Eyelash forming system and method based on laser cutting film

Abstract

The invention discloses an eyelash forming system and method based on a laser cutting film, relates to the technical field of eyelash forming processing, and aims to solve the technical problems of insufficient precision and hot melting deformation in a traditional eyelash process. Comprising a design input and modeling module, a process parameter intelligent optimization module, a cutting path intelligent planning module, a production process intelligent control module, a quality online monitoring and feedback module and a data management and analysis module. The process parameter intelligent optimization module is connected with the design input and modeling module and is used for receiving the initial design scheme and optimizing cutting process parameters of laser power, cutting speed, pulse frequency and light spot diameter by using a machine learning algorithm in combination with a laser cutting process knowledge base and thin film material characteristics; and the cutting path intelligent planning module is connected with the design input and modeling module and the process parameter intelligent optimization module. The eyelash processing device has the advantage of improving the eyelash processing qualification rate and the production efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of eyelash molding and processing, and more particularly to an eyelash molding system and method based on laser cutting of thin films. Background Art

[0002] In the field of eyelash production, traditional eyelash shaping processes face significant technical bottlenecks. Due to the delicate structure of eyelashes, which often include complex designs such as hollowing and multi-layer nesting, traditional methods have difficulty accurately controlling cutting accuracy and edge quality. This is especially true when dealing with thin film materials of varying thicknesses and melting points. Dimensional deviations, edge burrs, or material deformation during thermal melting are prone to occur, resulting in low product qualification rates and inefficient production. Furthermore, parameter adjustment in traditional processes relies on manual experience, making it difficult to adapt to diverse design requirements and environmental changes, and unable to achieve intelligent production control. This severely restricts the automation and refinement of eyelash manufacturing. In light of this, we propose an eyelash shaping system and method based on laser-cut thin films. Summary of the Invention

[0003] The purpose of the present invention is to provide an eyelash molding system and method based on laser cutting film to solve the technical problems of insufficient precision and thermal melting deformation in the traditional eyelash molding process.

[0004] To solve the above technical problems, the present invention provides the following technical solutions: an eyelash shaping system based on laser cutting film, comprising a design input and modeling module, a process parameter intelligent optimization module, a cutting path intelligent planning module, a production process intelligent control module, an online quality monitoring and feedback module, and a data management and analysis module; The process parameter intelligent optimization module is connected to the design input and modeling module, and is used to receive the initial design plan, combine the laser cutting process knowledge base with the film material properties, and use the machine learning algorithm to optimize the cutting process parameters such as laser power, cutting speed, pulse frequency, and spot diameter; The cutting path intelligent planning module is connected to the design input and modeling module and the process parameter intelligent optimization module, and is used to generate the laser cutting path using the path planning algorithm; The production process intelligent control module is connected with the cutting path intelligent planning module and the process parameter intelligent optimization module; The quality online monitoring and feedback module is connected to the production process intelligent control module; The data management and analysis module is connected with the design input and modeling module, the process parameter intelligent optimization module, the cutting path intelligent planning module, the production process intelligent control module, and the quality online monitoring and feedback module.

[0005] Preferably, the design input and modeling module has built-in graphics processing algorithms and model verification functions; The graphics processing algorithm combines adaptive multi-precision Bezier curve fitting technology with a deep learning edge refinement algorithm. For eyelash design areas of different complexity, the algorithm first identifies complex structural features through a deep learning algorithm, and then automatically adjusts the Bezier curve fitting accuracy parameters according to the feature type. At the same time, the graphics processing algorithm is also equipped with a feature point recognition and protection mechanism, which can identify key feature points in complex designs and protect them during the fitting process; In addition, the design input and modeling module is also equipped with a design-process linkage analysis unit, which performs correlation analysis between the constructed three-dimensional digital model and the process knowledge base of the process parameter intelligent optimization module, predicts in advance the process difficulties that may arise in the actual production of the design scheme, and generates a process feasibility report for feedback.

[0006] Preferably, the design-process linkage analysis unit establishes a data interaction channel with the cutting path intelligent planning module to transmit the structural features and process feasibility report of the three-dimensional digital model to the cutting path intelligent planning module; Based on the received information, the cutting path intelligent planning module uses an improved simulated annealing algorithm to establish a multi-dimensional comprehensive evaluation model that includes design structure complexity and process implementation difficulty, and searches for the optimal solution through continuous iteration.

[0007] Preferably, the laser cutting process knowledge base built into the process parameter intelligent optimization module contains characteristic parameters of several film material types, thicknesses, and melting points, and the machine learning algorithm adopts a reinforcement learning algorithm with cutting quality score as a reward function; The process parameter intelligent optimization module is also provided with an environmental perception and dynamic compensation submodule, which collects temperature and humidity data in the production environment and real-time characteristic data of the film material in real time; The reinforcement learning algorithm dynamically adjusts the optimization strategy of process parameters such as laser power, cutting speed, pulse frequency, and spot diameter based on the real-time data collected by the environmental perception and dynamic compensation submodule, and establishes a mapping relationship model between environmental parameters and process parameters; In addition, the process parameter intelligent optimization module is also provided with a process-quality correlation analysis unit, which analyzes the correlation between the current process parameters and cutting quality data in real time, combines historical production data, and uses machine learning algorithms to build a process-quality correlation model. When an abnormal quality trend is detected, the process parameters are preventively adjusted in advance.

[0008] Preferably, the process-quality correlation analysis unit also establishes a deep data interaction mechanism with the quality online monitoring and feedback module and the data management and analysis module; The online quality monitoring and feedback module transmits the real-time detected eyelash quality data, including detailed indicators of dimensional accuracy, shape integrity, and edge smoothness, to the process-quality correlation analysis unit in real time; The data management and analysis module integrates and processes the historical production process data, including design parameters, process parameters, equipment operation data, etc., and provides it to the process-quality correlation analysis unit; The process-quality correlation analysis unit constructs a process-quality correlation model based on multi-source data using a deep learning algorithm.

[0009] Preferably, the design input and modeling module is further provided with a complex structure preprocessing unit, which performs hierarchical analysis and feature extraction on the design file for the eyelash design with complex structure including hollowing and multi-layer nesting before performing graphic processing, identifies the structural features and mutual relationships at different levels, and generates a structural feature tree.

[0010] Preferably, the environmental perception and dynamic compensation submodule is also provided with a material batch matching algorithm, which compares the thin film material characteristic data collected in real time with the standard material characteristic data in the process knowledge base. When it is detected that the material batch difference exceeds a preset threshold, the most matching historical process parameters are automatically selected from the knowledge base as the initial optimization starting point, thereby shortening the convergence time of parameter optimization.

[0011] Preferably, the cutting path intelligent planning module also has a cutting path optimization scheme comparison function, which can generate multiple different cutting path optimization schemes, and compare and analyze the indicators of the total cutting path length, film material utilization rate, and expected cutting time of each scheme, and display the comparison results through a visual interface for users to select the optimal scheme.

[0012] Preferably, the data management and analysis module is also provided with a data deep mining and knowledge migration unit, which uses knowledge graph technology to model the data of the entire production process and construct a multi-dimensional knowledge network including design knowledge, process knowledge, and quality knowledge.

[0013] A method for shaping eyelashes based on laser cutting film, comprising the following steps: S1. Design Input and Modeling: Receive user-inputted design parameters for eyelash shape, size, and density, import graphic files, and convert them into 3D digital models. Use built-in graphic processing algorithms and a hierarchical design rule library to verify and correct the model, generate an initial design plan, and predict production process difficulties through design-process linkage analysis. S2. Process parameter optimization: Based on the initial design plan and combined with the laser cutting process knowledge base, a reinforcement learning algorithm is used to optimize the parameters of laser power, cutting speed, pulse frequency, and spot diameter. The environmental perception module collects temperature, humidity, and material property data in real time to dynamically adjust the strategy. The optimized parameters are then evaluated for robustness to ensure anti-interference capabilities. S3. Cutting Path Planning: Based on the 3D digital model and optimized process parameters, an improved simulated annealing algorithm is used to generate the cutting path. The design structure complexity and process implementation difficulty are incorporated into the objective function. Multiple schemes are compared, forming a scheme library for easy access. The path is adjusted in real time in conjunction with the production control module. S4. Production process control: Convert cutting paths and process parameters into control instructions, communicate with laser cutting equipment, receive real-time feedback from equipment operation and adjust instructions, and simultaneously perform production scheduling management to ensure smooth execution of the cutting process; S5. Online quality monitoring: Use image recognition technology to detect the quality indicators of eyelash size accuracy, shape integrity, and edge smoothness after cutting in real time. If any abnormality is found, it will be fed back to the production control module in time, and the detection data will be stored for subsequent analysis.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention uses the adaptive multi-precision Bezier curve fitting technology and deep learning edge refinement algorithm of the design input and modeling module, combined with the reinforcement learning dynamic adjustment mechanism of the process parameter intelligent optimization module, to accurately control the laser cutting parameters, significantly improve the dimensional accuracy and edge smoothness of eyelash cutting, and solve the problems of insufficient precision and hot melt deformation in traditional processes.

[0015] 2. The cutting path intelligent planning module of the present invention incorporates the design structure complexity and process feasibility into the path optimization objective function, and optimizes the cutting sequence through a multi-dimensional comprehensive evaluation model. While improving material utilization, it reduces the problem of low cutting efficiency caused by unreasonable path planning, and further enhances the processing adaptability to complex structures.

[0016] 3. The data management and analysis module of the present invention constructs a full-process knowledge network through knowledge graph technology, realizes the intelligent migration and update of design experience and process strategies, enables the system to predict quality anomalies based on historical data and adjust process parameters in advance, forming a closed-loop management of "design-production-optimization", and fundamentally solves the technical pain points of traditional processes that rely on manual experience and cannot be adaptively optimized. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of the system framework of the present invention. DETAILED DESCRIPTION

[0018] To facilitate those skilled in the art to understand the technical solution of the present invention, the technical solution of the present invention is further described with reference to the accompanying drawings.

[0019] Example 1, as Figure 1 As shown, the present invention provides an eyelash shaping system based on laser cutting film, comprising: The design input and modeling module is used to receive the design parameters of eyelash shape, size, and density input by the user, import the graphic file and convert it into a three-dimensional digital model, and generate an initial design plan after verifying and correcting the model; The process parameter intelligent optimization module is connected to the design input and modeling module. It is used to receive the initial design plan, combine the laser cutting process knowledge base with the film material properties, and use machine learning algorithms to optimize the cutting process parameters such as laser power, cutting speed, pulse frequency, and spot diameter; The cutting path intelligent planning module is connected with the design input and modeling module and the process parameter intelligent optimization module to generate the laser cutting path using the path planning algorithm based on the 3D digital model and optimization parameters; The production process intelligent control module is connected to the cutting path intelligent planning module and the process parameter intelligent optimization module. It is used to convert the cutting path and process parameters into control instructions to communicate with the laser cutting equipment, receive equipment feedback in real time, adjust instructions, and perform production scheduling; The online quality monitoring and feedback module is connected to the production process intelligent control module. It uses image recognition technology to detect the quality of the cut eyelashes in real time. If any problems are found, it will be fed back to the production process intelligent control module and the detection data will be stored; The data management and analysis module is connected to the design input and modeling module, the process parameter intelligent optimization module, the cutting path intelligent planning module, the production process intelligent control module, and the quality online monitoring and feedback module. It is used to store full-process data and generate various reports through big data analysis to provide decision support for production management, process optimization and equipment maintenance.

[0020] In an embodiment of the present invention, the design input and modeling module supports the import of graphic files in several formats and has built-in graphic processing algorithms and model verification functions; The graphics processing algorithm combines adaptive multi-precision Bezier curve fitting technology with a deep learning edge refinement algorithm. For eyelash design areas of varying complexity, the algorithm first identifies complex structural features through a deep learning algorithm, and then automatically adjusts the Bezier curve fitting accuracy parameters based on the feature type. Specifically, for eyelash design areas with complex structures such as hollowing and multi-layer nesting, the fitting accuracy parameters are dynamically adjusted to within the range of 0.01-0.1 pixels. At the same time, the graphics processing algorithm also has a feature point recognition and protection mechanism that can identify key feature points in complex designs and protect them during the fitting process, ensuring that the position accuracy error of feature points does not exceed 0.05 pixels. The adaptive multi-precision Bezier curve fitting technology is implemented by the following formula: ; in, is a point on the curve, are Bernstein basis functions, is the control point, is the curve order; Fitting accuracy parameters According to the design area complexity Dynamic adjustment, ,in, is the basic accuracy parameter, is the attenuation coefficient, Structural feature vector output by deep learning model calculate: ; in, is the eigenvector The weight, is the corresponding weight; The model verification function is based on an extensible design rule library with a hierarchical structure, including a basic design rule layer, a complex structure rule layer, and a customized rule layer. The complex structure rule layer sets special design rules for complex structures such as hollowing and multi-layer nesting. The customized rule layer allows users to add, modify, and delete design rules according to special production needs. The rule library also has conflict detection and automatic repair functions. When a newly added rule conflicts with an existing rule, the cause of the conflict is automatically analyzed and repair suggestions are generated. In addition, the design input and modeling module is also equipped with a design-process linkage analysis unit, which conducts correlation analysis between the constructed three-dimensional digital model and the process knowledge base of the process parameter intelligent optimization module, predicts in advance the process difficulties that may arise in the actual production of the design scheme, and generates a process feasibility report for feedback.

[0021] In an embodiment of the present invention, the design-process linkage analysis unit establishes a data exchange channel with the cutting path intelligent planning module to transmit the structural features and process feasibility report of the three-dimensional digital model to the cutting path intelligent planning module; Based on the received information, the cutting path intelligent planning module uses an improved simulated annealing algorithm. When planning the cutting path, in addition to considering the shortest total cutting path length and the highest film material utilization rate, it also incorporates the process feasibility of the design scheme into the objective function of the path planning. It establishes a multi-dimensional comprehensive evaluation model that includes the complexity of the design structure and the difficulty of process implementation. Through continuous iteration to search for the optimal solution, the comprehensive evaluation model is expressed by the following formula: ; in, is the cutting path plan, is the total path length, is the material utilization rate, Score the design's structural complexity. Score the difficulty of the process. , , , is the corresponding weight and ; Design structure complexity score Based on the design area complexity calculate: ; in, is the number of hollow areas, is the total number of regions, is the balance coefficient; Process implementation difficulty score Calculated by parameters in the process knowledge base: ; in, is the current process parameter, is the ideal process parameter, is the weight coefficient; At the same time, the cutting path intelligent planning module also has the path-model collaborative optimization function. When it is found during the path planning process that the existing model has structural features that are not conducive to cutting, the optimization suggestions are automatically fed back to the design input and modeling module to achieve collaborative optimization of the design model and the cutting path.

[0022] In an embodiment of the present invention, the laser cutting process knowledge base built into the process parameter intelligent optimization module contains characteristic parameters of several film material types, thicknesses, and melting points, and the machine learning algorithm uses a reinforcement learning algorithm with the cutting quality score as the reward function; The process parameter intelligent optimization module is also equipped with an environmental perception and dynamic compensation submodule, which includes temperature and humidity sensors and environmental data acquisition equipment for material property detection devices, and collects temperature and humidity data in the production environment and real-time property data of thin film materials in real time; The reinforcement learning algorithm dynamically adjusts the optimization strategy of process parameters such as laser power, cutting speed, pulse frequency, and spot diameter based on the real-time data collected by the environmental perception and dynamic compensation submodules, and establishes a mapping relationship model between environmental parameters and process parameters; The reinforcement learning algorithm uses the Q-learning framework to update the action-value function using the following formula: ; in, is the state, containing the environmental parameter vector and process parameter vector , , ,in, is the temperature, For humidity, is the material property vector, is the action, i.e. the adjustment amount of process parameters, is the reward function based on the cutting quality score calculate: ; in, and is the balance coefficient; Cut quality score Calculated through quality online monitoring data: ; in, As a quality indicator, and are the minimum and maximum values ​​of the corresponding indicators, is the weight; When the ambient temperature and humidity fluctuate by more than ±5% or the film material properties fluctuate by more than ±3%, the dynamic parameter optimization process is triggered to ensure the stability of cutting quality under different production conditions; At the same time, the process parameter intelligent optimization module is also equipped with a parameter robustness evaluation mechanism. Before outputting the optimized parameters, the robustness evaluation of the parameters is carried out to calculate the impact of parameter fluctuations within a certain range on the cutting quality, ensuring that the optimized parameters have strong anti-interference capabilities. In addition, the process parameter intelligent optimization module is also equipped with a process-quality correlation analysis unit, which analyzes the correlation between the current process parameters and cutting quality data in real time, combines historical production data, and uses machine learning algorithms to build a process-quality correlation model. When an abnormal quality trend is detected, the process parameters are preventively adjusted in advance.

[0023] In an embodiment of the present invention, the process-quality correlation analysis unit also establishes a deep data interaction mechanism with the quality online monitoring and feedback module and the data management and analysis module; The online quality monitoring and feedback module transmits the real-time detected eyelash quality data, including detailed indicators of dimensional accuracy, shape integrity, and edge smoothness, to the process-quality correlation analysis unit; The data management and analysis module integrates and processes historical production process data, including design parameters, process parameters, equipment operation data, etc., and provides it to the process-quality correlation analysis unit; The process-quality correlation analysis unit uses a deep learning algorithm to build a process-quality correlation model based on multi-source data. The model is expressed by the following formula: ;

[0024] in, Score the quality of the predictions, is the process parameter vector, is the environmental parameter vector, is the design parameter vector, and are the neural network weights and biases, is the activation function; The model is trained by minimizing the following loss function: ; in, Rate the actual quality, are model parameters, is the regularization coefficient; When abnormal quality trends are detected, the process parameter adjustment direction is calculated based on the model : ; in, is the learning rate, Score the quality of the target; The online quality monitoring and feedback module transmits the real-time detected eyelash quality data, including detailed indicators such as dimensional accuracy, shape integrity, and edge smoothness, to the process-quality correlation analysis unit in real time; the data management and analysis module integrates and processes historical production process data, including design parameters, process parameters, equipment operation data, etc., and provides it to the process-quality correlation analysis unit; the process-quality correlation analysis unit uses a deep learning algorithm based on multi-source data to continuously optimize and update the process-quality correlation model, thereby improving the model's prediction accuracy for quality anomalies and the targeted adjustment of process parameters; at the same time, when the online quality monitoring and feedback module detects a serious quality problem, the process-quality correlation analysis unit immediately activates the emergency process adjustment plan, and based on the solutions to similar historical quality problems, quickly generates an emergency process parameter adjustment plan, and transmits it to the production process intelligent control module for execution.

[0025] In an embodiment of the present invention, the design input and modeling module is further provided with a complex structure pre-processing unit. This unit performs hierarchical analysis and feature extraction on the design file before graphic processing for eyelash designs with complex structures including hollowing and multi-layer nesting, identifies structural features and their relationships at different levels, and generates a structural feature tree. The graphics processing algorithm adopts different processing strategies for different levels of structures based on the structural feature tree. It uses a high-precision processing mode for key structural layers and a fast processing mode for auxiliary structural layers, thereby improving processing efficiency while ensuring processing accuracy. The hierarchical parsing process is implemented through the following formula: ; in, For design files, For the Layer structure, is the number of layers; Importance score of each layer The calculation is as follows: ; in, For complexity, For structural relevance, is process sensitivity, , , is the weight coefficient; At the same time, the complex structure pre-processing unit also has a design simplification suggestion generation sub-unit. When it detects that the design complexity exceeds the system's preset threshold, it automatically generates design simplification suggestions, including suggestions for adjusting the size of hollow areas and optimizing the number of nested layers, and provides a comparison of the design effects before and after simplification. In addition, the complex structure preprocessing unit is also connected to the data management and analysis module to extract design experience and production data of similar complex structures from the historical design case library to assist users in design decision-making and process planning.

[0026] In an embodiment of the present invention, the environmental perception and dynamic compensation submodule is further provided with a material batch matching algorithm. This algorithm compares the real-time collected thin film material characteristic data with the standard material characteristic data in the process knowledge base. When it is detected that the material batch difference exceeds a preset threshold, the algorithm automatically selects the most matching historical process parameters from the knowledge base as the initial optimization starting point, thereby shortening the convergence time of parameter optimization. The material batch matching algorithm is implemented by the following formula: ; in, is the current material property vector, is the material property set in the process knowledge base, For the A vector of standard material properties; Match score Calculated as: ; in, is the best matching material property vector, is the scaling factor, when When the batch adjustment process is triggered, is the preset threshold; At the same time, the environmental perception and dynamic compensation submodule also has an environmental parameter prediction model. Based on historical environmental data and weather forecast information, it predicts the trend of environmental parameter changes in the future, adjusts the process parameter optimization strategy in advance, and implements proactive parameter optimization. The environmental parameter prediction model uses an LSTM network: ; in, for Step prediction value, is the time window size, is the model weight; In addition, the environmental perception and dynamic compensation submodule is also equipped with a sensor network self-organization and calibration mechanism, which can automatically identify faulty nodes and abnormal data in the sensor network, and ensure the accuracy of environmental data collection through redundant node replacement and data fusion technology; At the same time, the sensor network self-organization and calibration mechanism also has a regular automatic calibration program, which automatically calibrates the sensor according to the standard environmental parameters in the process knowledge base to ensure that the sensor measurement error does not exceed ±1%; In addition, the environmental perception and dynamic compensation submodule is also equipped with an environmental data security transmission protocol, which uses an encryption algorithm to encrypt the collected environmental data to prevent the data from being tampered with or leaked during transmission.

[0027] In an embodiment of the present invention, the cutting path intelligent planning module also has a cutting path optimization scheme comparison function, which can generate multiple different cutting path optimization schemes and compare and analyze the indicators of the total cutting path length, film material utilization rate, and estimated cutting time of each scheme. The comparison results are displayed through a visual interface for the user to select the optimal scheme. The indicator calculation is achieved by the following formula: ; in, is the cutting path plan, The first Points, is the number of path points; The material utilization rate is calculated as: ; in, is the cutting area, is the total area of ​​the film; The estimated cutting time is calculated as: ; in, From point arrive Cutting speed, For the Delay time for the first stop or turn; At the same time, this module supports saving commonly used cutting path optimization solutions to form a cutting path solution library. Users can directly call existing solutions in subsequent production to improve production efficiency. In addition, the cutting path intelligent planning module also establishes a real-time feedback mechanism with the production process intelligent control module. During the cutting process, when the production process intelligent control module detects abnormal equipment operation status or fluctuations in process parameters, it will promptly feedback the information to the cutting path intelligent planning module. The cutting path intelligent planning module dynamically adjusts the unexecuted cutting path based on the feedback information to ensure the smooth progress of the cutting process.

[0028] In an embodiment of the present invention, the data management and analysis module is further provided with a data deep mining and knowledge migration unit, which uses knowledge graph technology to model the data of the entire production process and construct a multi-dimensional knowledge network including design knowledge, process knowledge, and quality knowledge; By analyzing and reasoning about knowledge networks, we can uncover the potential relationships and patterns among data, and form knowledge of reusable design experience, process optimization strategies, and quality control methods. The knowledge graph is represented by entity-relationship-entity triples: ; in, is a collection of entities, is a set of relations; Knowledge reasoning is achieved through the following rules: ; The knowledge transfer process is quantified by the following formula: ; in, For the target task, is the source task, and Characterized by and is the weight, is the similarity function; The data deep mining and knowledge transfer unit also establishes a knowledge sharing interface with the design input and modeling module, the process parameter intelligent optimization module, the cutting path intelligent planning module, etc. When new design tasks or production requirements arise, it automatically searches the knowledge network and pushes relevant knowledge to the corresponding modules to assist the modules in decision-making and optimization. At the same time, the data deep mining and knowledge transfer unit also has the function of knowledge updating and evolution. According to the actual feedback and newly generated data in the production process, it continuously updates and improves the knowledge network to improve the accuracy and practicality of knowledge.

[0029] Example 2: A method for shaping eyelashes based on laser cutting of film, comprising the following steps: S1. Design Input and Modeling: Receive user-inputted design parameters for eyelash shape, size, and density, import graphic files, and convert them into 3D digital models. Use built-in graphic processing algorithms and a hierarchical design rule library to verify and correct the model, generate an initial design plan, and predict production process difficulties through design-process linkage analysis. S2. Process parameter optimization: Based on the initial design plan and combined with the laser cutting process knowledge base, a reinforcement learning algorithm is used to optimize the parameters of laser power, cutting speed, pulse frequency, and spot diameter. The environmental perception module collects temperature, humidity, and material property data in real time to dynamically adjust the strategy. The optimized parameters are then evaluated for robustness to ensure anti-interference capabilities. S3. Cutting Path Planning: Based on the 3D digital model and optimized process parameters, an improved simulated annealing algorithm is used to generate the cutting path. The design structure complexity and process implementation difficulty are incorporated into the objective function. Multiple schemes are compared, forming a scheme library for easy access. The path is adjusted in real time in conjunction with the production control module. S4. Production process control: Convert cutting paths and process parameters into control instructions, communicate with laser cutting equipment, receive real-time feedback from equipment operation and adjust instructions, and simultaneously perform production scheduling management to ensure smooth execution of the cutting process; S5. Online quality monitoring: Use image recognition technology to detect the quality indicators of eyelash size accuracy, shape integrity, and edge smoothness after cutting in real time. If any abnormality is found, it will be fed back to the production control module in time, and the detection data will be stored for subsequent analysis.

[0030] The embodiments disclosed in the present invention are preferred embodiments, but are not limited to them. Ordinary technicians in this field can easily understand the spirit of the present invention based on the above embodiments and make different extensions and changes. As long as they do not deviate from the spirit of the present invention, they are all within the scope of protection of the present invention.

Claims

1. An eyelash shaping system based on laser cutting film, characterized in that, It includes design input and modeling module, process parameter intelligent optimization module, cutting path intelligent planning module, production process intelligent control module, quality online monitoring and feedback module, and data management and analysis module; The process parameter intelligent optimization module is connected to the design input and modeling module, and is used to receive the initial design plan, combine the laser cutting process knowledge base with the film material properties, and use the machine learning algorithm to optimize the cutting process parameters such as laser power, cutting speed, pulse frequency, and spot diameter; The cutting path intelligent planning module is connected to the design input and modeling module and the process parameter intelligent optimization module, and is used to generate the laser cutting path using the path planning algorithm; The production process intelligent control module is connected with the cutting path intelligent planning module and the process parameter intelligent optimization module; The quality online monitoring and feedback module is connected to the production process intelligent control module; The data management and analysis module is connected with the design input and modeling module, the process parameter intelligent optimization module, the cutting path intelligent planning module, the production process intelligent control module, and the quality online monitoring and feedback module.

2. The eyelash shaping system based on laser cutting film according to claim 1, characterized in that: The design input and modeling module has built-in graphics processing algorithms and model verification functions; The graphics processing algorithm combines adaptive multi-precision Bezier curve fitting technology with a deep learning edge refinement algorithm. For eyelash design areas of different complexity, the algorithm first identifies complex structural features through a deep learning algorithm, and then automatically adjusts the Bezier curve fitting accuracy parameters according to the feature type. At the same time, the graphics processing algorithm is also equipped with a feature point recognition and protection mechanism, which can identify key feature points in complex designs and protect them during the fitting process; In addition, the design input and modeling module is also equipped with a design-process linkage analysis unit, which performs correlation analysis between the constructed three-dimensional digital model and the process knowledge base of the process parameter intelligent optimization module, predicts in advance the process difficulties that may arise in the actual production of the design scheme, and generates a process feasibility report for feedback.

3. The eyelash shaping system based on laser cutting film according to claim 2, characterized in that: The design-process linkage analysis unit establishes a data interaction channel with the cutting path intelligent planning module to transmit the structural features and process feasibility report of the three-dimensional digital model to the cutting path intelligent planning module; Based on the received information, the cutting path intelligent planning module uses an improved simulated annealing algorithm to establish a multi-dimensional comprehensive evaluation model that includes design structure complexity and process implementation difficulty, and searches for the optimal solution through continuous iteration.

4. The eyelash shaping system based on laser cutting film according to claim 1, characterized in that: The laser cutting process knowledge base built into the process parameter intelligent optimization module contains characteristic parameters of several film material types, thicknesses, and melting points. The machine learning algorithm adopts a reinforcement learning algorithm with cutting quality scores as a reward function. The process parameter intelligent optimization module is also provided with an environmental perception and dynamic compensation submodule, which collects temperature and humidity data in the production environment and real-time characteristic data of the film material in real time; The reinforcement learning algorithm dynamically adjusts the optimization strategy of process parameters such as laser power, cutting speed, pulse frequency, and spot diameter based on the real-time data collected by the environmental perception and dynamic compensation submodule, and establishes a mapping relationship model between environmental parameters and process parameters; In addition, the process parameter intelligent optimization module is also provided with a process-quality correlation analysis unit, which analyzes the correlation between the current process parameters and cutting quality data in real time, combines historical production data, and uses machine learning algorithms to build a process-quality correlation model. When an abnormal quality trend is detected, the process parameters are preventively adjusted in advance.

5. The eyelash shaping system based on laser cutting film according to claim 4, characterized in that: The process-quality correlation analysis unit also establishes a deep data interaction mechanism with the quality online monitoring and feedback module and the data management and analysis module; The online quality monitoring and feedback module transmits the real-time detected eyelash quality data, including detailed indicators of dimensional accuracy, shape integrity, and edge smoothness, to the process-quality correlation analysis unit in real time; The data management and analysis module integrates and processes the historical production process data, including design parameters, process parameters, equipment operation data, etc., and provides it to the process-quality correlation analysis unit; The process-quality correlation analysis unit constructs a process-quality correlation model based on multi-source data using a deep learning algorithm.

6. The eyelash shaping system based on laser cutting film according to claim 1, characterized in that: The design input and modeling module is also equipped with a complex structure preprocessing unit. This unit performs hierarchical analysis and feature extraction on the design file before graphic processing for eyelash designs with complex structures including hollowing and multi-layer nesting, identifies structural features and their relationships at different levels, and generates a structural feature tree.

7. The eyelash shaping system based on laser cutting film according to claim 4, characterized in that: The environmental perception and dynamic compensation submodule is also equipped with a material batch matching algorithm, which compares the thin film material characteristic data collected in real time with the standard material characteristic data in the process knowledge base. When it is detected that the material batch difference exceeds a preset threshold, the most matching historical process parameters are automatically selected from the knowledge base as the initial optimization starting point, shortening the convergence time of parameter optimization.

8. The eyelash shaping system based on laser cutting film according to claim 1, characterized in that: The cutting path intelligent planning module also has a cutting path optimization scheme comparison function, which can generate multiple different cutting path optimization schemes, and compare and analyze the total cutting path length, film material utilization rate, and estimated cutting time indicators of each scheme, and display the comparison results through a visual interface for users to select the optimal scheme.

9. The eyelash shaping system based on laser cutting film according to claim 1, characterized in that: The data management and analysis module is also equipped with a data deep mining and knowledge migration unit, which uses knowledge graph technology to model the data of the entire production process and build a multi-dimensional knowledge network that includes design knowledge, process knowledge, and quality knowledge.

10. A method for applying the eyelash shaping system based on laser cutting film according to claim 9, characterized in that: The following steps are involved: S1. Design Input and Modeling: Receive user-inputted design parameters for eyelash shape, size, and density, import graphic files, and convert them into 3D digital models. Use built-in graphic processing algorithms and a hierarchical design rule library to verify and correct the model, generate an initial design plan, and predict production process difficulties through design-process linkage analysis. S2. Process parameter optimization: Based on the initial design plan and combined with the laser cutting process knowledge base, a reinforcement learning algorithm is used to optimize the parameters of laser power, cutting speed, pulse frequency, and spot diameter. The environmental perception module collects temperature, humidity, and material property data in real time to dynamically adjust the strategy. The optimized parameters are then evaluated for robustness to ensure anti-interference capabilities. S3. Cutting Path Planning: Based on the 3D digital model and optimized process parameters, an improved simulated annealing algorithm is used to generate the cutting path. The design structure complexity and process implementation difficulty are incorporated into the objective function. Multiple schemes are compared, forming a scheme library for easy access. The path is adjusted in real time in conjunction with the production control module. S4. Production process control: Convert cutting paths and process parameters into control instructions, communicate with laser cutting equipment, receive real-time feedback from equipment operation and adjust instructions, and simultaneously perform production scheduling management to ensure smooth execution of the cutting process; S5. Online quality monitoring: Use image recognition technology to detect the quality indicators of eyelash size accuracy, shape integrity, and edge smoothness after cutting in real time. If any abnormality is found, it will be fed back to the production control module in time, and the detection data will be stored for subsequent analysis.

Images