously along this mathematically defined smooth curve, rather than moving point by point as in high curvature regions. This helps to reduce motion jitter and improve cutting speed and surface finish.

[0059] S43: At the junction of the high curvature transition zone and the smooth extension zone, extract the tangent direction of the end point of the cutting sub-path in the high curvature transition zone and the tangent direction of the starting point of the continuous cutting sub-path in the smooth extension zone to generate a transition connection segment. Furthermore, the transition connection segment can be generated by curve interpolation based on the tangent direction of the end point of the high curvature transition zone cutting sub-path and the tangent direction of the starting point of the flat extension zone continuously cutting sub-path, so as to generate a transition connection segment tangent to the tangent directions at both ends.

[0060] In this embodiment, for the end point of the cutting sub-path in the high curvature transition region, its tangent direction can be determined by the point and its previous sampling point, or by the last small path segment defined by the point; for the starting endpoint of the continuous cutting sub-path in the smooth extension region, its tangent direction can be calculated by the derivative of the spline fitting curve at the starting point; furthermore, curve interpolation is a mathematical method used to generate a smooth curve connecting two endpoints given their tangent directions. Common interpolation methods include Hermite interpolation, Bezier curves, or B-spline curves; through curve interpolation, a smooth transition connection segment can be generated, ensuring that the cutting path has tangent continuity at the junction, i.e., the direction does not change abruptly.

[0061] S44: The cutting sub-path, continuous cutting sub-path, and transition connection segment are sequentially connected on the irregular curved surface workpiece to form a cutting path.

[0062] In this embodiment, the spatial distribution order refers to the connection order of each cutting sub-path and transition connection segment determined based on the previous region type division and the actual geometric arrangement of each region on the irregular curved surface workpiece. For example, if a high curvature transition area is followed by a smooth extension area, then the cutting sub-path of the high curvature area, the connection segment, and the continuous cutting sub-path of the smooth area will be connected in this order. By connecting the independent path segments end to end, a complete cutting path can be formed. This process usually involves splicing and integrating the geometric data of each path segment, such as point coordinates, attitude angles, and feed directions, to generate a unified CNC code or motion command sequence.

[0063] For example, as a specific implementation, suppose an optical protective film on a car windshield with a complex curved surface needs to be cut. First, the three-dimensional contour data of the windshield is obtained through three-dimensional scanning, and a surface coordinate system is established. Discrete sampling points are extracted in the surface coordinate system, and the curvature change is calculated. Then, high curvature transition areas are defined, such as sharp bends at the edges of the windshield, and gently extending areas, such as the flat area in the center of the windshield. Further, in the high curvature transition areas, such as the A-pillar connection area of ​​the windshield, for each discrete sampling point in this area, the normal vector direction is calculated through local surface fitting. For example, quadratic surface fitting based on the least squares method can be used to obtain the normal vector. The normal vector direction is then converted into the attitude command of the laser cutting head at the end of the six-axis robot, ensuring that the laser beam is always perpendicular to the glass surface. Simultaneously, the direction of the line connecting adjacent sampling points can be used as the feed direction of the cutting head. For example, if sampling points P1 and P2 are adjacent, the vector direction from P1 to P2 is the feed direction for that segment of the cut. In this way, a series of tiny cutting sub-paths can be generated point-by-point, enabling the laser cutting head to accurately follow the complex contours of high-curvature areas. In gently extending areas, such as the central area of ​​the windshield, the coordinates of discrete sampling points within this area are extracted. Using these extracted coordinates as control points, a cubic B-spline curve fitting algorithm can be used to generate a smooth spline fitting curve. The spline-fitted curve can then be used as the continuous cutting trajectory of the laser cutting head. For example, by parameterizing the B-spline curve, a series of continuous interpolation points can be generated, and the laser cutting head will move smoothly along the generated interpolation points, thereby achieving efficient and high-quality cutting. When encountering the boundary between a high-curvature transition zone and a gently extending zone, such as the connecting line from the sharp bend of the A-pillar to the central flat area, the end point of the cutting sub-path in the high-curvature transition zone is extracted, for example, the tangent direction of the end point of the last point-by-point generated path segment, and the starting point of the continuous cutting sub-path in the gently extending zone, for example, the tangent direction of the starting point of the spline-fitted curve. The tangent direction of the end point can be calculated from its previous point and the point itself. The tangent direction at the starting point can be calculated using the derivative of the spline curve at that point. Then, the Hermite curve interpolation method can be used, with the coordinates of the two endpoints and the tangent direction extracted above as boundary conditions, to generate a smooth transition connection curve. This transition connection curve can ensure the directional continuity of the cutting path between different regions. Finally, all the generated high-curvature transition zone cutting sub-paths, smooth extension zone continuous cutting sub-paths, and transition connection segments are spliced ​​together according to their actual geometric order on the windshield. For example, first the path segment in the high-curvature zone, then the connection segment, and finally the path segment in the smooth zone. After the above cycle, a complete laser cutting path that can be executed by the CNC system can be formed.

[0064] Through the above technical solutions, this application can adjust the cutting path generation strategy according to the local curvature characteristics of irregular curved workpieces. Specifically, in high curvature regions, by controlling the attitude angle and feed direction of the laser cutting head point by point, it can be ensured that the laser beam always acts on the workpiece surface at the optimal angle, thereby improving cutting accuracy and the ability to follow complex curved contours. In flat regions, spline fitting is used to generate continuous cutting sub-paths, which can effectively improve cutting efficiency and surface finish. Furthermore, and more importantly, by generating tangent-continuous transition connection segments at the boundaries of different regions, the problems of uneven path connection and discontinuity in traditional segmented planning can be solved, avoiding the decrease in cutting quality or equipment impact caused by abrupt path changes, thereby improving the stability of the entire cutting process and the cutting quality of the final product.

[0065] In one embodiment, step S43 includes: S431: Extract the tangent direction and curvature value at the end point of the cutting sub-path in the high curvature transition zone, and the tangent direction and curvature value at the beginning point of the continuous cutting sub-path in the smooth extension zone, and use them as the tangent constraint condition and curvature constraint condition at both ends of the transition connection segment, respectively. In this embodiment, the tangent direction refers to the direction pointed to by the tangent at a certain endpoint of the cutting sub-path, i.e., the direction of the first derivative of the path curve at that endpoint, represented by a unit tangent vector. It describes the travel trend of the cutting path at that endpoint and is the geometric constraint basis for ensuring tangent continuity (G1 continuity) between the transition connection segment and adjacent cutting sub-paths at the junction. The curvature value refers to the magnitude of the curve curvature of the cutting sub-path at a certain endpoint, defined as the rate at which the tangent direction of the curve turns through an angle per unit arc length at that point. It reflects the degree of curvature of the path curve at that endpoint and is the geometric constraint basis for ensuring curvature continuity (G2 continuity) between the transition connection segment and adjacent cutting sub-paths at the junction. Among them, the tangent direction can ensure the directional smoothness of the path at the connection, i.e., first-order continuity, while the curvature value can further improve the smoothness of the curvature of the path at the connection, i.e., second-order continuity. Furthermore, the tangent direction and curvature value can be obtained from the mathematical model of the adjacent cutting sub-paths. For example, if the cutting sub-path is composed of splines... The curve or Bézier curve can be used to represent the curve. The tangent direction can be obtained by taking the first derivative of the curve equation, and the curvature value can be calculated by taking the second derivative or the curvature formula. The tangent constraint condition refers to the boundary constraint of the tangent direction at the end point of the cutting sub-path in the high curvature transition zone and the tangent direction at the beginning point of the continuous cutting sub-path in the smooth extension zone. It requires that the tangent direction at the two joint points of the transition section is consistent with the tangent direction at the corresponding adjacent cutting sub-path endpoints. This ensures that there is no abrupt change in tangent direction between the transition section and the cutting sub-paths on both sides, that is, it satisfies G1 geometric continuity. The curvature constraint condition refers to the boundary constraint of the curvature value at the end point of the cutting sub-path in the high curvature transition zone and the curvature value at the beginning point of the continuous cutting sub-path in the smooth extension zone. It requires that the curvature value at the two joint points of the transition section matches the curvature value at the corresponding adjacent cutting sub-path endpoints. This ensures that there is no abrupt change in curvature between the transition section and the cutting sub-paths on both sides, that is, it satisfies G2 geometric continuity.

[0066] S432: Using tangent constraints and curvature constraints as boundary conditions, construct a set of curve interpolation equations for the transition connection segment, solve for the transition connection curve that satisfies that the tangent directions at both ends are tangent and the curvature values ​​at both ends match the curvature values ​​at the endpoints of the adjacent cutting sub-paths, and use the transition connection curve as the transition connection segment. In this embodiment, the curve interpolation equation system refers to the system of equations established for the transition connection segment, using tangent constraints and curvature constraints as boundary conditions, to solve for the control parameters of the transition connection curve. It can typically be constructed using parametric curves such as quintic Hermite splines, cubic splines, or Bézier curves. By simultaneously solving the position conditions of the two endpoints, the tangent direction conditions, and the curvature value conditions, a linear or nonlinear system of equations about the curve control parameters is formed. Solving this system yields an analytical expression for the transition connection curve that satisfies all boundary constraints. The transition connection curve is a parametric curve obtained by solving the curve interpolation equation system, which simultaneously satisfies the tangent constraints and curvature constraints at both endpoints. The transition connection curve is used to connect the end point of the cutting sub-path in the high curvature transition zone with the starting point of the continuous cutting sub-path in the smooth extension zone. The transition connection curve achieves tangential and curvature continuity with the adjacent cutting sub-paths at both ends, ensuring smooth advancement of the cutting head at the path transition and avoiding motion jitter and cutting quality degradation caused by abrupt changes in tangent or curvature. The transition connection segment refers to the actual cutting path segment determined by the transition connection curve's geometry, located between the cutting sub-path in the high curvature transition zone and the continuous cutting sub-path in the smooth extension zone, serving as a connector and smooth transition. The transition connection segment is a key path component in the cutting path segmentation planning used to eliminate geometric discontinuities between adjacent sub-paths.

[0067] S433: Verify the curvature continuity of the transition connection segment along the arc length direction, and calculate the curvature residual between the curvature values ​​at the joint points at both ends of the transition connection segment and the curvature values ​​at the endpoints of the adjacent cutting sub-paths; In this embodiment, curvature continuity verification refers to determining whether there is a discontinuity or abrupt change by comparing the curvature of the transition connection curve at the junction point with the curvature of the adjacent cut sub-path at that point; curvature residual refers to the quantitative index of such discontinuity in curvature continuity verification, which can be an absolute difference or a relative difference.

[0068] S434: When the curvature residual exceeds the preset curvature continuity allowable threshold, the boundary conditions of the curve interpolation equation system are iteratively adjusted based on the curvature residual, and the transition connection curve is resolved until the curvature residual does not exceed the preset curvature continuity allowable threshold. The solved transition connection curve is then set as the transition connection segment.

[0069] In this embodiment, the curvature continuity tolerance threshold refers to a pre-set curvature residual judgment limit value. When the curvature residual does not exceed the curvature continuity tolerance threshold, the curvature continuity of the transition connection segment at the joint point is considered to meet the engineering requirements, and the current transition connection curve can be determined as the final transition connection segment. When the curvature residual exceeds the curvature continuity tolerance threshold, the boundary conditions of the curve interpolation equation system need to be iteratively adjusted and re-solved until the curvature residual converges to within the threshold range. Iterative adjustment means that when the curvature residual exceeds the curvature continuity tolerance threshold, the boundary condition parameters in the curve interpolation equation system are gradually corrected based on the magnitude and direction of the current curvature residual, and after each correction... The transition connection curve is re-solved, and the above correction and solution process is repeated until the curvature residual meets the convergence condition. The purpose of this iterative optimization process is to eliminate curvature matching errors through multiple approximations, and finally obtain a transition connection curve that meets the G2 continuity requirement. Furthermore, the iterative adjustment process can use numerical optimization algorithms, such as gradient descent, Newton's method, or genetic algorithms, to fine-tune the parameters or boundary conditions of the curve interpolation equation system according to the magnitude and direction of the curvature residual. For example, the tangent direction or curvature value can be slightly corrected, or the control points of the interpolation curve can be adjusted, until the curvature residual converges to within the preset allowable threshold, thereby ensuring that the final generated transition connection curve has a high degree of curvature continuity.

[0070] For example, as a specific implementation, when generating the transition connection segment, the tangent vector and curvature scalar at the end point of the cutting sub-path in the high curvature transition zone, and the tangent vector and curvature scalar at the starting end point of the continuous cutting sub-path in the smooth extension zone, can be extracted first as boundary conditions for constructing a quintic Hermite spline curve. The quintic Hermite spline curve can simultaneously satisfy the continuity requirements of position, first derivative, and second derivative. Specifically, a quintic polynomial equation system with six unknown coefficients can be established based on the positions of the two endpoints, the tangent vector, and the curvature value. Solving this equation system yields the parametric expression of the transition connection curve. Subsequently, the curvature continuity of the generated transition connection curve is verified. For example, the curvature value can be calculated along the arc length direction of the transition connection segment with a preset step size, and compared with the curvature values ​​at the endpoints of adjacent cutting sub-paths to calculate the curvature residual. If, at a certain junction, the absolute value of the difference between the curvature of the transition connection curve and the curvature of the adjacent cutting sub-path exceeds a preset value of 0.01 mm... -1If the allowable threshold is reached, iterative adjustment can be initiated. This iterative adjustment can employ an optimization strategy based on Newton's method. Based on the magnitude and direction of the curvature residual, the control points or boundary conditions of the fifth-order Hermite spline curve are fine-tuned. For example, the tangent vector or curvature value is scaled or angularly adjusted slightly. Then, the curve equation is resolved. The above iterative adjustment process is repeated until the curvature residual at all junctions is less than the preset allowable threshold, thereby ensuring that the final transition connection curve achieves C2 continuity geometrically.

[0071] Through the above technical solution, this application can effectively solve the problem of curvature discontinuity that may occur when relying solely on tangent direction for curve interpolation. Specifically, by introducing curvature constraints and combining them with an iterative optimization mechanism, it can ensure that the transition connection segment and the adjacent cutting sub-path have high-order geometric continuity at the junction, that is, not only is the position and direction smooth, but the curvature change is also smooth, thereby improving the overall smoothness and accuracy of the cutting path, avoiding vibration or impact caused by abrupt changes in the path in the transition area of ​​the laser cutting head, effectively improving the cutting quality of the optical protective film, reducing the scrap rate, and extending the service life of the cutting equipment.

[0072] In one embodiment, step S50 includes: S51: Combine the local thickness value and the surface normal angle as coupled input variables to form the parameter state vector of the current cutting position; In this embodiment, the local thickness value and the surface normal angle are combined as coupled input variables to form the parameter state vector of the current cutting position. This means encapsulating the two key factors affecting the laser cutting effect—the local thickness value and the surface normal angle—into two-dimensional or multi-dimensional input data points, thereby simplifying the complexity of parameter transmission and processing, and providing a standardized input format for parameter mapping and interpolation calculation. For example, the local thickness value and the surface normal angle can be directly stored in an array or structure containing two floating-point numbers as the parameter state vector; or, these two values ​​can be normalized and then combined into a vector to facilitate unified processing between parameters of different dimensions.

[0073] S52: Using the preset local thickness value range and the surface normal tilt angle range as coordinate axes, construct a two-dimensional parameter mapping table. Each node in the two-dimensional parameter mapping table stores the laser power calibration value and feed rate calibration value that match the corresponding local thickness value range and surface normal tilt angle range. In this embodiment, a two-dimensional parameter mapping table is constructed using a preset local thickness value range and a surface normal tilt angle range as coordinate axes. This refers to constructing a data structure typically represented as a matrix or lookup table, where rows and columns represent discrete ranges of two different input parameters. Each cell in the two-dimensional parameter mapping table stores the output value corresponding to that combination of input parameters. Furthermore, the two-dimensional parameter mapping table pre-stores the optimal cutting parameters for different thickness and tilt angle combinations, avoiding complex real-time calculations and improving response speed. It transforms the continuous parameter space into queryable discrete points through discretization, providing basic data for interpolation calculations. For example, through experimental calibration or simulation, a series of discrete points can be selected within the preset local thickness value and surface normal tilt angle range, and the optimal laser power and feed rate corresponding to each discrete point can be measured or calculated. The measured data is then stored in a two-dimensional array; alternatively, a sparse matrix or hash table can be used for storage. A mapping table, especially useful when the parameter space is large but the effective data points are few, saves storage space. Each node in the two-dimensional parameter mapping table stores laser power calibration values ​​and feed rate calibration values ​​that match the corresponding local thickness range and surface normal angle range. These stored values ​​are typically determined through prior experiments, simulations, or experience, representing reference values ​​for laser power and feed rate that achieve ideal cutting results under specific combinations of local thickness and surface normal angles. The laser power calibration values ​​and feed rate calibration values ​​are the core content of the two-dimensional parameter mapping table, representing the optimal cutting strategy at discrete parameter points and providing accurate reference data for interpolation calculations. For example, calibration values ​​can be stored as floating-point numbers, accurate to several decimal places, to ensure fine adjustment of cutting parameters; or they can be stored as integers and multiplied by a scaling factor to represent the actual physical quantity, adapting to the needs of different control systems.

[0074] S53: Map the local thickness value and the surface normal angle in the parameter state vector to the corresponding coordinate axes of the two-dimensional parameter mapping table to determine the node interval where the parameter state vector is located in the two-dimensional parameter mapping table; In this embodiment, determining the node interval of the parameter state vector in the two-dimensional parameter mapping table means aligning the real-time local thickness value and the surface normal tilt angle with the discrete coordinate axes of the two-dimensional parameter mapping table to find its position in the table. The node interval refers to the area enclosed by four adjacent nodes in the mapping table. By locating the position of the real-time parameter in the preset mapping table, four basic data points for interpolation are determined. For example, simple numerical comparison and interval judgment can be used to determine which preset interval the local thickness value and the surface normal tilt angle fall into, thereby determining their row and column indices in the two-dimensional parameter mapping table, thus locking in the rectangular area formed by the four adjacent nodes. Alternatively, binary search or other efficient search algorithms can be used to quickly locate the position of the parameter value on the discrete coordinate axes, especially when the mapping table is large.

[0075] S54: Based on the laser power calibration value and feed rate calibration value of the adjacent nodes in the node interval where the parameter state vector is located, bilinear interpolation calculation is performed according to the relative position of the local thickness value and the surface normal inclination angle in the node interval to obtain the laser power and feed rate required for the current cutting position. In this embodiment, bilinear interpolation is a method for interpolation on a two-dimensional grid. It estimates the value of any point within the grid by weighted averaging of four adjacent data points. In this embodiment, the four adjacent data points refer to the laser power and feed rate calibration values ​​stored at the four corner points of the node interval. Through bilinear interpolation, continuous and accurate laser power and feed rate can be smoothly estimated between preset discrete calibration data based on the real-time changing local thickness value and surface normal tilt angle, thereby achieving adaptive adjustment of cutting parameters and compensating for the deficiencies of discrete calibration data. For example, bilinear interpolation can be achieved through two linear interpolations. First, linear interpolation is performed on two edges parallel to a coordinate axis to obtain the values ​​of two intermediate points. Then, linear interpolation is performed on the two intermediate points to obtain the final interpolation result. Alternatively, the mathematical formula of bilinear interpolation can be used directly for calculation, taking into account the values ​​of the four corner points and the relative position weights of the point to be interpolated within the interval.

[0076] S55: The laser power and feed rate are output to the laser cutting control terminal in real time, and the laser cutting control terminal adjusts the laser cutting head in real time according to the laser power and feed rate.

[0077] In this embodiment, the laser cutting control terminal refers to the hardware and software system used to receive cutting parameter instructions and drive the laser cutting equipment to perform cutting operations; real-time output means that the calculated parameter values ​​can be transmitted to the control terminal in a timely manner to match the dynamic changes in the cutting process; step S55 converts the calculated optimized parameters into actual physical control signals, which are directly applied to the laser cutting head to ensure that the laser power and feed rate can accurately respond to the local thickness changes and surface normal inclination of the workpiece surface, thereby maintaining stable cutting quality; for example, the laser power and feed rate can be transmitted to the laser cutting control terminal in the form of digital signals through communication interfaces such as industrial Ethernet, CAN bus, RS-232 / 485; or, the parameter values ​​can also be transmitted to the control terminal through analog signals, and the D / A converter inside the control terminal converts them into control signals to drive the laser and motion axis.

[0078] Specifically, this application couples the local thickness value with the surface normal angle and uses a preset two-dimensional parameter mapping table for rapid lookup and bilinear interpolation. This effectively solves the technical challenge of determining the laser power and feed rate in real time and accurately during the cutting of complex curved surfaces. Compared to simple linear relationships or single-parameter control, the solution in this embodiment can more comprehensively consider the mutual influence of the two key parameters. By smoothly transitioning discrete calibration data through interpolation, continuous adjustment of laser power and feed rate can be achieved throughout the cutting path. This allows the laser cutting head to better adapt to the complex geometric features of irregular curved workpieces and the local thickness changes of the optical protective film, improving the stability of the cutting process and the consistency of cutting quality. It also avoids cutting defects caused by parameter mismatch, such as scorching, incomplete cutting, or rough cutting edges.

[0079] Through the above technical solution, this application can achieve adaptive control of laser cutting parameters for irregular curved surface optical protective films. Specifically, by using the local thickness value and the surface normal angle as coupling input variables and performing bilinear interpolation using a two-dimensional parameter mapping table, the limitations of traditional methods in handling multi-variable dynamic coupling can be overcome. This allows the laser power and feed rate to be smoothly and continuously adjusted according to the real-time changes in the local thickness of the workpiece surface and the surface normal angle, improving the matching accuracy of cutting parameters. Therefore, the cutting quality remains consistent throughout the entire cutting path, effectively avoiding problems such as incomplete cutting, rough edges, and excessive heat-affected zones caused by improper parameters. This improves the processing accuracy and yield of irregular curved surface optical protective films.

[0080] In one embodiment, step S60 includes: S61: Collect the actual contour coordinates of the cut edge at the current sampling position, and transform the actual contour coordinates to the surface coordinate system to obtain the actual cut contour coordinates at the current sampling position; In this embodiment, step S61 aims to obtain the true geometric information of the edge of the cut optical protective film. The actual contour coordinates refer to the set of discrete coordinate points collected by measuring the edge of the cut optical protective film at the current sampling position. These coordinates directly reflect the true geometric shape of the cut edge after laser cutting. The actual contour coordinates can be collected using various high-precision measurement devices. For example, a machine vision-based contour recognition system can be used to acquire images of the cut edge using an industrial camera, and image processing algorithms can be used to extract the coordinates of edge pixels. Alternatively, other methods can be employed. Laser displacement sensors or 3D scanners directly acquire 3D point cloud data of the cutting edge. Transforming the acquired original coordinate points to the surface coordinate system of the irregular curved workpiece ensures that the two can be on a unified reference datum when comparing the theoretical contour, thereby eliminating errors caused by inconsistencies in coordinate systems and improving the accuracy of deviation calculation. The surface coordinate system refers to a locally parametric coordinate system established based on the 3D contour data of the irregular curved workpiece, with the geometric shape of the surface itself as the reference. It is usually constructed by taking a reference point on the surface as the origin and the tangent and normal directions of the surface as the coordinate axes.

[0081] S62: Based on the path mileage parameters of the current sampling position on the cutting path, extract the theoretical contour coordinate points at the corresponding positions on the cutting path; In this embodiment, step S62 aims to obtain theoretical cutting contour information corresponding to the actual measurement point. The path mileage parameter refers to the path travel distance parameter accumulated along the arc length direction of the cutting path from the starting point to the current sampling position. It is usually represented by the arc length value s. Furthermore, the path mileage parameter can uniquely identify the absolute position of the current sampling position on the cutting path and is the basis for locating and extracting the corresponding theoretical contour coordinate point on the cutting path. Through the path mileage parameter, the theoretical contour coordinate point corresponding to the current actual sampling position in space can be retrieved or calculated from the pre-planned cutting path data. The obtained theoretical contour coordinate point represents the cutting edge shape and position under ideal conditions and is used as a benchmark for deviation comparison.

[0082] S63: Align and register the theoretical contour coordinates with the actual cutting contour coordinates in the curved surface coordinate system to obtain several corresponding point pairs. Use the normal distance between each corresponding point pair as the local contour deviation value, and count the local contour deviation values ​​of all corresponding point pairs at the current sampling position. Take the maximum value as the representative contour deviation value of the current sampling position. In this embodiment, step S63 aims to quantify the difference between the actual cutting contour and the theoretical contour. Alignment and registration refers to the process of making the coordinate points of the theoretical contour and the actual cutting contour coincide as much as possible geometrically in a surface coordinate system through spatial transformation. This can typically be achieved using an iterative nearest-point algorithm or a feature-point-based registration algorithm to achieve optimal matching between the actual contour point cloud and the theoretical contour point cloud in the surface coordinate system. The purpose is to eliminate systematic positional deviations caused by factors such as measurement reference offset and sensor installation errors, ensuring that subsequent point-by-point deviation calculations can accurately reflect the local geometric errors of the cutting contour. After registration, a correspondence between actual contour points and theoretical contour points can be established to form corresponding point pairs. A corresponding point pair refers to the pairing of geometrically corresponding points in the set of theoretical contour coordinate points and the set of actual cutting contour coordinate points after alignment and registration. Typically, a point on the theoretical contour and its nearest point on the actual cutting contour form a corresponding point pair. Each corresponding point pair is used to calculate the contour deviation value at that local location. The normal distance refers to the projected distance along the normal direction of the theoretical contour at that point between the actual cutting contour coordinate point and the theoretical contour coordinate point in a corresponding point pair. The normal distance is used... Using non-Euclidean straight-line distance as a deviation metric can more accurately reflect the actual offset of the cutting edge in the direction perpendicular to the theoretical contour, eliminating the interference of sampling errors along the tangent direction on the deviation calculation results. In this embodiment, the normal distance between each corresponding point pair is used as the local contour deviation value because, in curved surface cutting, the deviation perpendicular to the surface normal direction has the most direct and significant impact on the cutting quality. By statistically analyzing the local contour deviation values ​​of all corresponding point pairs at the current sampling position and selecting the maximum value as the representative value of the contour deviation, the most severe cutting deviation within the sampling area can be effectively captured. The local contour deviation value refers to a specific... The normal distance between corresponding point pairs quantitatively describes the magnitude of the geometric deviation of the actual cutting profile from the theoretical profile at that local location. In this embodiment, the local profile deviation value is the basic unit for evaluating the accuracy level of cutting quality at a single corresponding point. The representative value of the profile deviation refers to the maximum value among all corresponding point pairs at the current sampling position after statistical analysis of their local profile deviation values. This value is taken as the representative index of the overall profile deviation level at the sampling position. In this embodiment, using the maximum value as the representative value ensures a conservative assessment of the most severe local deviation at the current sampling position and avoids masking local out-of-tolerance situations due to mean smoothing effects.

[0083] S64: Along the travel direction of the cutting path, the representative values ​​of the contour deviation at several consecutive sampling positions are accumulated and statistically analyzed. The weighted average of the representative values ​​of the contour deviation at several consecutive sampling positions is calculated, and the weighted average is used as the comprehensive deviation comparison result of the current cutting stage.

[0084] In this embodiment, step S64 aims to provide a more representative overall deviation assessment. Since the representative value of the contour deviation at a single sampling location may be affected by local noise or instantaneous disturbances, it is insufficient to fully reflect the cutting quality. However, by accumulating and statistically analyzing the representative values ​​of the contour deviation at several consecutive sampling locations along the cutting path, instantaneous fluctuations can be smoothed out, resulting in more stable trend information. Accumulated statistics refer to the process of summarizing and collecting the representative values ​​of the contour deviation at several consecutive sampling locations along the cutting path. By accumulating and statistically analyzing the deviation representative values ​​at multiple consecutive sampling locations, the overall deviation distribution trend within a certain path range at the current cutting stage can be comprehensively reflected, avoiding excessive influence of occasional errors at a single sampling location on the deviation evaluation results. The weighted average refers to... The comprehensive deviation statistic is obtained by weighting the representative values ​​of the contour deviation at several consecutive sampling locations according to the preset weight coefficients corresponding to each sampling location. The weight coefficients are usually set according to the temporal sequence of the sampling locations or the importance of the path. For example, sampling points closer to the current location are given higher weights so that the comprehensive deviation comparison results can better reflect the recent cutting quality status. Furthermore, the weighted average of the representative values ​​of the contour deviation can be calculated, and different sampling locations can be assigned different weights according to actual needs, so as to more accurately reflect the overall deviation situation at the current cutting stage. This weighted average, as the comprehensive deviation comparison result, is used to compare with the preset allowable tolerance range to determine whether it is necessary to trigger local correction for the remaining part of the cutting path. It is the core judgment basis for closed-loop control of cutting quality.

[0085] For example, as a specific implementation, in the process of laser cutting irregular curved surface optical protective film, a high-resolution linear CCD camera combined with a laser line scanner can be used to acquire the actual contour of the cut edge at the current sampling position in real time. The laser line scanner scans along the cutting path at a fixed frequency to acquire three-dimensional point cloud data of the cutting edge. The acquired three-dimensional point cloud data is first transformed into the surface coordinate system of the irregular curved surface workpiece through a pre-calibrated transformation matrix between the camera and the workpiece coordinate system, forming the actual cutting contour coordinates. Simultaneously, the cutting control system can calculate and extract the theoretical contour coordinates corresponding to the actual sampling points in real time from the stored NURBS curve-formed cutting path model based on the path mileage parameters fed back by the encoder on the current laser head along the cutting path. To improve the accuracy of the comparison, a feature point-based method can be used. The registration algorithm, for example, selects several points with obvious geometric features on the actual contour and the theoretical contour for coarse registration, and then uses an improved ICP algorithm for fine alignment to obtain high-precision corresponding point pairs. For each corresponding point pair, the shortest distance from the actual point to the theoretical surface is calculated as the local contour deviation value. For example, if 100 corresponding point pairs are collected within a sampling position, the maximum value among these 100 local contour deviation values ​​is taken as the representative contour deviation value of the current sampling position. Furthermore, the representative contour deviation values ​​of the most recent 5 consecutive sampling positions are continuously monitored and stored. When calculating the weighted average, a linear weighting method can be used. For example, the weight of the most recent sampling position is 0.4, the second most recent is 0.3, the previous one is 0.2, and the two furthest ones are 0.05 each. The weighted representative contour deviation values ​​are accumulated to obtain the comprehensive deviation comparison result of the current cutting stage.

[0086] By transforming the actual contour coordinates into a curved surface coordinate system and aligning them with the theoretical contour coordinates, the accuracy of deviation calculation can be ensured. Furthermore, by using the normal distance as the local contour deviation value and selecting the maximum value as the representative value of the contour deviation, the most critical deviation information of the cutting edge can be effectively captured. More importantly, by accumulating and calculating the weighted average of the representative values ​​of the contour deviation at several consecutive sampling positions, a more stable and representative comprehensive deviation comparison result can be provided. This allows for a more reliable assessment when determining whether cutting path correction is needed, avoiding misjudgments or over-corrections caused by instantaneous fluctuations or local anomalies. This improves the precision, stability, and product quality of laser cutting of irregular curved surface optical protective films.

[0087] Further, as a specific implementation method, the local correction in step S70 may specifically include: using the path mileage parameter of the current laser cutting head at the time of triggering correction as a reference, determining the starting position of the local correction; based on the distribution of local contour deviation values ​​of each continuous sampling position in the comprehensive deviation comparison result, backtracking to determine the path interval where the deviation continuously exceeds the limit, and determining the remaining unexecuted path segment corresponding to the path interval as the scope of the local correction; wherein, the termination position of the correction scope can be set as the deviation prediction convergence point, or limited to a preset fixed correction path length; based on the comprehensive deviation comparison result, calculating... The path correction vector for the current cutting stage is generated by using the weighted average of the representative values ​​of the contour deviation at each sampling position as the basis for the correction amount, and the direction of the correction vector is the opposite of the normal offset direction of the actual cutting contour relative to the theoretical contour. The magnitude of the path correction vector can be determined by the product of the deviation and the correction gain coefficient, which can be pre-calibrated based on the workpiece material properties and cutting process parameters. Based on the path correction vector, the remaining cutting path within the local correction range is geometrically corrected. Local path correction can be achieved using one of the following methods: Overall path translation correction: The remaining cutting paths within the correction range are translated along the path correction vector direction. This is suitable for situations where the deviation distribution is relatively uniform. Local path replanning: Based on the actual cutting contour coordinates at the corrected starting position and the path termination target point, the geometric calculation of the local cutting path is re-performed to generate a new local cutting path segment, which is suitable for situations with large deviations or uneven distribution. Furthermore, to avoid discontinuity in the cutting head movement due to abrupt changes in position between the corrected path and the original cutting path, at the starting position of the correction, a smooth transition connection segment can be generated by curve interpolation based on the tangent directions of the end point of the path before correction and the starting point of the path after correction, to ensure that the corrected path and the original path meet the tangent continuity. After the local correction is completed, the cutting control system continues to drive the laser cutting head to perform cutting according to the corrected cutting path. At the same time, it continuously monitors the subsequent cutting quality in real time according to the contour back inspection and deviation comparison process in steps S61 to S64. If the comprehensive deviation comparison result exceeds the allowable tolerance range again, the local correction process can be triggered repeatedly until the entire cutting path is completed.

[0088] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0089] In one embodiment, an adaptive laser cutting control system for irregularly shaped curved optical protective films is provided. This adaptive laser cutting control system for irregularly shaped curved optical protective films corresponds one-to-one with the adaptive laser cutting control method for irregularly shaped curved optical protective films described in the above embodiments. The adaptive laser cutting control system for irregularly shaped curved optical protective films includes: The coordinate system establishment module is used to acquire the three-dimensional contour data of the irregular curved surface workpiece and establish the surface coordinate system corresponding to the irregular curved surface workpiece based on the three-dimensional contour data. The change calculation module is used to extract several discrete sampling points along the surface of an irregular curved workpiece in a curved coordinate system, and to calculate the curvature change between adjacent discrete sampling points. The region determination module is used to mark discrete sampling points whose curvature changes exceed a preset curvature threshold and their neighborhood as high curvature transition regions, and to mark discrete sampling points whose curvature changes do not exceed the preset curvature threshold and their neighborhood as smooth extension regions. The segmented planning module is used to segment the cutting path of the optical protective film based on the distribution of the high curvature transition zone and the smooth extension zone. The segmented planning includes: generating cutting sub-paths point by point in the high curvature transition zone with the normal vector direction of discrete sampling points as the reference; generating continuous cutting sub-paths in the smooth extension zone by spline fitting; and connecting the cutting sub-paths in a preset way to form the cutting path. The parameter determination module is used to collect the local thickness value of the optical protective film at the laser focus and the surface normal tilt angle corresponding to the current position in real time along the travel direction of the cutting path. The laser power and feed rate required for the current cutting position are determined by combining the local thickness value and the surface normal tilt angle. The deviation comparison module is used to perform contour back inspection on the edge of the completed optical protective film at a preset sampling interval during the cutting process based on the cutting path, and to compare the actual cutting contour obtained by contour back inspection with the theoretical contour at the corresponding position of the cutting path. The local correction module is used to locally correct the remaining unexecuted part of the cutting path when the deviation comparison result exceeds the preset allowable tolerance range, and then continue to execute the cutting until the entire cutting path is completed.

[0090] Specific limitations regarding the adaptive laser cutting control system for irregularly shaped curved optical protective films can be found in the above description of the adaptive laser cutting control method for irregularly shaped curved optical protective films, and will not be repeated here. Each module in the aforementioned adaptive laser cutting control system for irregularly shaped curved optical protective films can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0091] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements an adaptive laser cutting control method for irregularly shaped curved optical protective films.

[0092] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. An adaptive laser cutting control method for irregularly shaped curved surface optical protective films, applied to irregularly shaped curved surface workpieces and optical protective films attached to irregularly shaped curved surface workpieces, characterized in that, Including the following steps: Obtain the three-dimensional contour data of the irregular curved surface workpiece, and establish the surface coordinate system corresponding to the irregular curved surface workpiece based on the three-dimensional contour data; In the surface coordinate system, several discrete sampling points are extracted along the surface of the irregular curved workpiece, and the curvature change between adjacent discrete sampling points is calculated. Based on a preset curvature threshold, the discrete sampling points associated with curvature changes and their surrounding areas are divided into regions to determine high curvature transition zones and gently extending regions. Based on the distribution of the high curvature transition zone and the smooth extension zone, the cutting path of the optical protective film is segmented and planned. The segmented planning includes: generating cutting sub-paths point by point in the high curvature transition zone with the normal vector direction of discrete sampling points as the reference; generating continuous cutting sub-paths in the smooth extension zone by spline fitting; and connecting the cutting sub-paths in a preset manner to form the cutting path. Along the travel direction of the cutting path, the local thickness value of the optical protective film at the laser focus and the tilt angle of the surface normal corresponding to the current position are collected in real time. The laser power and feed rate required for the current cutting position are determined by combining the local thickness value and the tilt angle of the surface normal. During the cutting process based on the cutting path, the edge of the completed optical protective film is back-checked at a preset sampling interval, and the actual cutting contour obtained by the contour back-check is compared with the theoretical contour at the corresponding position of the cutting path. When the deviation comparison result exceeds the preset allowable tolerance range, the remaining unexecuted part of the cutting path is locally corrected and the cutting continues until the entire cutting path is completed.

2. The adaptive laser cutting control method for irregular curved surface optical protective film according to claim 1, characterized in that: The step of extracting several discrete sampling points along the surface of an irregularly shaped workpiece in a curved coordinate system and calculating the curvature change between adjacent discrete sampling points includes: In the surface coordinate system, based on the preset initial sampling interval, a number of initial discrete sampling points are extracted along the surface of the irregular surface workpiece, and a number of adjacent initial discrete sampling points within a preset neighborhood are extracted with each initial discrete sampling point as the center. Local surface patches are fitted to the adjacent initial discrete sampling points to obtain the local surface patches corresponding to each initial discrete sampling point. The first principal curvature and the second principal curvature of the local surface patch corresponding to each initial discrete sampling point are obtained respectively. The mean of the first principal curvature and the second principal curvature is determined as the representative curvature of the initial discrete sampling point. The absolute value of the difference between the representative curvatures of two adjacent initial discrete sampling points is determined as the initial curvature change between the two adjacent initial discrete sampling points. Interpolate and encrypt the interval between adjacent initial discrete sampling points where the initial curvature change exceeds the preset encryption trigger threshold, supplement several encrypted discrete sampling points, and re-execute local surface patch fitting and representative curvature calculation based on the encrypted discrete sampling points and the initial discrete sampling points to obtain the set of discrete sampling points and the corresponding curvature change.

3. The adaptive laser cutting control method for irregular curved surface optical protective film according to claim 2, characterized in that: The steps of calculating the first principal curvature and the second principal curvature of the local surface patch corresponding to each initial discrete sampling point, determining the mean of the first principal curvature and the second principal curvature as the representative curvature of the initial discrete sampling point, and determining the absolute value of the difference between the representative curvatures of two adjacent initial discrete sampling points as the initial curvature change between the two adjacent initial discrete sampling points include: Obtain the design surface model of the irregular curved workpiece, determine the theoretical surface normal vector at the location of each initial discrete sampling point based on the design surface model, and compare the theoretical surface normal vector with the measured normal vector to calculate the normal deviation angle; When the normal deviation angle exceeds the preset allowable threshold for normal deviation, the coordinates of the initial discrete sampling points are projected and corrected according to the theoretical surface normal vector. The corrected coordinates of the initial discrete sampling points are then substituted into the local surface patch for fitting to obtain the corrected local surface patch. The first principal curvature and the second principal curvature of the corrected local surface patch are obtained respectively. The mean of the first principal curvature and the second principal curvature is determined as the representative curvature of the initial discrete sampling point. The absolute value of the difference between the representative curvatures of two adjacent initial discrete sampling points is determined as the initial curvature change between the two adjacent initial discrete sampling points.

4. The adaptive laser cutting control method for irregular curved surface optical protective film according to claim 1, characterized in that: The step of dividing discrete sampling points and their neighborhoods associated with curvature changes based on a preset curvature threshold to determine high curvature transition zones and gently extending zones includes: Traverse all discrete sampling points, set adjacent discrete sampling point pairs whose curvature change exceeds the preset curvature threshold as high curvature associated point pairs, and take each discrete sampling point in each high curvature associated point pair as the center, expand outward to the preset first neighborhood radius, and set the covered area as the high curvature transition zone. Traverse all discrete sampling points, set all discrete sampling points outside the high curvature transition area as smooth correlation points, and expand the preset second neighborhood radius outward with each smooth correlation point as the center, and set the coverage area that does not overlap with the high curvature transition area as the smooth extension area. When there is an uncovered gap region between the high curvature transition region and the smooth extension region, calculate the distance ratio between each location point in the gap region and the nearest boundary of the high curvature transition region and the boundary of the smooth extension region, and determine the region type of the gap region based on the distance ratio.

5. The adaptive laser cutting control method for irregular curved surface optical protective film according to claim 1, characterized in that: The method of segmenting the cutting path of the optical protective film based on the distribution of the high curvature transition zone and the smooth extension zone includes: generating cutting sub-paths point by point in the high curvature transition zone using the normal vector direction of discrete sampling points as a reference; generating continuous cutting sub-paths in the smooth extension zone using spline fitting; and connecting the cutting sub-paths in a preset manner to form the cutting path. The normal vector direction of each discrete sampling point in the high curvature transition zone is extracted sequentially. The attitude angle of the laser cutting head is determined based on the normal vector direction of each discrete sampling point. The feed direction of the line connecting adjacent discrete sampling points is determined by the line direction of the corresponding cutting sub-path segment. The cutting sub-path is generated point by point according to the arrangement order of the discrete sampling points. Extract the coordinates of each discrete sampling point in the gently extending region, use the coordinates as control points to perform spline curve fitting, obtain the spline fitting curve, and use the spline fitting curve as the cutting trajectory to generate a continuous cutting sub-path. At the boundary between the high curvature transition zone and the smooth extension zone, the tangent direction of the end point of the cutting sub-path in the high curvature transition zone and the tangent direction of the starting point of the continuous cutting sub-path in the smooth extension zone are extracted to generate the transition connection segment. The cutting path is formed by sequentially connecting the cutting sub-path, continuous cutting sub-path, and transition connection segment on the irregular curved surface workpiece according to their spatial distribution order.

6. The adaptive laser cutting control method for irregular curved surface optical protective film according to claim 5, characterized in that: The step of extracting the tangent direction of the end point of the cutting sub-path in the high-curvature transition zone and the tangent direction of the starting point of the continuous cutting sub-path in the smooth extension zone at the boundary between the high-curvature transition zone and the smooth extension zone, and generating the transition connection segment, includes: The tangent direction and curvature value at the end point of the cutting sub-path in the high curvature transition zone, and the tangent direction and curvature value at the beginning point of the continuous cutting sub-path in the smooth extension zone are extracted and used as the tangent constraint condition and curvature constraint condition at both ends of the transition connection segment, respectively. Using tangent constraints and curvature constraints as boundary conditions, a set of curve interpolation equations for the transition connection segment is constructed. The transition connection curves that satisfy the condition that the tangent directions at both ends are tangent and the curvature values ​​at both ends match the curvature values ​​at the endpoints of the adjacent cutting sub-paths are solved. The transition connection curves are used as the transition connection segments. The curvature continuity of the transition connection segment along the arc length direction is verified, and the curvature residual between the curvature values ​​at the joint points at both ends of the transition connection segment and the curvature values ​​at the endpoints of the adjacent cutting sub-paths is calculated. When the curvature residual exceeds the preset curvature continuity allowable threshold, the boundary conditions of the curve interpolation equation system are iteratively adjusted based on the curvature residual, and the transition connection curve is resolved until the curvature residual does not exceed the preset curvature continuity allowable threshold. The solved transition connection curve is then set as the transition connection segment.

7. The adaptive laser cutting control method for irregular curved surface optical protective film according to claim 1, characterized in that: The step of real-time acquisition of the local thickness value of the optical protective film at the laser focal point and the surface normal tilt angle corresponding to the current position along the travel direction of the cutting path, and determining the laser power and feed rate required for the current cutting position by combining the local thickness value and the surface normal tilt angle, includes: The local thickness value and the surface normal inclination angle are combined as coupled input variables to form the parameter state vector of the current cutting position; Using the preset local thickness value range and the surface normal tilt angle range as coordinate axes, a two-dimensional parameter mapping table is constructed. Each node in the two-dimensional parameter mapping table stores the laser power calibration value and feed rate calibration value that match the corresponding local thickness value range and surface normal tilt angle range. Map the local thickness value and the surface normal inclination angle in the parameter state vector to the corresponding coordinate axes of the two-dimensional parameter mapping table to determine the node interval where the parameter state vector is located in the two-dimensional parameter mapping table. Based on the laser power calibration value and feed rate calibration value of the adjacent nodes in the node interval where the parameter state vector is located, bilinear interpolation calculation is performed according to the relative position of the local thickness value and the surface normal inclination angle in the node interval to obtain the laser power and feed rate required for the current cutting position. The laser power and feed rate are output to the laser cutting control terminal in real time, and the laser cutting control terminal adjusts the laser cutting head in real time according to the laser power and feed rate.

8. The adaptive laser cutting control method for irregular curved surface optical protective film according to claim 1, characterized in that: The step of performing contour re-inspection on the edge of the completed optical protective film at a preset sampling interval during the cutting process based on the cutting path, and comparing the actual cutting contour obtained from the contour re-inspection with the theoretical contour at the corresponding position of the cutting path, includes: Collect the actual contour coordinates of the cut edge at the current sampling position, and transform the actual contour coordinates to the surface coordinate system to obtain the actual cut contour coordinates at the current sampling position; Based on the path mileage parameters of the current sampling position on the cutting path, extract the theoretical contour coordinate points at the corresponding positions on the cutting path; Align and register the theoretical contour coordinates with the actual cutting contour coordinates in the surface coordinate system to obtain several corresponding point pairs. Use the normal distance between each corresponding point pair as the local contour deviation value, and count the local contour deviation values ​​of all corresponding point pairs at the current sampling position. Take the maximum value as the representative contour deviation value of the current sampling position. Along the direction of travel of the cutting path, the representative values ​​of the contour deviation at several consecutive sampling positions are accumulated and statistically analyzed. The weighted average of the representative values ​​of the contour deviation at several consecutive sampling positions is calculated, and the weighted average is used as the comprehensive deviation comparison result of the current cutting stage.

9. An adaptive laser cutting control system for irregularly shaped curved optical protective films, characterized in that, include: The coordinate system establishment module is used to acquire the three-dimensional contour data of the irregular curved surface workpiece and establish the surface coordinate system corresponding to the irregular curved surface workpiece based on the three-dimensional contour data. The change calculation module is used to extract several discrete sampling points along the surface of an irregular curved workpiece in a curved coordinate system, and to calculate the curvature change between adjacent discrete sampling points. The region determination module is used to mark discrete sampling points whose curvature changes exceed a preset curvature threshold and their neighborhood as high curvature transition regions, and to mark discrete sampling points whose curvature changes do not exceed the preset curvature threshold and their neighborhood as smooth extension regions. The segmented planning module is used to segment the cutting path of the optical protective film based on the distribution of the high curvature transition zone and the smooth extension zone. The segmented planning includes: generating cutting sub-paths point by point in the high curvature transition zone with the normal vector direction of discrete sampling points as the reference; generating continuous cutting sub-paths in the smooth extension zone by spline fitting; and connecting the cutting sub-paths in a preset way to form the cutting path. The parameter determination module is used to collect the local thickness value of the optical protective film at the laser focus and the surface normal tilt angle corresponding to the current position in real time along the travel direction of the cutting path. The laser power and feed rate required for the current cutting position are determined by combining the local thickness value and the surface normal tilt angle. The deviation comparison module is used to perform contour back inspection on the edge of the completed optical protective film at a preset sampling interval during the cutting process based on the cutting path, and to compare the actual cutting contour obtained by contour back inspection with the theoretical contour at the corresponding position of the cutting path. The local correction module is used to locally correct the remaining unexecuted part of the cutting path when the deviation comparison result exceeds the preset allowable tolerance range, and then continue to execute the cutting until the entire cutting path is completed.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the adaptive laser cutting control method for irregular curved surface optical protective film as described in any one of claims 1-8.

Images