Laser cutting control system and method for lens assembly

Abstract

The invention relates to the technical field of laser cutting control, in particular to a laser cutting control system and method for a lens assembly. The method comprises the following steps that a lens assembly is scanned in an omnibearing mode to obtain point cloud data, then abnormal points are removed, three-dimensional reconstruction is carried out, a lens assembly model is generated, cutting task data are obtained, a cutting target is recognized, optimal cutting space-time arrangement is carried out on the lens assembly model on the basis, and an optimized cutting blueprint is generated. Cutting path planning is conducted according to the optimal blueprint, candidate cutting paths are obtained and dynamically compensated, a compensation cutting path is formed, mapping of control parameters is conducted according to the compensation cutting path, change control parameters are obtained, then parameter replacement is conducted, and accurate intelligent laser cutting control is achieved. The laser cutting control method for the lens assembly is more accurate and more efficient.

Description

Technical Field

[0001] The present invention relates to the field of laser cutting control technology, and in particular to a laser cutting control system and method for lens components. Background Art

[0002] The processing accuracy and efficiency of precision optical components are increasingly becoming the focus of attention in the manufacturing industry, especially in the fields of high-end optical instruments, precision meters and medical equipment. The processing quality of lens components directly affects the performance and reliability of the final product. Traditional lens component cutting methods have prominent problems such as unstable processing accuracy, low cutting efficiency, and great damage to materials, which seriously restrict the processing quality and production efficiency of optical components. Traditional cutting technology has the problems of difficulty in accurately controlling the cutting path, inability to dynamically adjust cutting parameters, lack of accurate identification and modeling capabilities of complex lens component morphologies, and often relies on empirical operations. It is impossible to achieve comprehensive intelligent control of the lens component cutting process, resulting in fluctuations in cutting accuracy, poor processing consistency, and low material utilization, which increases production costs and the risk of defective products. Summary of the invention

[0003] Based on this, it is necessary to provide a laser cutting control system and method for lens components to solve at least one of the above technical problems.

[0004] To achieve the above object, a laser cutting control method for a lens assembly comprises the following steps: Step S1: scanning the lens assembly in all directions to obtain lens assembly point cloud data; removing abnormal points from the lens assembly point cloud data to obtain filtered lens assembly point cloud data; performing three-dimensional reconstruction on the filtered lens assembly point cloud data to generate a lens assembly model; Step S2: obtaining cutting task data; performing cutting target recognition on the cutting task data to obtain cutting target data; performing optimal cutting time-space arrangement on the lens component model based on the cutting target data to generate an optimal cutting blueprint; Step S3: planning a cutting path for the lens component model based on the preferred cutting blueprint to obtain a candidate cutting path; dynamically compensating the candidate cutting path to obtain a compensated cutting path; Step S4: mapping control parameters according to the compensation cutting path to obtain changed control parameters; performing parameter replacement based on the changed control parameters to execute the intelligent laser cutting control method.

[0005] The present invention realizes accurate digital capture of the surface geometric features of lens components through omni-directional high-precision scanning technology, introduces an intelligent recognition and elimination algorithm of abnormal points based on machine learning, significantly improves the quality and reliability of point cloud data, and transforms discrete point clouds into continuous 3D geometric models through advanced 3D reconstruction technology, breaking through the limitations of traditional reconstruction methods and laying an accurate data foundation for subsequent cutting processing. At the same time, a deep learning target recognition model is constructed in the data analysis link of cutting tasks to accurately understand the processing requirements of complex lens components, introduce a multi-dimensional optimization algorithm for cutting time and space arrangement, comprehensively consider multiple constraints such as material utilization, processing accuracy and cutting efficiency, realize intelligent planning and dynamic adjustment of cutting paths, and further base on An intelligent path planning algorithm is developed to optimize the cutting blueprint. By introducing multiple constraints and dynamic path generation technology, the cutting path of the lens component is accurately controlled, and a machine learning dynamic compensation model is constructed to perceive the slight deformation and stress distribution of the cutting process in real time, adaptively adjust the cutting path, and effectively suppress the accumulation of processing errors and material deformation. Finally, by constructing a control parameter mapping model based on machine learning, the precise parameter conversion of the compensated cutting path is realized, and an adaptive parameter substitution algorithm is introduced to dynamically adjust the laser cutting control parameters according to the real-time processing environment, breaking through the traditional static parameter setting mode, significantly improving the intelligence level of the laser cutting process, and realizing precise control of the lens component processing process, comprehensively improving the cutting quality and production efficiency.

[0006] Preferably, step S1 comprises the following steps: Step S11: performing omnidirectional laser scanning on the lens assembly to obtain original point cloud data; performing spatial registration on the original point cloud data to obtain lens assembly point cloud data; Step S12: performing local density analysis on the lens component point cloud data to obtain point cloud density distribution data; performing outlier detection based on the point cloud density distribution data to generate an abnormal point set; Step S13: performing an abnormal elimination process on the abnormal point set to generate point cloud data of the filter lens component; performing a normal vector estimation process on the point cloud data of the filter lens component to obtain a normal vector field; Step S14: Perform three-dimensional reconstruction on the filtered lens component point cloud data based on the normal vector field to generate a lens component model.

[0007] The present invention can obtain comprehensive original point cloud data by implementing all-round laser scanning on the lens component, and the implementation of spatial alignment can ensure data consistency under different viewing angles. The generated lens component point cloud data provides a reliable basis for subsequent processing, and the implementation of local density analysis on the lens component point cloud data can reveal the distribution characteristics of the point cloud. The obtained point cloud density distribution data provides a basis for outlier detection, and the implementation of outlier detection based on the point cloud density distribution data can accurately identify abnormal points in the data. The generated abnormal point set provides a target for elimination processing, and the implementation of abnormal elimination processing on the abnormal point set can improve data quality. The generated filtered lens component point cloud data provides a clear basis for subsequent normal vector estimation, and the implementation of normal vector estimation processing on the filtered lens component point cloud data can provide direction information for three-dimensional reconstruction. The obtained normal vector field provides support for effective reconstruction, and the implementation of three-dimensional reconstruction of the filtered lens component point cloud data based on the normal vector field can generate an accurate lens component model, ensure the efficiency and accuracy of subsequent laser cutting, and improve the quality and reliability of the entire cutting process.

[0008] Preferably, step S2 comprises the following steps: Step S21: obtaining cutting task data; classifying cutting targets on the cutting task data to obtain target classification results; Step S22: performing cutting target recognition on the cutting task data based on the target classification result to generate cutting target data; Step S23: performing feature matching on the lens component model according to the cutting target data to obtain a candidate cutting target model; Step S24: performing arrangement constraint processing on the candidate cutting models to obtain the constrained cutting conditions; performing optimal spatiotemporal arrangement processing on the candidate cutting models based on the constrained cutting conditions to generate an optimal cutting blueprint.

[0009] The present invention can ensure the information integrity of the cutting process by acquiring the cutting task data, and the implementation of cutting target classification based on the cutting task data can effectively distinguish different cutting requirements. The obtained target classification result provides a basis for subsequent identification, and the implementation of cutting target identification based on the target classification result can accurately extract the target to be cut. The generated cutting target data provides a basis for feature matching. The implementation of feature matching of the lens component model according to the cutting target data can realize the generation of candidate cutting target models. The obtained candidate cutting target model provides a clear reference for subsequent arrangement processing, and the implementation of arrangement constraint processing of the candidate cutting model can ensure the feasibility of the cutting process. The generated constrained cutting conditions provide direction for the optimal spatiotemporal arrangement. The implementation of optimal spatiotemporal arrangement processing of the candidate cutting model based on the constrained cutting conditions can form an efficient cutting plan. The generated optimal cutting blueprint provides clear guidance for subsequent cutting path planning, improves the accuracy and efficiency of laser cutting control, and ensures the cutting quality and consistency of the lens component.

[0010] Preferably, step S24 includes the following steps: Step S241: performing boundary condition constraints on the candidate cutting model to obtain boundary condition constraint data; performing material utilization rate constraints on the candidate cutting model to generate material constraint data; Step S242: performing constraint integration processing on the boundary condition constraint data and the material constraint data based on the preset constraint weight data to obtain the constraint cutting conditions; Step S243: performing optimal cutting target matching on the candidate cutting models based on the constraint cutting conditions to obtain the optimal cutting sequence; performing adjacent distance calculation on the optimal cutting sequence to generate a target spacing matrix; Step S244: performing a heat impact analysis according to the preferred cutting sequence to obtain heat accumulation distribution data; performing conflict area identification on the heat accumulation distribution data to generate a conflict timing point set; performing time reallocation on the preferred cutting sequence based on the conflict timing point set to generate timing optimization data; Step S245: Calculate the safety distance of the target spacing matrix to obtain gap safety distance data; perform spacing adjustment processing on the target spacing matrix according to the gap safety distance to generate spacing optimization data; Step S246: Integrate and map the timing optimization data and the spacing optimization data to generate a preferred cutting blueprint.

[0011] The present invention can ensure that the cutting operation is carried out within physical limitations by implementing boundary condition constraints on candidate cutting models. The generated boundary condition constraint data provides a basis for subsequent processing. The implementation of material utilization rate constraints on candidate cutting models can optimize material usage. The generated material constraint data provides a basis for constraint integration. The implementation of constraint integration processing of boundary condition constraint data and material constraint data based on preset constraint weight data can form comprehensive constrained cutting conditions. The obtained constrained cutting conditions provide conditional support for preferred cutting target matching. The implementation of preferred cutting target matching of candidate cutting models based on constrained cutting conditions can identify the optimal cutting order. The generated preferred cutting sequence provides a basis for subsequent distance calculation. The implementation of adjacent distance calculation of the preferred cutting sequence can identify the relative position relationship between cutting targets. The generated target spacing matrix provides a reference for thermal impact analysis. The implementation of thermal impact analysis can evaluate the thermal effect in the cutting process, and the obtained heat accumulation distribution data provides support for conflict area identification. The implementation of conflict area identification on the heat accumulation distribution data can identify potential cutting conflicts. The generated conflict timing point set provides a basis for time reallocation. The implementation of time reallocation of the preferred cutting sequence based on the conflict timing point set can optimize the cutting timing and generate timing optimization data. The implementation of safe distance calculation on the target spacing matrix can ensure the safety of the cutting process. The obtained gap safety distance data provides a basis for target spacing adjustment. The implementation of spacing adjustment processing on the target spacing matrix according to the gap safety distance can optimize the distance between cutting targets and generate spacing optimization data. Finally, the implementation of integrated mapping of timing optimization data and spacing optimization data can form a final preferred cutting blueprint, ensure the efficiency and safety of laser cutting, and improve the accuracy and overall quality of lens component cutting.

[0012] Preferably, step S245 includes the following steps: Calculate the space requirement of the target spacing matrix to obtain the space requirement matrix; perform effective utilization analysis based on the space requirement matrix to obtain effective utilization space data; Extract cutting gaps based on effective use of spatial data and target spacing matrix to generate cutting gap data; Calculate the safety distance of the cutting gap data to obtain the gap safety distance data; According to the gap safety distance data, the target spacing matrix is ​​mapped to the safety reserved space to generate the reserved space data required for safety; The target spacing matrix is ​​adjusted based on the reserved space data required for safety to generate spacing optimization data.

[0013] The present invention can evaluate the space required in the cutting process by implementing space requirement calculation on the target spacing matrix, and the generated space requirement matrix provides a basis for subsequent effective utilization analysis. The implementation of effective utilization analysis based on the space requirement matrix can optimize resource allocation, and the obtained effective utilization space data provides a basis for cutting gap extraction. The implementation of cutting gap extraction based on the effective utilization space data and the target spacing matrix can identify gaps suitable for cutting, and the generated cutting gap data provides support for safety distance calculation. The implementation of safety distance calculation on the cutting gap data can ensure the safety during the cutting process, and the obtained gap safety distance data provides a reference for the adjustment of the target spacing matrix. The implementation of safety reserved space mapping on the target spacing matrix according to the gap safety distance data can formulate a reasonable reserved space strategy, and the generation of safety required reserved space data provides a basis for interval adjustment. The implementation of interval adjustment processing on the target spacing matrix based on the safety required reserved space data can optimize the spacing between cutting targets, and the generated spacing optimization data provides a guarantee for the accuracy and safety of laser cutting, thereby improving the overall effect and quality control of lens component cutting.

[0014] Preferably, step S3 comprises the following steps: Step S31: performing path constraint mapping on the lens component model based on the preferred cutting blueprint to obtain path constraint data; Step S32: planning a cutting path for the lens component model according to the path constraint data to obtain a candidate cutting path; Step S33: performing cutting simulation based on the candidate cutting path to obtain simulated component cutting data; performing deformation prediction on the simulated component cutting data to generate cutting deformation prediction data; Step S34: performing stress distribution calculation based on the cutting deformation prediction data to generate stress distribution data; performing stress path compensation according to the stress distribution data to obtain a compensated cutting path.

[0015] The present invention can ensure that the cutting path meets the design specifications by implementing path constraint mapping for the lens component model based on the preferred cutting blueprint, the generated path constraint data provides the necessary constraint basis for cutting path planning, the implementation of cutting path planning for the lens component model according to the path constraint data can identify the effective cutting path, the obtained candidate cutting path provides a basis for subsequent simulation analysis, the implementation of cutting simulation based on the candidate cutting path can evaluate the cutting effect before actual cutting, the generated simulated component cutting data provides an important basis for deformation prediction, the implementation of deformation prediction for the simulated component cutting data can identify the deformation during the cutting process, the generated cutting deformation prediction data provides support for stress analysis, the implementation of stress distribution calculation based on the cutting deformation prediction data can reveal the stress condition inside the material during the cutting process, the generated stress distribution data provides a reference for path optimization, the implementation of stress path compensation according to the stress distribution data can adjust the cutting path to reduce the risk of deformation, the obtained compensated cutting path provides guarantee for the precise control of laser cutting, improves the quality and consistency of lens component cutting, and ensures the controllability and safety of the final cutting result.

[0016] Preferably, step S33 includes the following steps: Acquire laser cutting heat source data; simulate the cutting temperature field of the laser cutting heat source data to obtain simulated temperature field distribution; Perform cutting cooling field simulation on laser cutting heat source data to generate simulated cooling field distribution; perform stress calculation on candidate cutting paths to obtain stress field data; Perform cutting simulation on the candidate cutting path based on the simulated temperature field distribution, simulated cooling field distribution and stress field data to obtain simulated component cutting data; Perform local strain calculation on the simulated component cutting data to obtain the local strain field; perform thermal deformation evaluation based on the local strain field to generate local thermal deformation data; Perform global deformation accumulation calculation on the simulated component cutting data according to the local thermal deformation data to obtain the accumulated deformation field; Deformation prediction is performed based on the accumulated deformation field to generate cutting deformation prediction data.

[0017] The present invention can accurately reflect the characteristics of the heat source in the cutting process by acquiring the implementation of laser cutting heat source data, and can predict the change of temperature distribution in the cutting process by implementing the cutting temperature field simulation on the laser cutting heat source data, and the generated simulated temperature field distribution provides a basis for the evaluation of the cutting effect, and can analyze the cooling effect by implementing the cutting cooling field simulation on the laser cutting heat source data, and the generated simulated cooling field distribution provides a reference for thermal management, and can reveal the distribution of stress in the cutting process by implementing the stress calculation on the candidate cutting path, and the obtained stress field data provides support for the subsequent cutting simulation, and can comprehensively consider the thermal effect and stress influence by implementing the cutting simulation on the candidate cutting path based on the simulated temperature field distribution, the simulated cooling field distribution and the stress field data, and the obtained simulated component cutting data is used for cutting. The optimization of the cutting process is provided; the implementation of local strain calculation on the simulated component cutting data can identify the local deformation of the material during the cutting process; the generated local strain field provides data support for thermal deformation evaluation; the implementation of thermal deformation evaluation based on the local strain field can evaluate the thermal response of the material during the cutting process; the generated local thermal deformation data provides a basis for the global deformation accumulation calculation; the implementation of global deformation accumulation calculation on the simulated component cutting data according to the local thermal deformation data can obtain the overall deformation of the material; the obtained cumulative deformation field provides support for the optimization of the cutting path; the implementation of deformation prediction based on the cumulative deformation field can predict the final cutting effect; the generated cutting deformation prediction data provides an important reference for the adjustment of the subsequent cutting strategy, ensuring the accuracy and quality control of the lens component cutting.

[0018] Preferably, step S34 includes the following steps: Perform stress tensor analysis on the cutting deformation prediction data to obtain stress component data; Perform distribution calculation on stress component data to generate stress distribution data; Calculate the compensation amount for the stress distribution data to obtain the path compensation data; The simulated component cutting data is compensated and superimposed according to the path compensation data to obtain a compensated cutting path.

[0019] The present invention can deeply understand the stress state inside the material during the cutting process by implementing stress tensor analysis on the cutting deformation prediction data, and the obtained stress component data provides a basis for stress distribution calculation. The implementation of distribution calculation on the stress component data can reveal the changing trend of stress in the material, and the generated stress distribution data provides a basis for subsequent compensation processing. The implementation of compensation amount calculation on the stress distribution data can determine the correction amount required for the cutting path, and the obtained path compensation data provides specific parameters for compensation superposition processing. The implementation of compensation superposition processing on the simulated component cutting data according to the path compensation data can adjust the cutting path to adapt to the deformation of the material. The finally generated compensated cutting path provides guarantee for the precise control of laser cutting, improves the quality and consistency of lens component cutting, ensures the controllability and safety of the final cutting result, and further optimizes the cutting efficiency and material utilization.

[0020] Preferably, step S4 comprises the following steps: Step S41: performing control feature analysis on the compensation cutting path to obtain compensation control features; performing curvature extraction on the compensation cutting path to obtain path curvature data; Step S42: calculating the speed change rate of the path curvature data to obtain curvature speed change rate data; mapping the control parameters of the compensation control characteristics according to the curvature speed change rate data to obtain the change control parameters; Step S43: Perform multi-objective optimization modeling based on the change control parameters to generate a multi-objective optimization model; perform parameter replacement on the change control parameters according to the multi-objective optimization model to obtain an optimization control instruction; Step S44: Perform laser cutting control according to the optimized control instruction.

[0021] The present invention can identify key control parameters in the cutting process by implementing control feature analysis on the compensation cutting path, and the obtained compensation control features provide a basis for path optimization. The implementation of curvature extraction on the compensation cutting path can quantify the curvature of the cutting path, and the generated path curvature data provides a basis for speed adjustment. The implementation of speed change rate calculation on the path curvature data can evaluate the speed change characteristics during the cutting process, and the obtained curvature speed change rate data provides support for control parameter mapping. The implementation of control parameter mapping of the compensation control features according to the curvature speed change rate data can generate change control parameters that meet cutting requirements. The implementation of multi-objective optimization modeling based on the change control parameters can consider multiple optimization objectives at the same time, and the generated multi-objective optimization model provides a framework for improving the cutting effect. The implementation of parameter replacement of the change control parameters according to the multi-objective optimization model can generate targeted optimization control instructions. Finally, the implementation of laser cutting control according to the optimization control instructions can realize precise cutting operations, improve the efficiency and quality of lens component cutting, ensure the stability and safety of the cutting process, and further enhance the intelligence level and automation capabilities of the system.

[0022] The present invention also provides a laser cutting control system for a lens assembly, which is used to execute the laser cutting control method for a lens assembly as described above. The laser cutting control system for a lens assembly comprises: The lens reconstruction module is used to scan the lens assembly in azimuth to obtain the lens assembly point cloud data; remove abnormal points from the lens assembly point cloud data to obtain filtered lens assembly point cloud data; perform three-dimensional reconstruction on the filtered lens assembly point cloud data to generate a lens assembly model; The spatiotemporal arrangement module is used to obtain cutting task data; identify cutting targets on the cutting task data to obtain cutting target data; and perform optimal spatiotemporal arrangement of cutting on the lens component model based on the cutting target data to generate an optimal cutting blueprint; A dynamic compensation module is used to plan a cutting path for the lens component model based on the preferred cutting blueprint to obtain a candidate cutting path; and dynamically compensate the candidate cutting path to obtain a compensated cutting path; The parameter replacement module is used to map the control parameters according to the compensation cutting path to obtain the changed control parameters; and perform parameter replacement based on the changed control parameters to execute the intelligent laser cutting control method.

[0023] The present invention can realize a comprehensive scan of the lens assembly and obtain detailed point cloud data through the implementation of the lens reconstruction module. The implementation of abnormal point elimination can improve the accuracy of the data. The generated filtered lens assembly point cloud data provides a reliable basis for subsequent three-dimensional reconstruction. The implementation of three-dimensional reconstruction can generate an accurate lens assembly model, providing a clear reference for the cutting process. The implementation of the spatiotemporal arrangement module can effectively obtain cutting task data and realize accurate identification of cutting targets. The implementation of the spatiotemporal arrangement for optimal cutting of the lens assembly model based on the cutting target data can optimize the cutting plan. The generated optimal cutting blueprint provides clear guidance for cutting path planning. The implementation of the dynamic compensation module can accurately plan the cutting path. The generation of candidate cutting paths provides a basis for subsequent compensation. The implementation of dynamic compensation can ensure the accuracy and effectiveness of the cutting path. The implementation of the parameter replacement module can convert the compensated cutting path into an operable control parameter. The generation of the change control parameter lays the foundation for the precise control of laser cutting. The implementation of parameter replacement based on the change control parameter can perform efficient intelligent laser cutting control, thereby improving the accuracy and efficiency of lens assembly processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A schematic flow chart of the steps of a laser cutting control method for a lens assembly; Figure 2 Detailed implementation flow chart of step S2; Figure 3 is a schematic diagram of a detailed implementation process of step S3; The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0025] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0026] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0027] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0028] To achieve this, please refer to Figures 1 to 3 , a laser cutting control method for a lens assembly, comprising the following steps: Step S1: scanning the lens assembly in all directions to obtain lens assembly point cloud data; removing abnormal points from the lens assembly point cloud data to obtain filtered lens assembly point cloud data; performing three-dimensional reconstruction on the filtered lens assembly point cloud data to generate a lens assembly model; Step S2: obtaining cutting task data; performing cutting target recognition on the cutting task data to obtain cutting target data; performing optimal cutting time-space arrangement on the lens component model based on the cutting target data to generate an optimal cutting blueprint; Step S3: planning a cutting path for the lens component model based on the preferred cutting blueprint to obtain a candidate cutting path; dynamically compensating the candidate cutting path to obtain a compensated cutting path; Step S4: mapping control parameters according to the compensation cutting path to obtain changed control parameters; performing parameter replacement based on the changed control parameters to execute the intelligent laser cutting control method.

[0029] The present invention realizes accurate digital capture of the surface geometric features of lens components through omni-directional high-precision scanning technology, introduces an intelligent recognition and elimination algorithm of abnormal points based on machine learning, significantly improves the quality and reliability of point cloud data, and transforms discrete point clouds into continuous 3D geometric models through advanced 3D reconstruction technology, breaking through the limitations of traditional reconstruction methods and laying an accurate data foundation for subsequent cutting processing. At the same time, a deep learning target recognition model is constructed in the data analysis link of cutting tasks to accurately understand the processing requirements of complex lens components, introduce a multi-dimensional optimization algorithm for cutting time and space arrangement, comprehensively consider multiple constraints such as material utilization, processing accuracy and cutting efficiency, realize intelligent planning and dynamic adjustment of cutting paths, and further base on An intelligent path planning algorithm is developed to optimize the cutting blueprint. By introducing multiple constraints and dynamic path generation technology, the cutting path of the lens component is accurately controlled, and a machine learning dynamic compensation model is constructed to perceive the slight deformation and stress distribution of the cutting process in real time, adaptively adjust the cutting path, and effectively suppress the accumulation of processing errors and material deformation. Finally, by constructing a control parameter mapping model based on machine learning, the precise parameter conversion of the compensated cutting path is realized, and an adaptive parameter substitution algorithm is introduced to dynamically adjust the laser cutting control parameters according to the real-time processing environment, breaking through the traditional static parameter setting mode, significantly improving the intelligence level of the laser cutting process, and realizing precise control of the lens component processing process, comprehensively improving the cutting quality and production efficiency.

[0030] In the embodiment of the present invention, refer to Figure 1 , is a schematic flow chart of the steps of a laser cutting control method for a lens assembly of the present invention. In this example, the laser cutting control method for a lens assembly includes the following steps: Step S1: scanning the lens assembly in all directions to obtain lens assembly point cloud data; removing abnormal points from the lens assembly point cloud data to obtain filtered lens assembly point cloud data; performing three-dimensional reconstruction on the filtered lens assembly point cloud data to generate a lens assembly model; In this embodiment, the lens assembly is scanned in all directions by using a laser scanner or a structured light scanning device. The lens assembly is fixed by a high-speed rotating platform, and the point cloud data of the lens assembly is obtained by the scanning device. After the data is obtained, it is preliminarily processed, and the point cloud data is denoised by a filter to remove abnormal points. These abnormal points are caused by equipment errors, surface unevenness and other factors. The point cloud data after filtering is more accurate. Next, a three-dimensional reconstruction algorithm is used, such as a Delaunay triangulation algorithm, to perform three-dimensional reconstruction on the filtered point cloud data to generate a three-dimensional model of the lens assembly. The model accurately reflects the shape, edge, thickness and tiny details of the lens, which is convenient for subsequent cutting path planning and cutting quality control.

[0031] Step S2: obtaining cutting task data; performing cutting target recognition on the cutting task data to obtain cutting target data; performing optimal cutting time-space arrangement on the lens component model based on the cutting target data to generate an optimal cutting blueprint; In this embodiment, the current cutting task data is obtained from the production line database, including the cutting size, shape and cutting surface type required by the user. After obtaining the task data, a deep learning model or a pattern recognition algorithm is used to identify the cutting target, extract the key features of the cutting target, such as the cutting area, cutting shape, angle, etc., and accurately calibrate the target data according to the size and thickness of the lens component. Then, based on these cutting target data, a heuristic algorithm (such as a simulated annealing algorithm) is used to optimize the spatiotemporal arrangement of the cutting of the lens component model. Through multiple iterative optimizations, it is ensured that the cutting area accurately matches the lens component. The generated optimal cutting blueprint provides the best cutting solution, maximizes cutting efficiency and reduces waste.

[0032] Step S3: planning a cutting path for the lens component model based on the preferred cutting blueprint to obtain a candidate cutting path; dynamically compensating the candidate cutting path to obtain a compensated cutting path; In this embodiment, based on the generated preferred cutting blueprint, a path planning algorithm, such as the A* algorithm (A-starAlgorithm), is used to plan the cutting path of the lens component model. By calculating the optimization method of the cutting path, several candidate cutting paths are obtained. These paths will try to avoid damage to the lens component and increase the cutting speed. Next, these candidate cutting paths are dynamically compensated. The compensation parameters include laser focus position, cutting speed, cutting depth, etc. Through real-time feedback data, combined with factors such as temperature during the cutting process and lens material properties, the cutting path is dynamically adjusted to generate a compensated cutting path to ensure accurate processing of the lens material during the cutting process and avoid problems such as thermal expansion or focus offset.

[0033] Step S4: mapping control parameters according to the compensation cutting path to obtain changed control parameters; performing parameter replacement based on the changed control parameters to execute the intelligent laser cutting control method.

[0034] In this embodiment, according to the compensated cutting path, the laser cutting control system is used to perform control parameter mapping. First, according to the compensated cutting path, the corresponding control parameters are obtained. The control parameters include laser power, pulse frequency, cutting speed, focal position, etc. These parameters will be adjusted according to the thickness, material and real-time temperature changes of the lens. Then, these control parameters are mapped to the control system of the cutting machine through a calculation formula to generate change control parameters. Then, according to these change control parameters, the parameters of the cutting equipment are replaced to ensure that the laser power, cutting speed, focal point of the laser beam, etc. can be adjusted in real time to perform the cutting operation, thereby performing precise laser cutting on the lens to achieve the predetermined cutting effect and ensure that the cutting surface is smooth and the size is accurate.

[0035] Preferably, step S1 comprises the following steps: Step S11: performing omnidirectional laser scanning on the lens assembly to obtain original point cloud data; performing spatial registration on the original point cloud data to obtain lens assembly point cloud data; Step S12: performing local density analysis on the lens component point cloud data to obtain point cloud density distribution data; performing outlier detection based on the point cloud density distribution data to generate an abnormal point set; Step S13: performing an abnormal elimination process on the abnormal point set to generate point cloud data of the filter lens component; performing a normal vector estimation process on the point cloud data of the filter lens component to obtain a normal vector field; Step S14: Perform three-dimensional reconstruction on the filtered lens component point cloud data based on the normal vector field to generate a lens component model.

[0036] In this embodiment, a laser scanning device is used to perform an all-round scan on the lens assembly. The scanner uses a laser beam to perform a high-frequency scan on each surface of the lens. The lens assembly is fixed on a rotating platform during scanning. The rotation angle and speed of the platform are adjusted according to the size of the lens to ensure the integrity and accuracy of the scanned data. After the original point cloud data is scanned, the point cloud is processed using a spatial registration algorithm. The registration process uses an Iterative Closest Point (ICP) algorithm to accurately align data from scans at different angles to ensure that all scanned point clouds can be matched in the same coordinate system. The ICP algorithm is used to iteratively calculate and adjust the relative position of the point cloud data to obtain unified lens assembly point cloud data. The registered data is a clear and accurate point cloud model. Based on the lens assembly point cloud data, a local density analysis is performed. First, the point cloud data is spatially segmented. The point cloud data is divided into multiple small grid units using a voxel grid method (Voxel Grid). The number of point clouds in each grid is calculated to obtain the local point cloud density of each grid. Then, based on the local point cloud density data, a K-nearest neighbor algorithm (K-Nearest Neighbor) is used to calculate the local point cloud density of each grid. Neighbor) evaluates the local density of each point, and determines the density distribution of the point cloud data by calculating the number of neighboring points within a certain range around each point. Then, the density values ​​are counted, and the areas with obviously high or low density values ​​are marked to identify the existing abnormal points. On this basis, the abnormal value detection method, such as the Z-score-based detection method, is used to determine the points with abnormal density and mark them as abnormal point sets. The abnormal point sets are eliminated, and the RANSAC algorithm (Random Sample Consensus) is used to identify the points that do not meet the surface flatness and edge characteristics of the lens from the abnormal point set. For the local anomalies of the point cloud data, the RANSAC algorithm is used for fitting, and the points with large fitting errors are eliminated to ensure that the eliminated point cloud data are consistent with the actual surface morphology. Next, the normal vector of the filtered point cloud data is estimated, and PCA (Principal Component Analysis, principal component analysis) method, calculates the normal vector of each point, and extracts the principal component as the normal vector by calculating the covariance matrix of the local neighborhood of each point. The estimated processing of the normal vector can provide the surface normal direction of each point, and further provide accurate geometric information for subsequent 3D reconstruction. The normal vector field is the normal vector data set of each point. Based on the normal vector field, the filtered lens component point cloud data is reconstructed in 3D. First, according to the spatial relationship between the normal vector of each point and the local point cloud, the point cloud is interpolated using the Poisson Surface Reconstruction algorithm, and a smooth 3D surface model is generated through Poisson reconstruction. The algorithm iteratively solves the optimal solution, combines the normal vector with the point cloud data, and generates a 3D mesh model that conforms to the actual surface features.Next, the grid is refined using the Multigrid Method to obtain a high-resolution three-dimensional lens model. The generated lens component model has high-precision surface features and can accurately reflect the geometric shape and details of the lens, facilitating subsequent cutting path planning and control.

[0037] Preferably, step S2 comprises the following steps: Step S21: obtaining cutting task data; classifying cutting targets on the cutting task data to obtain target classification results; Step S22: performing cutting target recognition on the cutting task data based on the target classification result to generate cutting target data; Step S23: performing feature matching on the lens component model according to the cutting target data to obtain a candidate cutting target model; Step S24: performing arrangement constraint processing on the candidate cutting models to obtain the constrained cutting conditions; performing optimal spatiotemporal arrangement processing on the candidate cutting models based on the constrained cutting conditions to generate an optimal cutting blueprint.

[0038] In this embodiment, after obtaining the cutting task data, a high-precision sensor is used to collect the size, shape and surface feature information of the lens component. The cutting task data contains complete geometric data about the lens component, including information such as the thickness, curvature, and surface texture of the lens. The data is collected by a laser scanning device and stored in a point cloud data format. The data is then preprocessed, and a point cloud filtering technology (for example, a Statistical Outlier Removal (SOR) filter) is used to remove noise point cloud data to obtain more accurate cutting task data. Then, according to the specific requirements of the task, a classification algorithm in machine learning is used to classify the cutting target of the cutting task data. The classification basis includes the cutting area, the shape and position of the target. A support vector machine (SVM) algorithm is used to determine the classification model through the training set data. The classification result includes the cutting area and the non-cutting area. The target classification result is output. Based on the target classification result, the cutting task data is identified for cutting targets. First, the classification result is further processed to match the cutting task data with the geometric features corresponding to the target area. An edge detection-based algorithm (such as Canny EdgeDetection identifies the specific edge of the cutting target, and the edge information is further associated with the cutting target area in the point cloud data. The cutting target area is registered using the iterative closest point (ICP) algorithm to accurately determine the spatial position and direction of the cutting target. Based on this recognition process, the cutting target data is generated. The data contains the specific cutting area coordinate information, area shape, cutting depth and other cutting feature information. Based on the cutting target data, the lens component model is feature matched. First, feature points are extracted from the cutting target data. These feature points include the vertices, curvature points, surface normal vectors, etc. of the target edge. Fast point cloud registration algorithms (such as Fast Point During the matching process, the lens component model is compared with the feature points in the cutting target data one by one, and the Euclidean distance between each pair of points is calculated. Based on the matching results, the candidate target model that best meets the cutting requirements is selected. The candidate cutting target model is displayed graphically and provides corresponding geometric size, shape, position and other information. The feature matching process is based on an iterative optimization algorithm to finally obtain the candidate cutting target model. According to the candidate cutting target model, the arrangement constraint processing is performed on it. First, the arrangement constraint conditions are defined. These constraints include the spatial position of the cutting target area, the boundary of the lens component, the parallelism of the cutting path, and the cutting angle. Based on these constraints, the simulated annealing algorithm is used to optimize the arrangement of the candidate cutting targets.The simulated annealing algorithm gradually approaches the global optimal solution by continuously adjusting the position and direction of the cutting target on the lens component, and obtains the optimal arrangement scheme of the candidate cutting model. Next, based on the optimized arrangement conditions, the spatiotemporal constraint algorithm (for example, an arrangement algorithm based on dynamic programming) is used to optimize the spatiotemporal arrangement of the cutting targets. During the processing, the geometric characteristics of the lens component and the cutting path planning requirements are combined to dynamically adjust the spatial position to ensure that the arrangement of each cutting target will not conflict in the physical space. Finally, the optimal cutting blueprint is generated. The blueprint includes the layout information of the final cutting path to meet the cutting accuracy and timeliness requirements.

[0039] Preferably, step S24 comprises the following steps: Step S241: performing boundary condition constraints on the candidate cutting model to obtain boundary condition constraint data; performing material utilization rate constraints on the candidate cutting model to generate material constraint data; Step S242: performing constraint integration processing on the boundary condition constraint data and the material constraint data based on the preset constraint weight data to obtain the constraint cutting conditions; Step S243: performing optimal cutting target matching on the candidate cutting models based on the constraint cutting conditions to obtain the optimal cutting sequence; performing adjacent distance calculation on the optimal cutting sequence to generate a target spacing matrix; Step S244: performing a heat impact analysis according to the preferred cutting sequence to obtain heat accumulation distribution data; performing conflict area identification on the heat accumulation distribution data to generate a conflict timing point set; performing time reallocation on the preferred cutting sequence based on the conflict timing point set to generate timing optimization data; Step S245: Calculate the safety distance of the target spacing matrix to obtain gap safety distance data; perform spacing adjustment processing on the target spacing matrix according to the gap safety distance to generate spacing optimization data; Step S246: Integrate and map the timing optimization data and the spacing optimization data to generate a preferred cutting blueprint.

[0040] In this embodiment, when the boundary condition constraints are imposed on the candidate cutting model, the geometric boundary of the candidate cutting target model is firstly accurately analyzed, the shape of the cutting model is extracted by using a three-dimensional model processing tool, and its edge contour is analyzed and connected with the working area of ​​the cutting equipment to ensure that the layout of the cutting path is carried out within the working range of the equipment. The specific boundary constraint data includes the size and shape of the cutting target and its distance from the equipment boundary. The position is adjusted by using a geometric transformation algorithm to meet the motion range requirements of the cutting equipment. Then, the material utilization rate of the candidate cutting model is constrained. By analyzing the ratio between the material thickness of the lens component and the cutting area, a linear The planning algorithm calculates the maximum material utilization rate. By optimizing the distribution of cutting targets, the overlap between each cutting area is minimized to reduce material waste, generate material constraint data, obtain the preset boundary constraint and material constraint weight values, and adjust the weight value according to the priority of the cutting task. The weight of the boundary condition constraint mainly considers the equipment movement range and cutting path safety, and the weight of the material constraint tends to optimize the cutting area and maximize the material utilization rate. The two constraint data are integrated by the weighted summation method to obtain the comprehensive constraint data. In the integration process, a multi-objective optimization algorithm (for example, particle swarm optimization algorithm, PSO, Particle Swarm Optimization is used to deal with the conflict between boundary constraint data and material constraint data. The result after optimization ensures that the arrangement of each cutting target not only meets the equipment boundary requirements, but also maximizes the material utilization rate. Finally, the constraint cutting conditions are obtained. The constraint cutting conditions are analyzed to ensure that the geometric characteristics of each cutting target match the cutting capacity of the equipment. The dynamic programming (DP) algorithm is used to compare each candidate cutting target with the defined cutting conditions one by one, and the matching degree of each cutting target is calculated. Finally, the target with the highest matching degree is selected as the final cutting target to generate the preferred cutting sequence. Then, the adjacent distance of the preferred cutting sequence is calculated, and the distance between the cutting targets is calculated using the Euclidean distance to obtain the target spacing matrix. The target spacing matrix contains the shortest distance between every two cutting targets. Considering the thermal physical properties of the lens component, the heat distribution in the laser cutting process is simulated by the heat conduction model to simulate the heat accumulation of the laser in the cutting process and its influence on the material. The finite element analysis method (Finite Element Analysis (FEA) calculates the heat accumulation of each cutting area and obtains the heat accumulation distribution data. Then, the heat accumulation distribution data is used to identify the conflicting areas. First, the threshold of the heat affected area is defined, and the thermal analysis software is used to identify the areas where the temperature exceeds the threshold. These areas will cause material deformation or damage, generate conflicting time sequence point sets, and record the timestamps of these conflicting areas.Based on the conflicting timing point set, the time of the preferred cutting sequence is reallocated. By adjusting the cutting order and cutting time, material damage caused by excessive heat accumulation is avoided. The timing optimization data is generated, and the safety distance standard is defined to ensure that there is enough space between the cutting targets to avoid mutual interference. The minimum spacing between the targets is calculated using a geometric algorithm and compared with the preset safety distance to obtain the gap safety distance data. Then, the target spacing matrix is ​​adjusted according to the gap safety distance. The spatial distribution of the cutting targets is adjusted using an optimization algorithm (for example, a genetic algorithm) to ensure that the spacing of each target meets the safety standard and avoids crossing or interference of the cutting path. The spacing optimization data is generated, and the timing optimization data is merged with the spacing optimization data. The two are integrated by a weighted average method to obtain the final preferred cutting blueprint. The data in the blueprint includes the timing arrangement of the cutting path and the spatial layout of the target spacing to ensure the smooth progress of the cutting process. Computer-Aided Design (CAD) is used to calculate the spacing matrix of the target. The Laser Cutting Design system visualizes the optimized cutting blueprint and eventually generates a complete laser cutting control diagram, which is transmitted to the laser cutting equipment to ensure that the equipment can perform cutting operations according to the optimized cutting path and sequence, achieving efficient and accurate cutting results.

[0041] Preferably, step S245 includes the following steps: Calculate the space requirement of the target spacing matrix to obtain the space requirement matrix; perform effective utilization analysis based on the space requirement matrix to obtain effective utilization space data; Extract cutting gaps based on effective use of spatial data and target spacing matrix to generate cutting gap data; Calculate the safety distance of the cutting gap data to obtain the gap safety distance data; According to the gap safety distance data, the target spacing matrix is ​​mapped to the safety reserved space to generate the reserved space data required for safety; The target spacing matrix is ​​adjusted based on the reserved space data required for safety to generate spacing optimization data.

[0042] In this embodiment, the geometric shape and size data of each cutting target are obtained, and the boundary of the cutting target is accurately modeled. The target spacing matrix is ​​converted into grid data using a spatial gridding method, and the position and size of each cutting target are mapped to a three-dimensional spatial grid. The space required for each cutting target is calculated using a spatial analysis algorithm to obtain the free space between the target and other targets, and the spatial range occupied by each target is calculated, and a spatial demand matrix is ​​further generated. The spatial demand matrix reflects the spatial demand of each cutting target in the entire cutting area. The value of each element in the matrix represents the degree of spatial demand at the position. The data is organized into a multidimensional matrix, showing the spatial demand of each cutting target and its surrounding area. The data in the spatial demand matrix is ​​screened to identify the free part in the cutting area and calculate its utilization efficiency. The region growing algorithm is used. Algorithm) divides the free space, evaluates the availability of each area, and further distinguishes the effective space from the invalid space by calculating the space utilization rate in each area, screens out the space areas that can be effectively utilized, and uses the optimization algorithm (for example, the maximum minimum area method) to calculate the maximum available space of these areas, and finally generates effective space utilization data, indicating which parts of the given cutting area can be efficiently utilized and which parts have potential space waste. The effective space is further divided, and the contour detection algorithm (for example, Canny edge detection) is used to identify the gap area between the cutting targets. The size, shape and position of the gap area are confirmed by the step-by-step scanning method to ensure that the extracted gap meets the actual working requirements of the cutting equipment. In this process, the minimum size of each gap is calculated, and the gap size is ensured to be suitable for the requirements of the laser cutting path according to the target spacing matrix. Finally, the identified effective gap area is saved as cutting gap data, which contains information such as the gap position, size, shape, etc. between each cutting target. A safety distance threshold is defined. According to the physical characteristics and safety requirements of the cutting equipment, a geometric distance algorithm (for example, Manhattan distance method) is used to calculate the cutting gap. It is checked whether the actual distance between each cutting target meets the preset safety standard. The safety distance of each cutting target gap is calculated to determine which gaps meet the preset safety requirements and which need to be adjusted. The gap safety distance data is generated. The data contains the minimum safety distance between each pair of cutting targets and shows which gaps do not meet the safety requirements and need further adjustment. The gap safety distance of each cutting target is compared with its surrounding area. The gap safety distance data is combined with the target spacing matrix through a spatial mapping algorithm. The minimum constraint priority algorithm is used to calculate the amount of space that should be reserved around each target. According to the preset safety standards, it is ensured that the gaps between targets meet the safety requirements.The data of reserved space required for safety is generated through mapping. The data includes the minimum safe space that needs to be reserved around each target, ensuring that the space around each target is large enough to avoid interference during the cutting process. According to the data of reserved space required for safety, the distribution of cutting targets in space is adjusted, and the distance between cutting targets is dynamically adjusted using the greedy algorithm. By optimizing the distance between each pair of targets, the space utilization efficiency is maximized, while ensuring that the safe distance between each target meets the requirements. After adjustment, the spacing optimization data is generated. The data reflects the target spacing after the safety reserved space calculation, so that the target arrangement meets the safety requirements and can efficiently use the space. The spacing optimization data finally generated provides a safe and efficient spatial layout for cutting path planning. ,

[0043] Preferably, step S3 comprises the following steps: Step S31: performing path constraint mapping on the lens component model based on the preferred cutting blueprint to obtain path constraint data; Step S32: planning a cutting path for the lens component model according to the path constraint data to obtain a candidate cutting path; Step S33: performing cutting simulation based on the candidate cutting path to obtain simulated component cutting data; performing deformation prediction on the simulated component cutting data to generate cutting deformation prediction data; Step S34: performing stress distribution calculation based on the cutting deformation prediction data to generate stress distribution data; performing stress path compensation according to the stress distribution data to obtain a compensated cutting path.

[0044] In this embodiment, the lens component model is scanned by the calculated preferred cutting blueprint, and the cutting path in the blueprint is mapped into the three-dimensional space of the lens component model. The cutting path is calibrated in the cutting area using a spatial mapping algorithm (such as a Bézier curve mapping method) to determine the spatial position, shape and change trend of each cutting path. During the mapping process, accurate spatial alignment is performed for each path node according to the path requirements of the cutting blueprint and the geometric features of the lens component model to obtain the constraint conditions of each cutting path, thereby ensuring that the path is completely consistent with the geometric shape of the lens component. The path constraint data includes the geometric information of each cutting path and its motion trajectory during the cutting process, as well as the minimum safe distance between the paths. The sheet component model is divided into multiple cutting areas. Each cutting area is independently planned according to the path constraint data. The shortest path algorithm based on graph theory (such as Dijkstra algorithm) is used to determine the cutting path of each cutting area. The motion trajectory and working range of the cutting equipment are taken into account during the planning process. The paths between all cutting areas are globally optimized to ensure the continuity and cutting efficiency of the path. The curvature and intersection of the path are strictly controlled during planning. A set of candidate cutting paths are generated. These paths reflect the best path plan from the starting point to the end point of the cutting, and each candidate cutting path is consistent with the requirements in the path constraint data. The candidate cutting paths are simulated and finite element analysis (Finite Element The FEA (Full Field Analysis) technology is used to simulate the cutting process of lens components. Through the refined simulation of candidate cutting paths, the cutting data of simulated components are generated. The simulation data includes the cutting trajectory, cutting depth, cutting speed and other parameters of each path. Combined with the mechanical properties of the material, the force and heat of the cutting process are analyzed during the simulation process, and the stress distribution on each cutting path is calculated to predict the deformation that will occur. Based on the cutting simulation data, the stress-strain relationship of the material is used to further predict the cutting deformation. The lens deformation that occurs during the cutting process is predicted through the thermal expansion and stress concentration model, and the cutting deformation prediction data is generated. The data includes the deformation amplitude, deformation direction and deformation of each cutting path. The stress concentration area of ​​the lens component is modeled by the finite element method according to the cutting deformation prediction data. The stress distribution on the surface and inside of the material during the cutting process is simulated, and the stress distribution data on each cutting path during the cutting process is calculated. The simulation data of the cutting path is combined with the physical properties of the material to obtain the stress intensity and stress direction on each path and its influence on the lens deformation. Based on the stress distribution data, the cutting path is adjusted by the reverse compensation method. The stress path compensation is performed for the stress concentration area of ​​each path. During the compensation process, adjustments are made according to the rebound characteristics of the material and the geometric shape of the cutting path to ensure that the cutting path can effectively offset the stress changes generated during the cutting process and generate a compensated cutting path.The compensated path can more accurately reduce the deformation during the cutting process, ensuring the final cutting accuracy and quality of the lens components.

[0045] Preferably, step S33 includes the following steps: Acquire laser cutting heat source data; simulate the cutting temperature field of the laser cutting heat source data to obtain simulated temperature field distribution; Perform cutting cooling field simulation on laser cutting heat source data to generate simulated cooling field distribution; perform stress calculation on candidate cutting paths to obtain stress field data; Perform cutting simulation on the candidate cutting path based on the simulated temperature field distribution, simulated cooling field distribution and stress field data to obtain simulated component cutting data; Perform local strain calculation on the simulated component cutting data to obtain the local strain field; perform thermal deformation evaluation based on the local strain field to generate local thermal deformation data; Perform global deformation accumulation calculation on the simulated component cutting data according to the local thermal deformation data to obtain the accumulated deformation field; Deformation prediction is performed based on the accumulated deformation field to generate cutting deformation prediction data.

[0046] In this embodiment, the power output of the laser, the focusing mode of the laser beam, the laser pulse frequency, the cutting speed, the beam irradiation area and other parameters of the laser are recorded by the sensor in the laser cutting system, and these parameters are combined with the real-time monitored temperature data. The intensity and distribution of the laser heat source during the laser cutting process are calculated using a heat source model (such as a Gauss distribution model) to obtain laser heat source data. These data describe the heat input and distribution when the laser beam contacts the lens component. Afterwards, the laser heat source data is simulated for the cutting temperature field using the heat conduction equation and the finite element method (FEM). The temperature field distribution on the surface and inside of the lens is obtained by a numerical solution method. During the simulation process, high-precision meshing technology is used to perform a detailed spatial distribution analysis of the temperature field change, and the temperature data of the lens surface and different depths are output to obtain simulated temperature field distribution data. Based on the laser cutting heat source data, a cooling field model is established in consideration of the cooling methods in the cutting process, such as spray cooling, gas cooling or contact cooling. The heat exchange efficiency and temperature change when the cooling fluid contacts the lens surface are calculated by a numerical simulation method. Fluid dynamics simulation (Computational Fluid Dynamics, The cooling medium flow state is analyzed by CFD to obtain the velocity field, pressure field and temperature field data of the cooling medium. The calculation of the cooling field takes into account the temperature, flow rate and cooling efficiency of the cooling medium. Based on these parameters, the influence of cooling on the temperature distribution during the cutting process is calculated to generate simulated cooling field distribution data. The cooling field data includes the contact effect between the cooling medium and the lens at different time points. The simulated cooling field can truly reflect the distribution of the cooling medium during the cutting process. According to the simulated temperature field distribution data and the simulated cooling field distribution data, combined with the mechanical parameters of the lens material such as the thermal expansion coefficient and elastic modulus, the finite element analysis method (Finite Element Analysis, FEA) is used to calculate stress, a thermal-mechanical coupling model is established, and the stress distribution caused by thermal expansion during laser cutting is calculated. Considering the rapid change of temperature field during laser cutting, the spatiotemporal distribution of thermal stress is calculated, and stress field data is generated. The stress field data includes the stress intensity, direction and distribution of each point on the cutting path. By performing stress analysis on different cutting paths, the stress concentration area that will appear in the cutting process is obtained, and the temperature field distribution, cooling field distribution and stress field data are simulated. The finite element analysis method coupled with multiple physical fields is used to simulate the cutting of lens components. By coupling models such as heat conduction, fluid mechanics and mechanical stress, the material response on the cutting path is simulated. The temperature change, cooling effect and stress change of the material are taken into account during the simulation process, and the cutting data of the simulated component is generated. The data includes the temperature change, stress distribution and plastic deformation of the material during the cutting process. Through different simulations of the cutting path,The cutting quality, deformation and potential defective areas on each path can be obtained. According to the temperature and stress data obtained from the simulated component cutting data, the strain in the cutting area is calculated using the strain-stress relationship (such as Hooke's law), with special attention paid to the strain distribution in the area near the cutting path. The strains in different areas are accurately calculated using the finite element method to obtain local strain field data. The local strain field data reflects the slight deformation of the material near the cutting path during the cutting process, and can accurately describe the material strain in the cutting area, including elastic strain and plastic strain. According to the simulation results, areas with large local strains are identified, which are potential risk points for cutting deformation. Based on the local strain field data, combined with the thermal expansion characteristics and deformation behavior of the material, the thermal deformation is evaluated. The thermal-mechanical coupling analysis model is used to combine the temperature field with the strain field to calculate the deformation caused by temperature changes in the local area, with special attention paid to areas with large thermal expansion. Based on the thermal expansion coefficient of the material and the stress-strain model, the thermal deformation behavior of different areas during the cutting process is simulated. Generate local thermal deformation data, which includes the thermal deformation amount, deformation direction and timeliness of the cutting area. In the global deformation accumulation calculation, based on the local thermal deformation data, the thermal deformation amounts of each local area are accumulated by numerical integration to obtain the global deformation field. Considering the influence of the deformation of each area on the overall structure, the total deformation amount of the entire lens assembly during the cutting process is calculated by combining the temperature field, cooling field and stress field on the cutting path to obtain the cumulative deformation field data. The cumulative deformation field reflects the deformation of the entire lens assembly during the cutting process and can accurately display the deformation trend of each area. Based on the cumulative deformation field data, the thermal-mechanical coupling model is used to predict the deformation of the cutting path, with special attention paid to the overall or local deformation that occurs during the cutting process. The prediction algorithm is used to analyze the deformation trends of different cutting paths to generate cutting deformation prediction data. The prediction data includes the deformation amount and deformation direction that will occur during the cutting process and its influence on the quality of the final product. The prediction results will provide guidance for cutting path optimization to ensure that the cutting accuracy and quality meet the requirements.

[0047] Preferably, step S34 includes the following steps: Perform stress tensor analysis on the cutting deformation prediction data to obtain stress component data; Perform distribution calculation on stress component data to generate stress distribution data; Calculate the compensation amount for the stress distribution data to obtain the path compensation data; The simulated component cutting data is compensated and superimposed according to the path compensation data to obtain a compensated cutting path.

[0048] In this embodiment, based on the temperature field, cooling field and stress field information in the cutting deformation prediction data, a finite element analysis method (Finite Element Analysis, FEA) performs a detailed analysis of the stress of each cutting path, calculates the stress distribution of the lens caused by temperature changes, material expansion and laser beam interaction during the laser cutting process, and uses the mathematical model of stress tensor (for example, Cauchy stress tensor) to decompose the stress in each cutting area to obtain stress components along different directions, including normal stress and shear stress. By solving the principal stress value and principal direction of the tensor, the stress component data of each point in the cutting process are obtained. Based on the stress component data, the finite element solution method is used to calculate the distribution of stress in space in each cutting area. The stress data of different areas are spatially weighted averaged by the numerical integration method, and the global stress distribution data is generated by combining the geometric shape of the cutting path, material properties, cooling effect and other factors. The stress distribution data includes the stress values ​​of each cutting path and its surrounding areas and the gradient change of stress. The calculated stress field reflects the stress state of the lens component during the laser cutting process. Based on the stress distribution data, considering the stress distribution and deformation trend of different areas, the thermal-mechanical coupling model is used to calculate the compensation amount. First, the high stress areas in the stress field are analyzed. These areas are often cutting deformation. The key point is to introduce the path correction factor, adjust the stress field data according to the stress size, deformation trend and specific requirements of the cutting path, and calculate the compensation amount of each cutting path. The compensation amount calculation involves the relationship between the stress adjustment value and the geometric position of the cutting path. The path is optimized by the integral method to generate path compensation data. The path compensation data contains the compensation amount of each cutting path. These compensation amounts will be applied to the simulated component cutting data in the subsequent steps, so as to make necessary corrections to the cutting path. The path compensation data is combined with the simulated component cutting data, and the compensation amount is applied to the coordinate points of the simulated cutting path by the numerical calculation method. For each point of each cutting path, the coordinates of the cutting path are adjusted according to the calculated compensation amount, and compensation is superimposed point by point. The adjusted path will take into account the actual stress and deformation to ensure that the influence of deformation and stress is effectively controlled during the actual cutting process. After this process, the generated compensated cutting path is the corrected path, which is suitable for actual laser cutting operations. These compensated cutting paths will be used for subsequent control signal generation to ensure that the lens component reaches the expected shape and size during the laser cutting process.

[0049] Preferably, step S4 comprises the following steps: Step S41: performing control feature analysis on the compensation cutting path to obtain compensation control features; performing curvature extraction on the compensation cutting path to obtain path curvature data; Step S42: calculating the speed change rate of the path curvature data to obtain curvature speed change rate data; mapping the control parameters of the compensation control characteristics according to the curvature speed change rate data to obtain the change control parameters; Step S43: Perform multi-objective optimization modeling based on the change control parameters to generate a multi-objective optimization model; perform parameter replacement on the change control parameters according to the multi-objective optimization model to obtain an optimization control instruction; Step S44: Perform laser cutting control according to the optimized control instruction.

[0050] In this embodiment, the compensation cutting path data is obtained, the path is represented by a coordinate point set, and the path is segmented using a numerical method. The second-order difference method is used to perform control feature analysis, and the characteristic data such as the curvature and smoothness of each path segment are calculated. A suitable algorithm is selected to model the shape of the path, and the shape changes of each segment in the path are analyzed, including bends, broken lines, etc. The control feature data is further extracted according to the geometric features of the path segment. The obtained control features include the curvature, tangent direction, angle, etc. of the path. The curvature of the path is then extracted. Based on the discrete point set of the path, a continuous difference algorithm is used to calculate the curvature of the path at each point. The curvature calculation result is the change rate of each path segment, the curvature extreme value, etc. By processing these data, the The curvature data of the path is obtained by performing first-order numerical differentiation on the curvature data of the path, and the curvature change rate of the path at each point is calculated. The curvature data is differentiated using a time step or a path length unit to obtain the speed change rate of each path point. These change rates reflect the change of the path curvature over time during the cutting process. Then, according to the calculated curvature speed change rate data, an adaptive control model is used to associate the curvature change rate with control parameters. The control parameters include laser power, cutting speed, etc. The curvature speed change rate is mapped to the corresponding control parameters through the control model to obtain the change control parameters. These parameters are used to adjust the laser power and cutting speed during the laser cutting process to ensure the accuracy of the cutting process. To ensure the accuracy and stability of the cutting process, the interpolation method is used in the mapping process to ensure the smooth transition of the control parameters of each path segment to avoid cutting errors caused by mutations. Multiple objective functions are considered, including cutting quality, cutting time, energy consumption, etc. Genetic algorithms are used to perform multi-objective optimization of the objective functions. The effects of different control parameter combinations are evaluated through the optimization algorithm. The initial population is generated according to the changed control parameters, and each population is evaluated to calculate the fitness value. The fitness value reflects the comprehensive effect of each group of parameters. After multiple optimization iterations, the parameter combination with higher fitness is selected for crossover and mutation operations, the parameter set is updated, and finally the optimized control model is generated. Then, the changed control parameters are replaced according to the optimization model, and the optimized The results are replaced with new control parameters to obtain the final optimized control instructions. These instructions are used in the subsequent laser cutting process. The optimization process ensures the balance between various goals in the cutting process, so that the cutting efficiency and quality are optimized. The optimized control instructions are transmitted to the laser cutting control system. The control system adjusts the laser power, frequency, cutting speed and other parameters according to the instructions to ensure that the cutting conditions of each path segment meet the preset requirements, monitors the temperature, pressure and other parameters in the cutting process in real time, and dynamically adjusts the laser output power through sensor feedback data to maintain the stability and accuracy of the cutting. During the cutting process, the control system also makes adaptive adjustments according to environmental changes. Finally, through closed-loop control, efficient and accurate cutting of the cutting path is ensured.And avoid any overheating or error, complete the laser cutting task of the entire lens assembly.

[0051] The present invention also provides a laser cutting control system for a lens assembly, which is used to execute the laser cutting control method for a lens assembly as described above. The laser cutting control system for a lens assembly comprises: The lens reconstruction module is used to scan the lens assembly in azimuth to obtain the lens assembly point cloud data; remove abnormal points from the lens assembly point cloud data to obtain filtered lens assembly point cloud data; perform three-dimensional reconstruction on the filtered lens assembly point cloud data to generate a lens assembly model; The spatiotemporal arrangement module is used to obtain cutting task data; identify cutting targets on the cutting task data to obtain cutting target data; and perform optimal spatiotemporal arrangement of cutting on the lens component model based on the cutting target data to generate an optimal cutting blueprint; A dynamic compensation module is used to plan a cutting path for the lens component model based on the preferred cutting blueprint to obtain a candidate cutting path; and dynamically compensate the candidate cutting path to obtain a compensated cutting path; The parameter replacement module is used to map the control parameters according to the compensation cutting path to obtain the changed control parameters; and perform parameter replacement based on the changed control parameters to execute the intelligent laser cutting control method.

[0052] The present invention can realize a comprehensive scan of the lens assembly and obtain detailed point cloud data through the implementation of the lens reconstruction module. The implementation of abnormal point elimination can improve the accuracy of the data. The generated filtered lens assembly point cloud data provides a reliable basis for subsequent three-dimensional reconstruction. The implementation of three-dimensional reconstruction can generate an accurate lens assembly model, providing a clear reference for the cutting process. The implementation of the spatiotemporal arrangement module can effectively obtain cutting task data and realize accurate identification of cutting targets. The implementation of the spatiotemporal arrangement for optimal cutting of the lens assembly model based on the cutting target data can optimize the cutting plan. The generated optimal cutting blueprint provides clear guidance for cutting path planning. The implementation of the dynamic compensation module can accurately plan the cutting path. The generation of candidate cutting paths provides a basis for subsequent compensation. The implementation of dynamic compensation can ensure the accuracy and effectiveness of the cutting path. The implementation of the parameter replacement module can convert the compensated cutting path into an operable control parameter. The generation of the change control parameter lays the foundation for the precise control of laser cutting. The implementation of parameter replacement based on the change control parameter can perform efficient intelligent laser cutting control, thereby improving the accuracy and efficiency of lens assembly processing.

[0053] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0054] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A laser cutting control method for a lens assembly, characterized in that: The following steps are involved: Step S1: scanning the lens assembly in all directions to obtain lens assembly point cloud data; removing abnormal points from the lens assembly point cloud data to obtain filtered lens assembly point cloud data; performing three-dimensional reconstruction on the filtered lens assembly point cloud data to generate a lens assembly model; Step S2: Obtain cutting task data; Perform cutting target recognition on the cutting task data to obtain cutting target data; Based on the cutting target data, the lens component model is optimally arranged in time and space to generate an optimal cutting blueprint; Step S3: planning a cutting path for the lens component model based on the preferred cutting blueprint to obtain a candidate cutting path; dynamically compensating the candidate cutting path to obtain a compensated cutting path; Step S4: mapping control parameters according to the compensation cutting path to obtain changed control parameters; performing parameter replacement based on the changed control parameters to execute the intelligent laser cutting control method.

2. The laser cutting control method for lens components according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: performing omnidirectional laser scanning on the lens assembly to obtain original point cloud data; performing spatial registration on the original point cloud data to obtain lens assembly point cloud data; Step S12: performing local density analysis on the lens component point cloud data to obtain point cloud density distribution data; performing outlier detection based on the point cloud density distribution data to generate an abnormal point set; Step S13: performing an abnormal elimination process on the abnormal point set to generate point cloud data of the filter lens component; performing a normal vector estimation process on the point cloud data of the filter lens component to obtain a normal vector field; Step S14: Perform three-dimensional reconstruction on the filtered lens component point cloud data based on the normal vector field to generate a lens component model.

3. The laser cutting control method for lens components according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: obtaining cutting task data; classifying cutting targets on the cutting task data to obtain target classification results; Step S22: performing cutting target recognition on the cutting task data based on the target classification result to generate cutting target data; Step S23: performing feature matching on the lens component model according to the cutting target data to obtain a candidate cutting target model; Step S24: performing arrangement constraint processing on the candidate cutting models to obtain the constrained cutting conditions; performing optimal spatiotemporal arrangement processing on the candidate cutting models based on the constrained cutting conditions to generate an optimal cutting blueprint.

4. The laser cutting control method for lens components according to claim 3, characterized in that: Step S24 includes the following steps: Step S241: performing boundary condition constraints on the candidate cutting model to obtain boundary condition constraint data; performing material utilization rate constraints on the candidate cutting model to generate material constraint data; Step S242: performing constraint integration processing on the boundary condition constraint data and the material constraint data based on the preset constraint weight data to obtain the constraint cutting conditions; Step S243: performing optimal cutting target matching on the candidate cutting models based on the constraint cutting conditions to obtain the optimal cutting sequence; performing adjacent distance calculation on the optimal cutting sequence to generate a target spacing matrix; Step S244: performing a heat impact analysis according to the preferred cutting sequence to obtain heat accumulation distribution data; performing conflict area identification on the heat accumulation distribution data to generate a conflict timing point set; performing time reallocation on the preferred cutting sequence based on the conflict timing point set to generate timing optimization data; Step S245: Calculate the safety distance of the target spacing matrix to obtain gap safety distance data; perform spacing adjustment processing on the target spacing matrix according to the gap safety distance to generate spacing optimization data; Step S246: Integrate and map the timing optimization data and the spacing optimization data to generate a preferred cutting blueprint.

5. The laser cutting control method for lens components according to claim 4, characterized in that: Step S245 includes the following steps: Calculate the space requirement of the target spacing matrix to obtain the space requirement matrix; perform effective utilization analysis based on the space requirement matrix to obtain effective utilization space data; Extract cutting gaps based on effective use of spatial data and target spacing matrix to generate cutting gap data; Calculate the safety distance of the cutting gap data to obtain the gap safety distance data; According to the gap safety distance data, the target spacing matrix is ​​mapped to the safety reserved space to generate the reserved space data required for safety; The target spacing matrix is ​​adjusted based on the reserved space data required for safety to generate spacing optimization data.

6. The laser cutting control method for lens components according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: performing path constraint mapping on the lens component model based on the preferred cutting blueprint to obtain path constraint data; Step S32: planning a cutting path for the lens component model according to the path constraint data to obtain a candidate cutting path; Step S33: performing cutting simulation based on the candidate cutting path to obtain simulated component cutting data; performing deformation prediction on the simulated component cutting data to generate cutting deformation prediction data; Step S34: performing stress distribution calculation based on the cutting deformation prediction data to generate stress distribution data; performing stress path compensation according to the stress distribution data to obtain a compensated cutting path.

7. The laser cutting control method for lens components according to claim 6, characterized in that: Step S33 includes the following steps: Acquire laser cutting heat source data; simulate the cutting temperature field of the laser cutting heat source data to obtain simulated temperature field distribution; Perform cutting cooling field simulation on laser cutting heat source data to generate simulated cooling field distribution; perform stress calculation on candidate cutting paths to obtain stress field data; Perform cutting simulation on the candidate cutting path based on the simulated temperature field distribution, simulated cooling field distribution and stress field data to obtain simulated component cutting data; Perform local strain calculation on the simulated component cutting data to obtain the local strain field; perform thermal deformation evaluation based on the local strain field to generate local thermal deformation data; Perform global deformation accumulation calculation on the simulated component cutting data according to the local thermal deformation data to obtain the accumulated deformation field; Deformation prediction is performed based on the accumulated deformation field to generate cutting deformation prediction data.

8. The laser cutting control method for lens components according to claim 6, characterized in that: Step S34 includes the following steps: Perform stress tensor analysis on the cutting deformation prediction data to obtain stress component data; Perform distribution calculation on stress component data to generate stress distribution data; Calculate the compensation amount for the stress distribution data to obtain the path compensation data; The simulated component cutting data is compensated and superimposed according to the path compensation data to obtain a compensated cutting path.

9. The laser cutting control method for lens components according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: performing control feature analysis on the compensation cutting path to obtain compensation control features; performing curvature extraction on the compensation cutting path to obtain path curvature data; Step S42: calculating the speed change rate of the path curvature data to obtain curvature speed change rate data; mapping the control parameters of the compensation control characteristics according to the curvature speed change rate data to obtain the change control parameters; Step S43: Perform multi-objective optimization modeling based on the change control parameters to generate a multi-objective optimization model; perform parameter replacement on the change control parameters according to the multi-objective optimization model to obtain an optimization control instruction; Step S44: Perform laser cutting control according to the optimized control instruction.

10. A laser cutting control system for lens components, characterized in that: For executing the laser cutting control method for a lens assembly as claimed in claim 1, the laser cutting control system for the lens assembly comprises: The lens reconstruction module is used to scan the lens assembly in azimuth to obtain the lens assembly point cloud data; remove abnormal points from the lens assembly point cloud data to obtain filtered lens assembly point cloud data; perform three-dimensional reconstruction on the filtered lens assembly point cloud data to generate a lens assembly model; The spatiotemporal arrangement module is used to obtain cutting task data; identify cutting targets on the cutting task data to obtain cutting target data; and perform optimal spatiotemporal arrangement of cutting on the lens component model based on the cutting target data to generate an optimal cutting blueprint; A dynamic compensation module is used to plan a cutting path for the lens component model based on the preferred cutting blueprint to obtain a candidate cutting path; and dynamically compensate the candidate cutting path to obtain a compensated cutting path; The parameter replacement module is used to map the control parameters according to the compensation cutting path to obtain the changed control parameters; and perform parameter replacement based on the changed control parameters to execute the intelligent laser cutting control method.

Images