A method for AI optimization of foam-shaped die-cutting path
By optimizing the die-cutting path of irregular foam shapes using AI algorithms, combined with image processing and sensor feedback, the problems of low efficiency and poor quality in traditional die-cutting technology are solved, achieving efficient and stable foam cutting production.
Patent Information
- Application Number
- CN202610476262.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-04-10
- Publication Date
- 2026-07-10
AI Technical Summary
Existing foam die-cutting technology relies on manual experience or fixed CNC programs, resulting in frequent tool turning, excessive idle strokes, poor material adaptability, lack of real-time adjustment capabilities, high verification costs, and difficulty in achieving high-quality mass production.
AI algorithms combined with image processing and sensor feedback are used to optimize the die-cutting path of irregular foam shapes. This includes image acquisition and preprocessing, feature recognition, path planning, dynamic parameter adjustment and closed-loop control, to achieve globally optimal path planning and real-time compensation.
It significantly improves production efficiency, increases material utilization by 18%, reduces cutting time by 22%, ensures consistent cutting quality, reduces scrap rate, extends equipment life and improves process stability.
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent manufacturing technology, and more specifically, to an AI optimization method for irregular die-cutting paths of foam. Background Technology
[0002] Foam materials are widely used in cushioning, sealing, and packaging applications in industries such as electronics, automotive, and medical. Cutting irregularly shaped foam requires high precision and efficiency. Traditional die-cutting processes rely on manual experience or fixed CNC programs, which have the following limitations: Path planning relies on experience: Cutting paths planned by manual programming or simple algorithms often result in frequent tool turns, many idle strokes, low efficiency, and do not take into account the impact of material physical properties (such as hardness and springback) on cutting quality. Poor adaptability: When faced with differences in texture, uneven thickness, or tendency to deform in foam, fixed parameters are prone to cutting deviations, burrs, or material waste, making it difficult to achieve high-quality mass production. Lack of real-time adjustment capability: Dynamic factors such as tool wear and material deformation during the cutting process cannot be compensated for in real time, requiring machine stoppage and manual intervention, which affects production continuity; High verification costs: The feasibility of the path depends on physical trial cutting, which increases material and time costs; While some existing technologies employ genetic algorithms for path optimization, these methods are largely limited to theoretical simulations and lack deep integration with material property databases, real-time sensor feedback, and closed-loop control, thus limiting their practical application effectiveness. Therefore, an AI-based path optimization method for irregularly shaped foam die-cutting is urgently needed to address these issues. Summary of the Invention
[0003] The purpose of this invention is to provide an AI optimization method for irregular die-cutting paths of foam to solve the problems mentioned in the background art. To achieve the above objectives, the present invention provides the following technical solution: an AI optimization method for irregular die-cutting paths of foam, comprising the following steps: S1. First, image acquisition and preprocessing are performed. High-precision industrial cameras are used to acquire high-definition images of foam raw materials. Image processing technology is used to extract irregular contour features and remove noise interference to generate a clear digital contour model. S2. Next, feature recognition and data annotation are performed on the digital contour model generated in step S1. By marking the physical properties of material hardness, thickness, and deformation tendency, a database linking material properties and cutting parameters is established. S3. Intelligent path planning: Global path planning is performed by improving AI algorithms and combining them with genetic algorithms to reduce the number of tool turns and generate the globally optimal planned route. S4. Dynamically and adaptively adjust the tool parameters, automatically adjust the cutting speed, pressure, and tool compensation parameters to achieve closed-loop control of "cutting-feedback-adjustment"; S5. Perform simulation verification and iterative optimization of the die-cutting path. Conduct 3D cutting simulation in a virtual environment to evaluate the efficiency and accuracy of the path. Use the generated data to feed back to the AI model for iterative training and continuously optimize the path algorithm until it meets the threshold required for production. S6. The final optimized toolpath is converted into CNC code and synchronized to the die-cutting equipment control system through the IoT interface to achieve automated execution. It is also equipped with a monitoring module to record production data in real time and support long-term performance analysis and model updates. As a preferred technical solution of the present invention, the image processing technology used in step S1 includes one or more of grayscale conversion, binarization and edge detection. This technology can achieve the effect of automatically identifying foam edges, textures and defect areas. As a preferred technical solution of the present invention, the feature recognition in step S2 is based on the automatic identification of irregular contour features by convolutional neural network, and provides a structural basis for subsequent path planning. As a preferred technical solution of the present invention, the global path planning in step S3 needs to consider tool wear, material springback and thermal deformation factors, and also needs to calculate the optimal cutting sequence and infeed angle in real time to ensure path feasibility. As a preferred technical solution of the present invention, the adaptive adjustment in step S4 includes real-time analysis of tool wear and material deformation during the cutting process, and dynamic adjustment of cutting speed, pressure and feed rate to compensate for errors and maintain the consistency of cutting quality. As a preferred technical solution of the present invention, the "cutting-feedback-adjustment" closed-loop control in step S4 is specifically as follows: by integrating force sensors and vision sensors on the tool or material support platform, cutting resistance and contour deviation are monitored in real time, and the AI model dynamically fine-tunes the cutting parameters of the next segment based on the feedback. As a preferred technical solution of the present invention, the threshold for meeting production requirements in step S5 is set as a material utilization rate increase of ≥18% and a cutting time reduction of ≥22%. As a preferred technical solution of the present invention, in step S6, the data recorded in real time by the monitoring module includes: the deviation between the actual cutting path and the planned path, the pass rate of each batch of products, the tool life decay curve, and the equipment energy consumption data. These data are aggregated into the cloud analysis platform to trigger the periodic automatic update warning of the AI model. As a preferred technical solution of the present invention, the path efficiency and accuracy evaluated in the simulation verification are quantified as a key component of the fitness function in the hybrid genetic algorithm, and are used to directly guide the evolution direction of the path in iterative optimization. Compared with the prior art, the present invention has the following beneficial effects: This invention is an AI optimization method for irregular die-cutting paths of foam. By optimizing the path through AI algorithms, this invention reduces tool turning and idle stroke. Combined with real-time parameter adjustment, it can shorten the cutting time by ≥22%, increase material utilization by ≥18%, and significantly improve production efficiency. Based on closed-loop control of material property database and sensor feedback, it dynamically compensates for tool wear and material deformation, ensuring the consistency of cutting quality of different batches of foam and reducing scrap rate. This invention is an AI optimization method for die-cutting irregular foam shapes. The invention uses IoT monitoring and cloud data analysis to support tool life prediction, energy consumption optimization, and autonomous AI model updates, thereby extending equipment life and improving long-term process stability. It integrates image processing, CNN feature recognition, hybrid genetic algorithms, and closed-loop control to form an intelligent solution covering the entire chain of "perception-planning-execution-optimization", which is suitable for efficient die-cutting production of complex irregular foam shapes. Detailed Implementation Example An example includes the following steps: S1. First, image acquisition and preprocessing are performed. High-precision industrial cameras are used to acquire high-definition images of foam raw materials. Image processing technology is used to extract irregular contour features and remove noise interference to generate a clear digital contour model. In this embodiment, the image processing techniques used in step S1 include one or more of grayscale conversion, binarization, and edge detection. This technique can achieve the effect of automatically identifying foam edges, textures, and defect areas. S2. Next, feature recognition and data annotation are performed on the digital contour model generated in step S1. By marking the physical properties of material hardness, thickness, and deformation tendency, a database linking material properties and cutting parameters is established. In this embodiment, the feature recognition in step S2 is based on the automatic identification of irregular contour features by convolutional neural network, and provides a structural basis for subsequent path planning. S3. Intelligent path planning: Global path planning is performed by improving AI algorithms and combining them with genetic algorithms to reduce the number of tool turns and generate the globally optimal planned route. In this embodiment, the global path planning in step S3 needs to consider factors such as tool wear, material springback and thermal deformation, and also needs to calculate the optimal cutting sequence and feed angle in real time to ensure path feasibility. S4. Dynamically and adaptively adjust the tool parameters, automatically adjust the cutting speed, pressure, and tool compensation parameters to achieve closed-loop control of "cutting-feedback-adjustment"; In this embodiment, the adaptive adjustment in step S4 includes real-time analysis of tool wear and material deformation during the cutting process, and dynamic adjustment of cutting speed, pressure and feed rate to compensate for errors and maintain consistent cutting quality. In this embodiment, the "cutting-feedback-adjustment" closed-loop control in step S4 is specifically as follows: by integrating force sensors and vision sensors on the tool or material support platform, cutting resistance and contour deviation are monitored in real time, and the AI model dynamically fine-tunes the cutting parameters for the next segment based on the feedback. S5. Perform simulation verification and iterative optimization of the die-cutting path. Conduct 3D cutting simulation in a virtual environment to evaluate the efficiency and accuracy of the path. Use the generated data to feed back to the AI model for iterative training and continuously optimize the path algorithm until it meets the threshold required for production. In this embodiment, the threshold for meeting production requirements in step S5 is set as a material utilization rate increase of ≥18% and a cutting time reduction of ≥22%. S6. The final optimized toolpath is converted into CNC code and synchronized to the die-cutting equipment control system through the Internet of Things interface to achieve automated execution. It is also equipped with a monitoring module to record production data in real time and support long-term performance analysis and model updates. In this embodiment, in step S6, the data recorded in real time by the monitoring module includes: the deviation between the actual cutting path and the planned path, the pass rate of each batch of products, the tool life decay curve, and equipment energy consumption data. This data is aggregated into the cloud analysis platform to trigger the periodic automatic update warning of the AI model. In this embodiment, the path efficiency and accuracy evaluated in the simulation verification are quantified as key components of the fitness function in the hybrid genetic algorithm, and are used to directly guide the evolution direction of the path in iterative optimization. Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for optimizing the die-cutting path of irregularly shaped foam using AI, characterized in that, Includes the following steps: S1. First, image acquisition and preprocessing are performed. High-precision industrial cameras are used to acquire high-definition images of foam raw materials. Image processing technology is used to extract irregular contour features and remove noise interference to generate a clear digital contour model. S2. Next, feature recognition and data annotation are performed on the digital contour model generated in step S1. By marking the physical properties of material hardness, thickness, and deformation tendency, a database linking material properties and cutting parameters is established. S3. Intelligent path planning: Global path planning is performed by improving AI algorithms and combining them with genetic algorithms to reduce the number of tool turns and generate the globally optimal planned route. S4. Dynamically and adaptively adjust the tool parameters, automatically adjust the cutting speed, pressure, and tool compensation parameters to achieve "cutting-feedback-adjustment" closed-loop control; S5. Perform simulation verification and iterative optimization of the die-cutting path. Conduct 3D cutting simulation in a virtual environment to evaluate the efficiency and accuracy of the path. Use the generated data to feed back to the AI model for iterative training and continuously optimize the path algorithm until it meets the threshold required for production. S6. The final optimized toolpath is converted into CNC code and synchronized to the die-cutting equipment control system through the IoT interface to achieve automated execution. It is also equipped with a monitoring module to record production data in real time and support long-term performance analysis and model updates.
2. The AI optimization method for irregular die-cutting path of foam according to claim 1, characterized in that, The image processing techniques used in step S1 include one or more of grayscale conversion, binarization, and edge detection. This technique can automatically identify foam edges, textures, and defect areas.
3. The AI optimization method for irregular die-cutting path of foam according to claim 1, characterized in that, In step S2, feature recognition is based on the automatic identification of irregular contour features using a convolutional neural network, providing a structural basis for subsequent path planning.
4. The AI optimization method for irregular die-cutting path of foam according to claim 1, characterized in that, In step S3, the global path planning needs to consider factors such as tool wear, material springback, and thermal deformation. It also needs to calculate the optimal cutting sequence and feed angle in real time to ensure path feasibility.
5. The AI optimization method for irregular die-cutting path of foam according to claim 1, characterized in that, The adaptive adjustment in step S4 includes real-time analysis of tool wear and material deformation during the cutting process, and dynamic adjustment of cutting speed, pressure and feed rate to compensate for errors and maintain consistent cutting quality.
6. The AI optimization method for irregular die-cutting path of foam according to claim 1, characterized in that, The "cutting-feedback-adjustment" closed-loop control in step S4 specifically involves integrating force sensors and vision sensors on the tool or material support platform to monitor cutting resistance and contour deviation in real time. The AI model then dynamically fine-tunes the cutting parameters for the next segment based on this feedback.
7. The AI optimization method for irregular die-cutting path of foam according to claim 1, characterized in that, In step S5, the threshold for meeting production requirements is set as follows: material utilization rate increased by ≥18% and cutting time shortened by ≥22%.
8. The AI optimization method for irregular die-cutting path of foam according to claim 1, characterized in that, In step S6, the data recorded in real time by the monitoring module includes: the deviation between the actual cutting path and the planned path, the pass rate of each batch of products, the tool life decay curve, and equipment energy consumption data. This data is aggregated into the cloud analysis platform to trigger the periodic automatic update and early warning of the AI model.
9. The AI optimization method for irregular die-cutting path of foam according to claim 1, characterized in that, The path efficiency and accuracy evaluated in the simulation verification are quantified as a key component of the fitness function in the hybrid genetic algorithm, and are used to directly guide the evolutionary direction of the path in iterative optimization.