Intelligent folding and packaging integrated complete equipment for soft cotton towels

Abstract

The invention is suitable for the technical field of intelligent cotton towel production, and provides intelligent cotton towel folding and packaging integrated complete equipment which comprises an intelligent folding assembly additionally arranged on a cotton towel packaging machine. A shaping assembly and a monitoring and positioning system are additionally arranged on the intelligent folding assembly, and the monitoring and positioning system comprises a multiple synchronous detection module and a central processing decision module; the multiple synchronous detection module is in communication connection with the central processing decision module and is used for collecting and transmitting state data of the soft cotton towel in real time; the central processing decision module is in communication connection with the molding assembly and is used for analyzing the data and generating a correction instruction; and the multiple synchronous detection module and the central processing decision module are used for monitoring and feeding back the state of the soft cotton towel in real time, timely finding and accurately positioning a position offset point or a position where a wrinkle is generated, and triggering the shaping assembly to the position offset point or the wrinkle position for correction.

Description

Technical Field

[0001] This invention relates to the field of intelligent production technology for cotton soft towels, and more specifically, to a complete set of equipment for intelligent folding and packaging of cotton soft towels. Background Technology

[0002] Cotton soft wipes, a hygiene product made from pure cotton or non-woven fabric, have seen a continuous increase in market demand in recent years due to their softness, excellent absorbency, and lint-free properties. To meet the requirements of large-scale, high-efficiency production, the post-processing of cotton soft wipes (mainly including folding, cutting, and packaging) has been widely automated. Currently, leading equipment manufacturers in the industry can provide everything from single machines to complete production lines. With the development of industrial automation and intelligent technologies, existing cotton soft wipe folding and packaging equipment has made significant progress in speed and integrated control. In the folding stage, equipment typically uses multi-roller coordination and mechanical claw structures to achieve longitudinal and transverse folding. In the packaging stage, high-speed full servo control has become a hallmark of high-end equipment in the industry. For example, by integrating programmable logic controllers and multiple servo motors and applying EtherCAT bus technology, precise synchronous movement of various execution units (such as pushing, sealing, and bag making) is achieved.

[0003] Despite improvements in speed and automation in existing equipment, a long-standing and unresolved technical challenge remains the real-time sensing and precise control of the cotton tissue material's condition. Cotton tissues are soft and easily affected by external factors such as mechanical vibration, tension fluctuations, or guide roller misalignment during high-speed conveying and feeding into the folding rollers, leading to positional shifts or wrinkles. If these defects cannot be detected accurately and in real-time on the production line, misaligned or wrinkled cotton tissues will directly enter the folding station, affecting not only the neatness of the folds but also causing problems such as material jams, irregular packaging, and even increased scrap rates in subsequent packaging stages.

[0004] While existing technologies include some correction or flattening devices (such as using flattening rollers to increase friction and smooth the cotton towel), they are essentially passive or mechanical correction methods, lacking real-time monitoring and feedback of the material's state. This means that the devices cannot identify and locate problems immediately upon occurrence, thus failing to achieve dynamic and precise compensation and adjustment, resulting in a delayed and uncertain correction effect. For example, in the patent application No. 202422143728.1, "A Folding Roller Structure for a Cotton Towel Folding Machine," the solution is "an arc-shaped ring limits the two sides of the cotton towel, making it less likely for the cotton towel to shift position when fed into the folding roller," without involving any active correction system based on real-time monitoring. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies: when cotton towels are fed into the folding rollers, they are easily affected by external factors such as mechanical vibration, tension fluctuations, or misalignment of the guide rollers, which can lead to positional deviation or wrinkles. If these issues are not detected in time, they can affect the folding effect of the cotton towels and the quality of subsequent packaging. Although existing technologies have devices for correcting cotton towels, they cannot detect and accurately position them in time to precisely solve the problems of positional deviation and cotton towel wrinkles. Therefore, this invention proposes an integrated set of intelligent folding and packaging equipment for cotton towels.

[0006] The specific technical solution is as follows: A complete set of intelligent folding and packaging equipment for cotton towels, including an intelligent folding assembly, which is installed on a cotton towel packaging machine; the intelligent folding assembly is equipped with a shaping assembly and a monitoring and positioning system, the monitoring and positioning system including a multi-synchronous detection module and a central processing and decision module; the multi-synchronous detection module and the central processing and decision module are communicatively connected to collect and transmit the status data of the cotton towel in real time; the central processing and decision module is communicatively connected to the shaping assembly to analyze the data and generate correction instructions; the multi-synchronous detection module and the central processing and decision module are used to monitor and provide feedback on the status of the cotton towel in real time, promptly detect and accurately locate the position deviation point or the position where wrinkles occur, and trigger the shaping assembly to correct at the position deviation point or wrinkle position.

[0007] A further technical solution of the present invention is that the molding assembly includes a molding body, which is mounted on the intelligent folding assembly. The molding body includes a molding cover and two molding adjustment components. The molding cover includes a crossbeam and two storage boxes. The two molding adjustment components correspond one-to-one with the two storage boxes. One end of the molding adjustment component is fitted inside the storage box, and the other end extends on the crossbeam. The two storage boxes are distributed at both ends of the crossbeam and are mounted on the intelligent folding assembly via mounting plates.

[0008] In a further technical solution of the present invention, a positioning cavity is provided on the crossbeam, a guide groove is provided on the crossbeam inside the positioning cavity, and multiple connecting grooves are provided on the guide groove. The multiple connecting grooves are arrayed on the inner wall of the positioning cavity; the connecting grooves connect the guide groove and the positioning cavity; multiple storage grooves are provided on the inner side of the guide groove, and the multiple storage grooves correspond one-to-one with the multiple connecting grooves.

[0009] In a further technical solution of the present invention, the molding adjustment component includes an arc-shaped limiting plate and multiple corrective legs. The corrective legs are assembled on the arc-shaped limiting plate via a connecting plate. The size of the arc-shaped limiting plate matches the guide groove. The arc-shaped limiting plate is movably assembled inside the guide groove and can slide along the guide groove. The multiple corrective legs correspond one-to-one with multiple storage slots, and the corrective legs are movably assembled inside the storage slots. An electric telescopic rod B is assembled on the arc-shaped limiting plate and is installed inside the storage box.

[0010] The corrective support leg includes an electric telescopic rod A and a corrective head. One end of the electric telescopic rod A is mounted on a connecting plate, and the other end is mounted on the corrective head. The corrective head includes an electric telescopic rod C and a corrective wheel. The electric telescopic rod C is mounted on the electric telescopic rod A via an L-shaped positioning plate A. The electric telescopic rod C and the electric telescopic rod A are vertically distributed. The corrective wheel is rotatably mounted on an L-shaped positioning plate B, which is fixed at the end of the electric telescopic rod A. A chain box is mounted on the side of the L-shaped positioning plate B. The chain box contains a drive sprocket and a driven sprocket, which are connected by a chain. The driven sprocket is mounted on the corrective wheel via a rotating shaft, and the drive sprocket is mounted on the output shaft of a motor via a drive shaft. The motor is mounted at the end of the L-shaped positioning plate B.

[0011] In a further technical solution of the present invention, the molding assembly further includes a cleaning cover and an exhaust pipe assembly. The cleaning cover is installed on the crossbeam and connects the storage slot and the connecting slot. The exhaust pipe assembly is installed on the cleaning cover and is externally connected to the vacuum cleaner through its included connecting pipe.

[0012] Compared with the prior art, the present invention has the following advantages: The folding rollers and cotton towel packaging mechanism inside the intelligent folding assembly form an integrated device for folding and packaging cotton towels. During the folding and packaging process, multiple synchronous detection modules collect and transmit the status data of the cotton towels in real time. The central processing and decision module analyzes the data and generates correction instructions. This enables real-time monitoring and feedback of the cotton towel status, timely detection and accurate positioning of position deviation points or wrinkle locations, and triggers the shaping assembly to correct at the position deviation point or wrinkle location. The multi-synchronous detection module and the central processing decision module monitor and provide feedback on the status of the cotton towel in real time. After timely detection and accurate location of position deviation points or wrinkle locations, the shaping assembly is triggered. The shaping assembly activates two shaping adjustment components on the shaping body. The shaping adjustment components activate electric telescopic rod B. Electric telescopic rod B adjusts the position of the arc-shaped limiting plate and multiple straightening support legs along the positioning cavity by telescopic movement, so that the arc-shaped limiting plate can limit cotton towels of different widths. The straightening support legs drive electric telescopic rods A and C to telescopic movement to adjust the straightening wheel, so that the straightening wheel can be accurately positioned at the position deviation point or wrinkle location. Then the motor is activated, and the motor drives the straightening wheel through the chain box to perform accurate point correction. After the vacuum cleaner is started, it will mobilize the air flow inside the cleaning hood, exhaust pipe assembly, storage tank and connection tank to thoroughly clean and recycle the cotton lint and other adhering substances on the surface of the cotton towel, ensuring the hygiene of the cotton towel and improving its quality. By combining a high-speed global camera with a high-precision linear scan camera group in a relay-style working mode, a seamless connection was achieved from rapid anomaly detection to precise measurement and characterization; this ensured timely detection and significantly improved the quality and processing speed of key data through focused scanning, laying a solid foundation for accurate positioning. The self-calibration and compensation submodule actively counteracts inherent interference sources in industrial environments, such as equipment vibration, thermal drift, motion fuzziness, and changes in ambient light. This fundamentally ensures the long-term stability and absolute reliability of the output data from all sensors under any operating conditions, preventing accurate positioning from drifting due to changes in time or environment. Edge computing is used to process massive real-time data streams, enabling highly timely preprocessing and initial judgment to meet timeliness requirements; at the same time, the cloud / server carries complex multimodal fusion models and deep analysis to achieve the ultimate pursuit of accuracy; the two work together to balance real-time response and deep intelligence. The monitoring and positioning system not only analyzes the defects of the current frame, but also combines its historical trajectory to predict trends and filter out interference, making decisions more forward-looking. Combined with the continuous learning of the dynamic knowledge base, the system can continuously accumulate production experience, become more intelligent with use, and eventually evolve towards predictive maintenance and self-optimization of process parameters, surpassing the traditional passive response system. In this invention, the monitoring and positioning system and the molding assembly form a high-speed real-time closed loop. After wrinkles or deviations are detected and accurately located by the multiple synchronous detection modules, the correction wheel of the molding assembly can reach the point within milliseconds to perform physical correction. The massive amount of "defect-correction" success data generated by this closed loop is continuously recorded and analyzed by the dynamic knowledge base. It can not only correct current defects, but also use machine learning to infer the process root cause of defects. For example, it may be found that cotton towels of a certain material are prone to producing a fixed type of wrinkles when passing through a certain pair of folding rollers under a certain humidity environment. Based on this, the monitoring and positioning system can provide early warnings and even automatically fine-tune the speed, temperature or tension parameters of the folding rollers to prevent defects from occurring at the source, realizing a leap from online treatment to pre-emptive prevention. The vacuuming and cleaning system, comprised of the vacuum cleaner, cleaning hood, storage compartment, and connecting compartment, generates crucial synergistic gains with the monitoring and positioning system and molding assembly. The vacuum cleaner promptly removes lint, ensuring product hygiene. For ensuring sensing accuracy, floating lint can adhere to the camera lens or create background interference, severely impacting the reliability of image analysis algorithms. Continuous clean airflow provides a stable and clean observation environment for the vision sensor, ensuring that the self-calibration and compensation submodule can focus on addressing more fundamental issues such as equipment vibration and thermal drift, resulting in an order-of-magnitude improvement in the reliability of "accurate positioning." For ensuring execution effectiveness, if lint adheres to the surface of the cotton towel or the calibrating wheel itself when it flattens or pushes the towel, it may cause secondary contamination or slippage. The cleaning system maintains the cleanliness of the execution interface, ensuring that every physical correction action is direct and effective, improving the correction success rate. The adjustable arc-shaped limiting plate and multi-degree-of-freedom straightening legs of the shaping assembly are inherently flexible. When combined with an intelligent decision-making module that can output precise three-dimensional coordinates, defect types, and sizes, they generate strong adaptability. The system can handle cotton towels of different widths, defect locations, and types. The central processing module can plan non-standard composite correction paths and forces for the straightening wheels based on the microscopic morphology of the defects (such as whether the wrinkles are single-peaked or multi-peaked, and the angle of edge warping). For example, for a slanted wrinkle, the system can instruct two adjacent straightening legs to work together, one responsible for traction and the other for rolling, to complete a complex smoothing action. This gives a hardware system the ability to handle an infinite number of defect morphologies through software definition, greatly expanding the application boundaries of the equipment and its ability to handle complex processes. The architecture, which uses edge computing to process real-time streams and cloud-based deep analysis, has demonstrated value exceeding instantaneous performance in long-term operation, balancing real-time response with the accuracy of deep analysis. The cloud aggregates anonymized defect and process data from different production lines and factories, enabling the dynamic knowledge base to learn across a wide range of dimensions. The resulting optimization models have the ability to generalize across production lines, materials, and environments. When a new cotton towel raw material is introduced to a production line, the system can quickly call up a basic model with similar features from the cloud model library for fine-tuning, significantly shortening the process debugging cycle of the new product. In addition, through the analysis of full lifecycle data, the system can predict the wear trends of key components (such as specific bearings and motors) in advance, achieving true predictive maintenance and avoiding unplanned downtime. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the integrated intelligent folding and packaging equipment for cotton towels of the present invention. Figure 2 for Figure 1 Schematic diagram of the middle molding assembly; Figure 3 for Figure 2 Schematic diagram of the structure of the mid-mounted plastic assembly after it has been flipped over; Figure 4 for Figure 3 Schematic diagram of the structure of the plastic shaping cover; Figure 5 for Figure 4 Cross-sectional view of the central shaping cover; Figure 6 for Figure 3 Schematic diagram of the structure of the plastic molding adjustment component; Figure 7 for Figure 6 A schematic diagram of the structure of the plastic molding adjustment component after it has been flipped over; Figure 8 for Figure 6 Schematic diagram of the structure of the central corrective support leg; Figure 9 for Figure 8 A schematic diagram of the structure of the corrective head.

[0014] In the attached diagram: Intelligent folding assembly 1, shaping assembly 2, cleaning cover 3, exhaust duct assembly 4, shaping body 5; Shaping cover 51, shaping adjustment part 52; 511 crossbeam, 512 storage box, 513 mounting plate, 514 positioning cavity, 515 connecting groove, 516 storage groove, 517 guide groove; Arc-shaped limiting plate 521, connecting plate 522, straightening support leg 523, electric telescopic rod A 524, straightening head 525, electric telescopic rod B 526; L-shaped positioning plate A5251, electric telescopic rod C5252, motor 5253, chain box 5254, L-shaped positioning plate B5255, straightening wheel 5256. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0016] In the embodiments of this invention, please refer to Figure 1 A complete set of intelligent folding and packaging equipment for cotton towels includes an intelligent folding assembly 1, which is installed on a cotton towel packaging machine. Therefore, the folding roller and cotton towel packaging mechanism inside the intelligent folding assembly 1 form an integrated device for folding and packaging cotton towels. Both the folding roller and the cotton towel packaging machine are existing technologies and are not fully illustrated; their detailed structures can be found in existing literature and journals, and they can also be purchased directly from the market. Those skilled in the art can choose to purchase them according to their needs. These are not the subjects of this invention and will not be described in detail here. The intelligent folding assembly 1 is equipped with a shaping assembly 2 and a monitoring and positioning system. The monitoring and positioning system includes a multi-synchronous detection module and a central processing and decision-making module. The multi-synchronous detection module and the central processing and decision-making module are connected to each other to collect and transmit the status data of the cotton towel in real time. The central processing and decision-making module is connected to the shaping assembly 2 to analyze the data and generate correction instructions. The multi-synchronous detection module and the central processing decision module are used to monitor and provide feedback on the status of the cotton towel in real time, promptly detect and accurately locate the position offset point or the location of the wrinkle, and trigger the shaping assembly 2 to correct at the position offset point or the wrinkle location.

[0017] Therefore, while existing technologies have devices for correcting cotton towels, they cannot promptly detect and accurately locate the problem to precisely resolve positional deviations and cotton towel wrinkles. This application achieves the following: the folding roller and cotton towel packaging mechanism inside the intelligent folding assembly 1 form an integrated device for folding and packaging cotton towels. During the folding and packaging process, multiple synchronous detection modules collect and transmit the cotton towel's status data in real time, and the central processing and decision-making module analyzes the data and generates correction instructions. This enables real-time monitoring and feedback of the cotton towel's status, timely detection and accurate location of positional deviations or wrinkle locations, triggering the shaping assembly 2 to correct the positional deviations or wrinkle locations.

[0018] In the embodiments of this invention, please refer to Figures 1-4 The molding assembly 2 includes a molding body 5, which is mounted on the intelligent folding assembly 1. The molding body 5 includes a molding cover 51 and two molding adjustment components 52. The molding cover 51 includes a crossbeam 511 and two storage boxes 512. The two molding adjustment components 52 correspond one-to-one with the two storage boxes 512. One end of the molding adjustment component 52 is fitted inside the storage box 512, and the other end extends on the crossbeam 511. The two storage boxes 512 are distributed at both ends of the crossbeam 511 and are mounted on the intelligent folding assembly 1 through a mounting plate 513.

[0019] For further details, please refer to Figure 4 and Figure 5 The crossbeam 511 has a positioning cavity 514 (the specific positioning cavity 514 is arc-shaped and matches the shape of the folding roller). Inside the positioning cavity 514, the crossbeam 511 has a guide groove 517. The guide groove 517 has multiple connecting grooves 515, which are arrayed on the inner wall of the positioning cavity 514. The connecting grooves 515 connect the guide groove 517 and the positioning cavity 514. The guide groove 517 has multiple storage grooves 516 on its inner side, which correspond one-to-one with the multiple connecting grooves 515.

[0020] Please see Figures 4-7 The shaping adjustment component 52 includes an arc-shaped limiting plate 521 and multiple corrective legs 523. The corrective legs 523 are assembled on the arc-shaped limiting plate 521 through a connecting plate 522. The size of the arc-shaped limiting plate 521 matches the guide groove 517. The arc-shaped limiting plate 521 is movably assembled inside the guide groove 517 and can slide along the guide groove 517. Multiple straightening legs 523 correspond one-to-one with multiple storage slots 516, and the straightening legs 523 are movably assembled inside the storage slots 516; an electric telescopic rod B526 is assembled on the arc-shaped limiting plate 521, and the electric telescopic rod B526 is installed inside the storage box 512.

[0021] Please see Figures 7-9 The corrective support leg 523 includes an electric telescopic rod A524 and a corrective head 525. One end of the electric telescopic rod A524 is mounted on the connecting plate 522, and the other end is assembled on the corrective head 525. The corrective head 525 includes an electric telescopic rod C5252 and a corrective wheel 5256. The electric telescopic rod C5252 is assembled on the electric telescopic rod A524 through an L-shaped positioning plate A5251. The electric telescopic rod C5252 and the electric telescopic rod A524 are vertically distributed. The straightening wheel 5256 is rotatably mounted on the L-shaped positioning plate B5255, which is fixed to the end of the electric telescopic rod A524. A chain box 5254 is mounted on the side of the L-shaped positioning plate B5255. The chain box 5254 contains a drive sprocket and a driven sprocket, which are connected by a chain. The driven sprocket is mounted on the straightening wheel 5256 via a rotating shaft, and the drive sprocket is mounted on the output shaft of the motor 5253 via a drive shaft. The motor 5253 is mounted at the end of the L-shaped positioning plate B5255.

[0022] Motor 5253, electric telescopic pole A524, electric telescopic pole B526 and electric telescopic pole C5252 are all existing technologies, which can be purchased directly on the market and can be used directly; they can also be purchased arbitrarily by those skilled in the art according to their own needs; they are not what this invention is meant to protect, and will not be described in detail here. Therefore, after the multi-synchronous detection module and the central processing decision module monitor and provide feedback on the status of the cotton towel in real time, and promptly detect and accurately locate the positional deviation point or the position where wrinkles occur, the shaping assembly 2 is triggered. The shaping assembly 2 activates the two shaping adjustment components 52 on the shaping body 5. The shaping adjustment component 52 activates the electric telescopic rod B526. The electric telescopic rod B526 adjusts the position of the arc-shaped limiting plate 521 and multiple corrective legs 523 along the positioning cavity 514 by telescopic movement, so that the arc-shaped limiting plate 521 can limit the cotton towels of different widths. The corrective legs 523 drive the electric telescopic rods A524 and C5252 to telescopically adjust the corrective wheel 5256, so that the corrective wheel 5256 can be accurately positioned at the positional deviation point or the position of the wrinkle. Then the motor 5253 is activated. The motor 5253 drives the corrective wheel 5256 to perform accurate point correction through the chain box 5254.

[0023] In the embodiments of this invention, please refer to Figure 1 and Figure 2The molding assembly 2 also includes a cleaning cover 3 and an exhaust pipe assembly 4. The cleaning cover 3 is installed on the crossbeam 511 and connects the storage groove 516 and the connecting groove 515. The exhaust pipe assembly 4 is installed on the cleaning cover 3 and is externally connected to the vacuum cleaner through the connecting pipe it contains.

[0024] Vacuum cleaners are existing technology, readily available on the market and directly applicable; they can be freely selected by those skilled in the art according to their needs; they are not the subject of this invention and will not be described in detail here. Therefore, after the vacuum cleaner is started, it will mobilize the air flow inside the cleaning cover 3, the exhaust pipe group 4, the storage tank 516 and the connecting tank 515 to thoroughly clean and recycle the cotton lint and other adhering substances on the surface of the cotton towel, ensuring the hygiene of the cotton towel and improving its quality.

[0025] In this embodiment of the invention, the multiple synchronization detection module includes: The first vision sensing unit is a high-speed global area array camera located upstream of the production line. It is used to capture the overall image of the cotton towel in the preparatory area before it enters the folding station, and to perform preliminary defect area screening and early warning. The second visual sensing unit is a high-resolution line scan camera group located downstream of the preparatory area and adjacent to the entrance of the smart folding assembly 1. It is used to perform millisecond-level continuous scanning of suspected defect areas or key edges after preliminary screening to obtain sub-pixel-level precision contour and surface texture detail data. The dynamic ranging sensing unit, which is at least one laser triangulation ranging sensor or structured light projector, is set at the same station as the line scan camera group to synchronously acquire information on the micro-depth change of the surface of the cotton towel in the normal direction, so as to quantify the fold height and edge warping. The first visual sensing unit, the second visual sensing unit, and the dynamic ranging sensing unit achieve strict synchronization of data acquisition through hardware trigger signals, and send the data to the central processing and decision-making module after unifying the timestamp. The self-calibration and compensation submodule, which is integrated into each sensing unit or used as an independent unit, is used to periodically perform reference pattern calibration, automatically correct sensor internal parameter errors caused by equipment vibration or thermal drift, receive speed signals from the production line encoder in real time, and perform motion blur compensation on the acquired images, and dynamically adjust the brightness of the auxiliary lighting unit based on feedback from the ambient light intensity sensor to ensure stable image quality.

[0026] In this embodiment of the invention, the central processing and decision-making module adopts an architecture of edge computing and cloud collaboration, specifically including: Edge computing nodes, deployed on the production line, integrate a streaming data processing engine to perform real-time preprocessing, feature extraction, and preliminary fusion of raw high-dimensional sensor data uploaded by multiple synchronous detection modules, generating lightweight intermediate results. A cloud-based or local server analysis unit, connected to the edge computing node network, is used to receive the intermediate results and run an advanced defect identification and decision-making model. The advanced defect identification and decision-making model includes: A multimodal feature fusion network is used to deeply fuse image features, contour features, and depth features from different sensors at the feature layer to construct a unified feature descriptor for defects. The spatiotemporal context analyzer is used to combine current defect features with their historical trajectories in consecutive frames to predict the evolution trend of defects and distinguish between persistent defects and transient disturbances. The precise localization engine, based on fused features and contextual information, outputs the defect type, precise 3D coordinates, geometric dimensions, severity level, and confidence level in a 3D spatial coordinate system.

[0027] Furthermore, the central processing and decision-making module also includes a dynamic knowledge base and optimization loops, which are used for: It continuously stores all defect cases generated during the production process, the corresponding raw sensor data, decision instructions, and the final correction results feedback; By using offline or online learning algorithms, the advanced defect identification and decision-making model is incrementally trained and its parameters are optimized periodically, enabling the system to have production line adaptation and defect prediction capabilities.

[0028] The initial fusion performed by the edge computing node specifically includes: Rapid region of interest localization is performed using data from the first visual sensing unit; The second visual sensing unit and the dynamic ranging sensing unit are driven to perform focused scanning of the ROI region, thereby achieving on-demand allocation of computing resources and maximizing processing efficiency.

[0029] Therefore, by combining a high-speed global camera with a high-precision linear array camera group in a relay-style working mode, a seamless connection was achieved from rapid anomaly detection to precise measurement and characterization; this ensured timely detection and significantly improved the quality and processing speed of key data through focused scanning, laying a solid foundation for accurate positioning. The self-calibration and compensation submodule actively counteracts inherent interference sources in industrial environments, such as equipment vibration, thermal drift, motion fuzziness, and changes in ambient light. This fundamentally ensures the long-term stability and absolute reliability of the output data from all sensors under any operating conditions, preventing accurate positioning from drifting due to changes in time or environment. Edge computing is used to process massive real-time data streams, enabling highly timely preprocessing and initial judgment to meet timeliness requirements; at the same time, the cloud / server carries complex multimodal fusion models and deep analysis to achieve the ultimate pursuit of accuracy; the two work together to balance real-time response and deep intelligence. The monitoring and positioning system not only analyzes the defects of the current frame, but also combines its historical trajectory to predict trends and filter out interference, making decisions more forward-looking. Combined with the continuous learning of the dynamic knowledge base, the system can continuously accumulate production experience, becoming more intelligent with use, and eventually evolving towards predictive maintenance and self-optimization of process parameters, surpassing the traditional passive response system.

[0030] In summary: In this invention, the monitoring and positioning system and the molding assembly 2 form a high-speed real-time closed loop. After wrinkles or deviations are detected and accurately located by the multiple synchronous detection modules, the correction wheel 5256 of the molding assembly 2 can reach the point within milliseconds to perform physical correction. The massive amount of "defect-correction" success data generated by this closed loop is continuously recorded and analyzed by the dynamic knowledge base. It can not only correct current defects, but also use machine learning to infer the process root cause of defects. For example, it may be found that cotton towels of a certain material are prone to producing fixed types of wrinkles when passing through a pair of folding rollers under a certain humidity environment. Based on this, the monitoring and positioning system can provide early warnings and even automatically fine-tune the speed, temperature or tension parameters of the folding rollers to prevent defects from occurring at the source, realizing a leap from online treatment to pre-emptive prevention.

[0031] The vacuum cleaning system, comprised of the vacuum cleaner, cleaning hood 3, storage slot 516, and connecting slot 515, generates crucial synergistic gains with the monitoring and positioning system and molding assembly 2. The vacuum cleaner promptly removes lint, ensuring product hygiene. For ensuring sensing accuracy, floating lint can adhere to the camera lens or create background interference, severely impacting the reliability of image analysis algorithms. Continuous clean airflow provides a stable and clean observation environment for the visual sensor, ensuring that the self-calibration and compensation submodule can focus on addressing more fundamental issues such as equipment vibration and thermal drift, resulting in an order-of-magnitude improvement in the reliability of "accurate positioning." For ensuring execution effectiveness, if lint adheres to the surface of the cotton towel or the correction wheel itself when it flattens or pushes it, it may cause secondary contamination or slippage. The cleaning system maintains the cleanliness of the execution interface, ensuring that every physical correction action is direct and effective, improving the correction success rate.

[0032] The adjustable arc-shaped limiting plate 521 and the multi-degree-of-freedom straightening support leg 523 of the molding assembly 2 are inherently flexible. When combined with an intelligent decision-making module that can output precise three-dimensional coordinates, defect types, and sizes, they generate strong adaptability. The system can handle cotton towels of different widths, defect locations, and types. The central processing module can plan non-standard composite correction paths and forces for the straightening wheel based on the microscopic morphology of the defect (such as whether the wrinkle is single-peaked or multi-peaked, and the angle of edge warping). For example, for a slanted wrinkle, the system can instruct two adjacent straightening support legs to work together, one responsible for traction and the other for rolling, to complete a complex smoothing action. This gives a hardware system the ability to handle an infinite number of defect morphologies through software definition, greatly expanding the application boundaries of the equipment and its ability to handle complex processes.

[0033] The architecture, which uses edge computing to process real-time streams and cloud-based deep analysis, has demonstrated value exceeding instantaneous performance in long-term operation, balancing real-time response with the accuracy of deep analysis. The cloud aggregates anonymized defect and process data from different production lines and factories, enabling the dynamic knowledge base to learn across a wide range of dimensions. The resulting optimization models have the ability to generalize across production lines, materials, and environments. When a new cotton towel raw material is introduced to a production line, the system can quickly call up a basic model with similar features from the cloud model library for fine-tuning, significantly shortening the process debugging cycle of the new product. In addition, through the analysis of full lifecycle data, the system can predict the wear trends of key components (such as specific bearings and motors) in advance, achieving true predictive maintenance and avoiding unplanned downtime.

[0034] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is 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 complete set of equipment for intelligent folding and packaging of cotton towels, comprising an intelligent folding assembly (1), wherein the intelligent folding assembly (1) is mounted on a cotton towel packaging machine; characterized in that, The intelligent folding assembly (1) is equipped with a molding assembly (2) and a monitoring and positioning system. The monitoring and positioning system includes a multi-synchronous detection module and a central processing and decision module. The multi-synchronous detection module and the central processing decision module are connected in communication to collect and transmit the status data of the cotton towel in real time; the central processing decision module is connected in communication with the shaping assembly (2) to analyze the data and generate correction instructions; The multi-synchronous detection module and the central processing decision module are used to monitor and provide feedback on the status of the cotton towel in real time, promptly detect and accurately locate the position offset point or the position where wrinkles occur, and trigger the shaping assembly (2) to correct at the position offset point or the position of the wrinkles.

2. The integrated intelligent folding and packaging equipment for cotton towels according to claim 1, characterized in that, The molding assembly (2) includes a molding body (5), which is mounted on the intelligent folding assembly (1). The molding body (5) includes a molding cover (51) and two molding adjustment parts (52). The molding cover (51) includes a crossbeam (511) and two storage boxes (512). The two molding adjustment parts (52) correspond one-to-one with the two storage boxes (512). One end of the molding adjustment part (52) is fitted inside the storage box (512), and the other end extends on the crossbeam (511). The two storage boxes (512) are distributed at both ends of the crossbeam (511), and the storage boxes (512) are mounted on the intelligent folding assembly (1) through a mounting plate (513).

3. The integrated intelligent folding and packaging equipment for cotton towels according to claim 2, characterized in that, The crossbeam (511) has a positioning cavity (514), and the crossbeam (511) inside the positioning cavity (514) has a guide groove (517). The guide groove (517) has multiple connecting grooves (515), and the multiple connecting grooves (515) are arrayed on the inner wall of the positioning cavity (514). The connecting grooves (515) connect the guide groove (517) and the positioning cavity (514). The guide groove (517) has multiple storage grooves (516) on its inner side, and the multiple storage grooves (516) correspond one-to-one with the multiple connecting grooves (515).

4. The integrated intelligent folding and packaging equipment for cotton towels according to claim 3, characterized in that, The molding adjustment component (52) includes an arc-shaped limiting plate (521) and multiple straightening legs (523). The straightening legs (523) are assembled on the arc-shaped limiting plate (521) through a connecting plate (522). The size of the arc-shaped limiting plate (521) matches the guide groove (517). The arc-shaped limiting plate (521) is movably assembled inside the guide groove (517) and can slide along the guide groove (517). Multiple corrective legs (523) correspond one-to-one with multiple storage slots (516), and the corrective legs (523) are movably assembled inside the storage slots (516); an electric telescopic rod B (526) is assembled on the arc-shaped limiting plate (521), and the electric telescopic rod B (526) is installed inside the storage box (512).

5. The integrated intelligent folding and packaging equipment for cotton towels according to claim 4, characterized in that, The corrective support leg (523) includes an electric telescopic rod A (524) and a corrective head (525). One end of the electric telescopic rod A (524) is mounted on the connecting plate (522), and the other end is mounted on the corrective head (525). The corrective head (525) includes an electric telescopic rod C (5252) and a corrective wheel (5256). The electric telescopic rod C (5252) is mounted on the electric telescopic rod A (524) through an L-shaped positioning plate A (5251). The electric telescopic rod C (5252) and the electric telescopic rod A (524) are vertically distributed. The straightening wheel (5256) is rotatably mounted on the L-shaped positioning plate B (5255), which is fixed at the end of the electric telescopic rod A (524). A chain box (5254) is mounted on the side of the L-shaped positioning plate B (5255). The chain box (5254) is equipped with a drive sprocket and a driven sprocket. The drive sprocket and the driven sprocket are connected by a chain. The driven sprocket is mounted on the straightening wheel (5256) through a rotating shaft. The drive sprocket is mounted on the output shaft of the motor (5253) through a drive shaft. The motor (5253) is installed at the end of the L-shaped positioning plate B (5255).

6. The integrated intelligent folding and packaging equipment for cotton towels according to claim 1, characterized in that, The molding assembly (2) also includes a cleaning cover (3) and an exhaust pipe assembly (4). The cleaning cover (3) is mounted on the crossbeam (511) and connects the storage slot (516) and the connecting slot (515). The exhaust pipe assembly (4) is mounted on the cleaning cover (3) and is connected to the vacuum cleaner through the connecting pipe it contains.

7. The integrated intelligent folding and packaging equipment for cotton towels according to claim 1, characterized in that, The multiple synchronization detection module includes: The first vision sensing unit is a high-speed global area array camera located upstream of the production line. It is used to capture the overall image of the cotton towel in the preparatory area before it enters the folding station, and to perform preliminary defect area screening and early warning. The second visual sensing unit is a high-resolution line array camera group located downstream of the preparatory area and adjacent to the entrance of the smart folding assembly (1). It is used to perform millisecond-level continuous scanning of suspected defect areas or key edges that have been preliminarily screened, and to obtain sub-pixel-level precision contour and surface texture detail data. The dynamic ranging sensing unit, which is at least one laser triangulation ranging sensor or structured light projector, is set at the same station as the line scan camera group to synchronously acquire information on the micro-depth change of the surface of the cotton towel in the normal direction, so as to quantify the fold height and edge warping. The first visual sensing unit, the second visual sensing unit, and the dynamic ranging sensing unit achieve strict synchronization of data acquisition through hardware trigger signals, and send the data to the central processing and decision-making module after unifying the timestamp. The self-calibration and compensation submodule, which is integrated into each sensing unit or used as an independent unit, is used to periodically perform reference pattern calibration, automatically correct sensor internal parameter errors caused by equipment vibration or thermal drift, receive speed signals from the production line encoder in real time, and perform motion blur compensation on the acquired images, and dynamically adjust the brightness of the auxiliary lighting unit based on feedback from the ambient light intensity sensor to ensure stable image quality.

8. The integrated intelligent folding and packaging equipment for cotton towels according to claim 1, characterized in that, The central processing and decision-making module adopts an architecture that combines edge computing and cloud collaboration, specifically including: Edge computing nodes, deployed on the production line, integrate a streaming data processing engine to perform real-time preprocessing, feature extraction, and preliminary fusion of raw high-dimensional sensor data uploaded by multiple synchronous detection modules, generating lightweight intermediate results. A cloud-based or local server analysis unit, connected to the edge computing node network, is used to receive the intermediate results and run an advanced defect identification and decision-making model. The advanced defect identification and decision-making model includes: A multimodal feature fusion network is used to deeply fuse image features, contour features, and depth features from different sensors at the feature layer to construct a unified feature descriptor for defects. The spatiotemporal context analyzer is used to combine current defect features with their historical trajectories in consecutive frames to predict the evolution trend of defects and distinguish between persistent defects and transient disturbances. The precise localization engine, based on fused features and contextual information, outputs the defect type, precise 3D coordinates, geometric dimensions, severity level, and confidence level in a 3D spatial coordinate system.

9. The integrated intelligent folding and packaging equipment for cotton towels according to claim 8, characterized in that, The central processing and decision-making module also includes a dynamic knowledge base and optimization loops, which are used for: It continuously stores all defect cases generated during the production process, the corresponding raw sensor data, decision instructions, and the final correction results feedback; By using offline or online learning algorithms, the advanced defect identification and decision-making model is incrementally trained and its parameters are optimized periodically, enabling the system to have production line adaptation and defect prediction capabilities.

10. The integrated intelligent folding and packaging equipment for cotton towels according to claim 8, characterized in that, The initial fusion performed by the edge computing node specifically includes: Rapid region of interest localization is performed using data from the first visual sensing unit; The second visual sensing unit and the dynamic ranging sensing unit are driven to perform focused scanning of the ROI region, thereby achieving on-demand allocation of computing resources and maximizing processing efficiency.

Images