Automatic sewing and cutting management system for underwear processing

Abstract

The invention relates to the technical field of intelligent manufacturing and management of textiles and garments, and discloses an automatic sewing and cutting management system for underwear processing. The system comprises a central control module, a material conveying and positioning sub-module, a machine vision acquisition sub-module, a multi-axis linkage sewing mechanical arm sub-module and a cutting execution sub-module. According to the system, image data of raw materials and semi-finished products are collected through a machine vision collection sub-module before sewing and after sewing respectively, the initial deformation quantity and the secondary deformation quantity are calculated through a central control module, then a dynamic sewing path and an actual cutting path are generated, and a mechanical arm is controlled to execute self-adaptive sewing and accurate cutting; a sensing-decision-execution closed-loop intelligent control strategy is adopted, the problem of deformation accumulation of the high-elasticity fabric in the processing process is effectively solved, the precision and efficiency of underwear production and the material utilization rate are further improved, and small-batch multi-variety flexible automatic production is achieved.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing and management technology in textiles and apparel, and more specifically, to an automated sewing and cutting management system for underwear processing. Background Technology

[0002] Traditional underwear manufacturing relies heavily on manual operation, resulting in low production efficiency, poor consistency, and a high dependence on skilled workers. Existing automated equipment mostly only completes single processes, such as simple cutting or simple sewing, requiring manual intervention between processes. This makes it unable to handle the deformation of highly elastic fabrics during processing, leading to low precision in the finished product. In particular, the stretching and twisting deformations generated during sewing directly affect the accuracy of subsequent cutting processes, causing dimensional deviations and material waste. For example, a machine vision-based garment cutting system and method (patent publication number CN119932889A) suffers from this problem. To address the problems in existing technologies, an automated sewing and cutting management system for underwear processing is provided, enabling flexible and automated production of high-quality underwear. Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the purpose of this invention is to provide an automated sewing and cutting management system for underwear processing, so as to solve the problems existing in the background technology.

[0004] The above-mentioned technical objective of the present invention is achieved through the following technical solution: an automated sewing and cutting management system for underwear processing, comprising: The central control module has a built-in digital twin model, which is used to store several standard pattern data for producing underwear, and to receive, process and control the coordinated execution of each sub-module. The material conveying and positioning submodule is used to fix and convey raw underwear materials and semi-finished underwear products; The machine vision acquisition submodule is used to acquire and transmit image data of the raw underwear material before sewing and the semi-finished underwear product after sewing. The multi-axis linkage sewing robot arm submodule is used to perform sewing operations on the underwear raw materials on the material conveying and positioning submodule; The cutting execution submodule is used to perform cutting operations on the semi-finished underwear products on the material conveying and positioning submodule. The central control module is configured as follows: Based on the image data of the underwear raw material before sewing collected by the machine vision acquisition submodule, the initial deformation of the sheet material is calculated, and then a dynamic sewing path adapted to the initial deformation is generated, and a first control command is generated for the sewing robotic arm module to execute. Based on the image data of the sewn underwear semi-finished product acquired by the machine vision acquisition submodule, the secondary deformation caused by the sewing operation is calculated, and then an actual cutting path to compensate for the secondary deformation is generated, and a second control command is generated for the cutting execution module to execute.

[0005] Optionally, the central control module includes: The main control and scheduling unit is responsible for coordinating the timing control and task scheduling of the control system workflow. According to the preset process formula, it triggers the various actions of visual scanning, robotic arm movement, sewing, and cutting in an orderly manner to ensure that all sub-modules work together in the correct order. Digital twin units are used to store standardized model data of underwear styles to be produced, which can be retrieved during production; The adaptive path planning unit, based on a built-in integrated algorithm, is used to analyze the image data acquired by the machine vision acquisition submodule, and then generate the stitching path or cutting path for the corresponding process.

[0006] Optionally, the machine vision acquisition submodule includes: The industrial camera imaging unit includes several industrial cameras that can switch between different wavelength light sources to enhance the ability to identify feature points of different colors, textures and transparent underwear materials. They are set up at the sewing station and the cutting station respectively to collect image data of underwear materials before sewing and image data of underwear semi-finished products after sewing in real time. The trigger shooting unit includes several pressure sensors, which are respectively set on the material conveying and positioning submodule to identify the entry of underwear raw materials or underwear semi-finished products, so as to trigger the real-time data acquisition of the industrial camera shooting unit. The imaging lighting unit uses a ring-shaped LED shadowless light source and is respectively set at the sewing station and the cutting station to enhance the imaging effect of the industrial camera imaging unit; The image preprocessing and output unit is used to preprocess the image data acquired by the industrial camera capturing unit using an image preprocessing method, and then use feature extraction and recognition algorithms to convert the preprocessed image data into engineering data and output it to the central control module.

[0007] Optionally, the material conveying and positioning submodule includes: Material placement platform, used for placing lingerie raw materials and semi-finished lingerie products; The flexible coarse positioning unit, composed of electromagnetic matrix pins, is set on the material placement platform and is used to dynamically form a first positioning area for coarse positioning based on the shape and size of the underwear raw materials and underwear semi-finished products. The multi-zone vacuum adsorption precision positioning unit consists of several vacuum chamber partitions for vacuum adsorption of underwear raw materials and underwear semi-finished products, which are respectively set on the material placement platform. Each vacuum chamber partition is independently controlled to dynamically travel to the second positioning zone for precision positioning, so as to adapt to underwear raw materials and underwear semi-finished products of different sizes and shapes. The servo drive and transmission unit is responsible for moving the material placement platform so that the underwear raw materials and underwear semi-finished products can pass through the sewing and cutting processes in sequence.

[0008] Optionally, the multi-axis linkage suture robotic arm submodule includes: The robotic arm sewing unit uses a six-axis articulated robotic arm, which is equipped with an integrated sewing machine head for sewing. The task parsing and control unit is used to receive and parse the first control command, decompose it into movement commands for each joint of the robotic arm suturing unit, so as to realize the suturing process; The safety protection unit, consisting of several torque sensors, is installed on the robotic arm sewing unit. When a foreign object or human body is detected, it promptly performs an emergency stop.

[0009] Optionally, the trimming execution submodule includes: The robotic arm cutting unit uses a six-axis articulated robotic arm, and an end effector for cutting is installed on the robotic arm; The task parsing and control unit is used to receive and parse the second control command, and decompose it into movement commands for each joint of the robotic arm cutting unit to realize the cutting process.

[0010] Optionally, the wrist of the robotic arm suturing unit is equipped with a six-dimensional force sensor for real-time monitoring of the needle and thread tension during the suturing process; The central control module is further configured to receive feedback data from the six-dimensional force sensor and then dynamically adjust the sewing speed, stitch length, and presser foot pressure of the integrated sewing machine head to form a closed-loop control for the sewing tension, thereby actively suppressing material deformation caused by the sewing process itself.

[0011] Optionally, the wrist of the robotic arm cutting unit is equipped with a three-dimensional laser profilometer, which is used to perform a final scan of the surface of the semi-finished underwear before the cutting end actuator is activated, and transmits the scan results to the central control module. The central control module compares the scan results with the expected path at the micrometer level, and then performs final path calibration before cutting.

[0012] An underwear processing method based on the above-mentioned automated sewing and cutting management system for underwear processing includes: Step S1, Underwear raw material feeding: The underwear raw material is placed on the material conveying and positioning sub-module by manual or machine personnel. The underwear raw material is fixed and conveyed by the flexible coarse positioning unit and the multi-zone vacuum adsorption fine positioning unit on the material conveying and positioning sub-module. Step S2, planning the sewing path of underwear raw materials: The machine vision acquisition submodule takes pictures of the underwear raw materials laid out before sewing and transmits the pre-processed image data to the central control module. The adaptive path planning unit in the central control module calculates the initial deformation based on the built-in integrated algorithm. Based on the initial deformation, the pre-stored standard sewing path is corrected to generate a dynamic sewing path, and then the first control command of the sewing process is generated and transmitted. Step S3: Underwear raw materials are sewn into semi-finished underwear: The multi-axis linkage sewing robot sub-module parses and executes the first control command transmitted by the central control module, thereby sewing the underwear raw materials into semi-finished underwear. Step S4: Underwear semi-finished product cutting path planning: The machine vision acquisition submodule captures images of the sewn underwear semi-finished product and transmits the pre-processed image data to the central control module. The adaptive path planning unit in the central control module calculates the quadratic deformation variable generated by the sewing process based on the built-in integrated algorithm. Based on the quadratic deformation variable, it performs elastic mapping and rearrangement of the pre-stored standard cutting path to generate the actual cutting path, and then generates and transmits the second control command for the cutting process. Step S5: Cutting the semi-finished underwear into finished underwear: The cutting execution submodule parses and executes the second control instruction transmitted by the central control module, thereby cutting the semi-finished underwear into finished underwear. Step S6, Finished Underwear Conveying: The material conveying and positioning submodule conveys the processed underwear to the subsequent processes.

[0013] In summary, the present invention has the following beneficial effects: 1. During the lingerie processing, two visual inspections before and after sewing, along with real-time path compensation, further eliminate deformation errors caused by material elasticity and processing stress, improving the lingerie product size qualification rate. At the same time, it achieves seamless integration and automation of sewing and cutting processes, reducing manual intervention and semi-finished product handling time, and further improving production efficiency.

[0014] 2. The central control module is equipped with a digital twin module, which can store underwear product data according to actual applications, enabling seamless switching of product models without changing physical molds. It can achieve rapid production changeover for small batches and multiple varieties, greatly enhancing the company's ability to respond to market changes. It digitizes and automates complex sewing and cutting processes, reduces the skill requirements for operators, and solves the problem of insufficient intelligence and automation in the garment manufacturing industry. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the system flow logic of the present invention; Figure 2 This is a schematic diagram of the method steps of the present invention; Figure 3 This is a schematic diagram of the workflow of the central control module of the present invention; Figure 4 This is a schematic diagram of the workflow of the machine vision acquisition submodule of the present invention; Figure 5 This is a schematic diagram of the workflow of the material conveying and positioning submodule of the present invention; Figure 6 This is a schematic diagram of the fine-tuning steps of the multi-axis linkage stitching robot arm sub-module of the present invention; Figure 7 This is a schematic diagram of the trimming path adjustment process of the trimming execution submodule of the present invention. Detailed Implementation

[0016] To make the objectives, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Several embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein.

[0017] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances. The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature.

[0018] In this invention, unless otherwise expressly specified and limited, "above" or "below" a second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of a second feature includes the first feature being directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" of a second feature includes the first feature being directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature. The terms "vertical," "horizontal," "left," "right," "above," "below," and similar expressions are for illustrative purposes only and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed or operated in a specific orientation, and therefore should not be construed as limiting the invention.

[0019] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] This invention provides an automated sewing and cutting management system for underwear processing, such as... Figure 1 As shown, it includes: The central control module has a built-in digital twin model, which is used to store several standard pattern data for producing underwear, and to receive, process and control the coordinated execution of each sub-module. The material conveying and positioning submodule is used to fix and convey raw underwear materials and semi-finished underwear products; The machine vision acquisition submodule is used to acquire and transmit image data of the raw underwear material before sewing and the semi-finished underwear product after sewing. The multi-axis linkage sewing robot arm submodule is used to perform sewing operations on the underwear raw materials on the material conveying and positioning submodule; The cutting execution submodule is used to perform cutting operations on the semi-finished underwear products on the material conveying and positioning submodule. The central control module is configured as follows: Based on the image data of the underwear raw material before sewing collected by the machine vision acquisition submodule, the initial deformation of the sheet material is calculated, and then a dynamic sewing path adapted to the initial deformation is generated, and a first control command is generated for the sewing robotic arm module to execute. Based on the image data of the sewn underwear semi-finished product acquired by the machine vision acquisition submodule, the secondary deformation caused by the sewing operation is calculated, and then an actual cutting path to compensate for the secondary deformation is generated, and a second control command is generated for the cutting execution module to execute.

[0021] Furthermore, the central control module includes: The main control and scheduling unit is responsible for coordinating the timing control and task scheduling of the control system workflow. According to the preset process formula, it triggers the various actions of visual scanning, robotic arm movement, sewing, and cutting in an orderly manner to ensure that all sub-modules work together in the correct order. Digital twin units are used to store standardized model data of underwear styles to be produced, which can be retrieved during production; The adaptive path planning unit, based on a built-in integrated algorithm, is used to analyze the image data acquired by the machine vision acquisition submodule, and then generate the stitching path or cutting path for the corresponding process.

[0022] In the specific implementation process, such as Figure 3 As shown, the main control and scheduling unit in the central control module is responsible for coordinating the operation of all sub-modules, performing data processing, decision-making, and issuing instructions; The digital twin unit, as the core database and virtual mapping unit of the system, not only stores static CAD data, but also constructs dynamic virtual models that can interact with physical entities in real time. These models include, but are not limited to: a product model library containing standard model data for all underwear styles to be produced, including 2D contours, 3D surfaces, stitch lines, feature point coordinates, and material properties, including the anisotropic elastic modulus of the underwear style; and a process parameter library containing optimal sewing parameters for different fabrics and styles, including but not limited to speed, stitch length, and tension, as well as cutting parameters for underwear styles, including but not limited to power and speed. In addition, it also includes a historical data lake, which records all sensor data, deformations, adjustment parameters and final quality results in each production process for big data analysis and model optimization. The adaptive path planning unit, as the core unit of this system, is used to convert visual data into actual path planning, and then generate corresponding control commands. During the generation of the first control command, namely the planning of the sewing process and path: the machine vision acquisition submodule takes pictures of the underwear material at the pre-sewing station and transmits them to the adaptive path planning unit after preprocessing. The adaptive path planning unit accurately extracts the pixel coordinates of the actual feature points on the underwear material. The actual feature points can be pre-punched positioning holes, printed marking points, or contour inflection points of the underwear material itself. After extracting the feature points, the standard coordinates of the corresponding theoretical feature points of the underwear material in the digital twin unit are called. Then, based on the iterative nearest point IPC point set registration algorithm, the optimal spatial transformation parameters required to match the actual feature point set to the theoretical feature point set are calculated. The transformation is a transformation matrix containing translation Tx,Ty, rotation R and scaling Sx,Sy parameters. The scaling factors Sx,Sy are the quantized values ​​T1 of the initial deformation of the underwear material. Then, using the inverse of the calculated transformation matrix T1, a reverse transformation is performed on each point on the pre-stored standard stitching path in the digital twin, as shown below:

[0023] This step then generates This is represented as a dynamic stitching path, which geometrically compensates for the initial deformation of the material, so that the stitching trajectory of the robotic arm can perfectly fit the actual position of the current piece of material. Finally, it is encapsulated into the first control command and sent to the multi-axis linkage stitching robotic arm sub-module. During the generation of the second control instruction, namely the planning of the cutting process and path: the machine vision acquisition submodule takes pictures of the semi-finished underwear at the sewing station and transmits them to the adaptive path planning unit after preprocessing. The adaptive path planning unit accurately extracts feature points, which are either the sewing stitches themselves or new contour lines. After extracting the feature points, the nearest point IPC point set registration algorithm is used to match them with the theoretical outline of the final product in the digital twin, and a new transformation matrix T2 is calculated. The transformation matrix T2 quantifies the quadratic deformation caused by the stitching operation. In the cutting path planning, the processing methods of T2 and T1 are distinguished. The transformation matrix T2 is regarded as a deformation field. A nonlinear mapping algorithm is adopted. Based on the distribution of the quadratic deformation, point-by-point, non-uniform dynamic compensation is performed on each point on the standard cutting path. This generates a new actual cutting path that can accurately offset all deformation errors accumulated from the fabric laying to the sewing process, ensuring that the final cut contour is completely consistent with the design size. The optimized actual cutting path is encapsulated as a second control command and sent to the controller of the cutting execution submodule to drive it to complete the precise cutting. The adaptive path planning unit achieves closed-loop control from perceiving reality to predictive decision-making and then to precise execution through two visual feedbacks, two point set registration calculations, and two path compensation generations, transforming abstract image pixel information into concrete, executable mechanical motion commands.

[0024] Furthermore, the machine vision acquisition submodule includes: The industrial camera imaging unit includes several industrial cameras that can switch between different wavelength light sources to enhance the ability to identify feature points of different colors, textures and transparent underwear materials. They are set up at the sewing station and the cutting station respectively to collect image data of underwear materials before sewing and image data of underwear semi-finished products after sewing in real time. The trigger shooting unit includes several pressure sensors, which are respectively set on the material conveying and positioning submodule to identify the entry of underwear raw materials or underwear semi-finished products, so as to trigger the real-time data acquisition of the industrial camera shooting unit. The imaging lighting unit uses a ring-shaped LED shadowless light source and is respectively set at the sewing station and the cutting station to enhance the imaging effect of the industrial camera imaging unit; The image preprocessing and output unit is used to preprocess the image data acquired by the industrial camera capturing unit using an image preprocessing method, and then use feature extraction and recognition algorithms to convert the preprocessed image data into engineering data and output it to the central control module.

[0025] In the specific implementation process, such as Figure 4 As shown, the machine vision acquisition submodule is an automated module that handles the process from physical signal acquisition to intelligent data output. During the triggering and image acquisition process, when the material conveying platform precisely delivers the underwear raw materials or semi-finished products to the sewing or cutting station, the pressure sensor installed on the platform positioning pin or edge detects a stable pressure signal, indicating that the material is in place; the sensor of the triggering unit sends a digital signal to the main control scheduling unit of the central control module, and the main control scheduling unit then sends a hard trigger signal to the industrial camera at the corresponding station. The industrial camera exposes the image the moment it receives the trigger signal, capturing a high-quality image; at the same time, the ring-shaped LED shadowless light source of the shooting lighting unit is synchronously controlled by the main control scheduling unit to provide optimal lighting conditions, ensuring that the image is uniform, shadowless, and has high contrast; During image preprocessing, the acquired raw image data is transmitted to the image preprocessing and output unit, where a preset preprocessing algorithm is called to process the raw image. Specific steps include: flat field correction, used to eliminate lens vignetting and uneven lighting; filtering and denoising, using median filtering or Gaussian filtering to suppress random noise from the fabric texture itself or the environment, while preserving key edges; image enhancement, using contrast stretching or histogram equalization to further enhance target features, such as the distinction between edges, markers, and background; and binarized edge detection, used to convert the image into a black and white binary image and use operators such as Canny to extract clear contour lines. In the process of feature extraction and recognition, image information is converted into quantified engineering data. Based on a preset feature model, the preset feature model adopts a convolutional neural network model trained by deep learning. This model is trained using massive underwear material image data and can robustly identify stitching, alignment marks and contour features under various interference conditions. For underwear raw materials before sewing, the main identification lies in artificial marking features, such as positioning holes and printed "+" marks, and natural features, such as the outline corner points of the underwear raw materials, and then calculating the sub-pixel precision coordinates of these feature points. For the sewn-up semi-finished underwear, the main identification process features are the direction of the stitches and the overall outline. Then, in the subsequent path planning, the deformation is quantified by comparing the sewn stitches with the theoretical stitches in the digital twin. Finally, the extracted feature information, such as the list of point coordinates and the mathematical equation of the contour, is packaged into a standardized data format and sent to the adaptive path planning unit of the central control module. The path planning unit uses this data to calculate the deformation variables and generate a new execution path.

[0026] In the first example of the shooting, for example, in the production process of a certain underwear, a dark-colored fabric with complex textures was used. The complex black pattern had extremely low contrast with the background under ordinary visible light, and the traditional combination of camera and flash could not reliably identify the alignment mark point. During the actual shooting process, the main control scheduling unit learned from the digital twin unit that the current lingerie material was black lace, and automatically called the visual acquisition formula optimized for this type of fabric. When the lingerie raw materials arrive at the corresponding workstation, the triggering unit activates the multispectral industrial camera. Instead of capturing ordinary RGB images, the camera switches to the near-infrared band, such as an 850nm light source, for imaging. Different materials have different absorption and reflection characteristics for different wavelengths of light. The polyester or nylon components in lace fabric have significantly different reflectivity in the near-infrared band compared to the background, thus making the pattern clearly visible. The image preprocessing unit utilizes this characteristic to easily extract the nodes of the lace pattern as feature points, accurately calculate their coordinates, and output the data to the central control module. This solves the problem that traditional vision systems cannot handle complex fabrics. In Example 2 of the photography, when processing transparent silicone shoulder straps or glossy satin materials, the material surface produces strong reflections, forming highlights, which makes it impossible to effectively identify the true feature information. During the actual shooting process, the main control scheduling unit learns from the digital twin unit that the current underwear material is a glossy satin material and automatically calls up the visual acquisition formula optimized for this type of fabric. The shooting lighting unit activates a special lighting mode, using a low-angle ring-shaped shadowless light source to supplement the lighting of the underwear raw materials and semi-finished underwear. The low-angle lighting makes the light almost parallel to the surface of the material, which can highlight the small contour undulations, such as edges and markings, as bright features, while flat surfaces are displayed as dark fields, effectively suppressing mirror reflections. After the camera acquires high-quality images, the preprocessing algorithm in the integrated algorithm can be used to further remove residual reflection interference through highlight detection and repair algorithms, and finally extract clear and accurate feature data.

[0027] Furthermore, the material conveying and positioning submodule includes: Material placement platform, used for placing lingerie raw materials and semi-finished lingerie products; The flexible coarse positioning unit, composed of electromagnetic matrix pins, is set on the material placement platform and is used to dynamically form a first positioning area for coarse positioning based on the shape and size of the underwear raw materials and underwear semi-finished products. The multi-zone vacuum adsorption precision positioning unit consists of several vacuum chamber partitions for vacuum adsorption of underwear raw materials and underwear semi-finished products, which are respectively set on the material placement platform. Each vacuum chamber partition is independently controlled to dynamically travel to the second positioning zone for precision positioning, so as to adapt to underwear raw materials and underwear semi-finished products of different sizes and shapes. The servo drive and transmission unit is responsible for moving the material placement platform so that the underwear raw materials and underwear semi-finished products can pass through the sewing and cutting processes in sequence.

[0028] Specifically, such as Figure 5 As shown, the material placement platform is made of a highly flat plate, such as anodized aluminum alloy or stainless steel, with a dense array of tiny adsorption pores on the surface. Below the plate, it is divided into multiple independent vacuum chamber zones. Each vacuum chamber zone is equipped with a vacuum generator controlled by a solenoid valve, thereby achieving vacuum adsorption of underwear raw materials and semi-finished underwear products. The independent control of multiple zones allows the system to activate the suction force of only the zone carrying the sheet material, while the zone without material is turned off. This ensures effective fixation and saves energy. Furthermore, it can adapt to sheets of different sizes and shapes, achieving on-demand adsorption and thus enabling precise positioning of underwear raw materials and semi-finished underwear products. The flexible coarse positioning unit consists of electromagnetic matrix pins. A matrix of holes is embedded inside or at the edge of the material placement platform. Each hole contains a metal pin that can be raised and lowered by an electromagnet, thus forming an electromagnetic matrix pin matrix. The central control module controls the pins at specific coordinate positions to rise according to the pattern being produced, forming a physical boundary that matches the outline of the underwear raw material. This allows for initial positioning and limiting of the underwear raw material, thereby achieving coarse positioning of the underwear raw material and the semi-finished underwear. The servo drive and transmission unit is responsible for driving the entire platform to move between workstations. It uses a servo motor with a high-precision linear module, preferably a ball screw or synchronous belt, to ensure that the positional accuracy of the platform's movement and stopping is kept within a good error range. In the first embodiment of material positioning transportation, the operator inputs the production order into the central control module. The central control module retrieves the pattern data of the corresponding style and generates positioning instructions and adsorption zone instructions accordingly. Specific pins in the electromagnetic matrix on the material placement platform rise according to the program, forming the outline of a virtual mold on the material placement platform. At the same time, the solenoid valve corresponding to the adsorption zone enters the ready-to-trigger state. The operator quickly and accurately places the underwear raw material into the roughly correct position by using the outline of the virtual mold, and the tooling eliminates the main positional and angular errors. The material placement platform is also equipped with a detection grating. When the operator presses the confirmation button or the system senses the removal of the hand through the grating, only the zone solenoid valves of the area covered by the sheet are activated. The powerful vacuum suction pulls the fabric evenly and flatly onto the platform surface through the micropores. This process can effectively smooth out the internal stress wrinkles of the flat fabric caused by weaving and cutting, so that it is in a flat and uniform tension state, which provides more convenience for subsequent high-precision visual inspection and processing. The central control module controls the servo drive module to precisely move the entire platform, along with the firmly fixed sheet material, to the front of the sewing station, which is directly below the first vision station. After the platform stops, the vision system performs scanning and calculations. After completing the visual scanning and sewing operations, the central control module controls the platform to move and transport the semi-finished underwear to the sewing station, which is the second visual station. Throughout the transportation process, the vacuum adsorption remains on to ensure that the semi-finished product does not slip or loosen during the movement. It is then scanned again at the second visual station before proceeding with the subsequent cutting process. After all processing steps are completed, the central control module closes the vacuum adsorption solenoid valve, releasing the vacuum. The finished underwear is no longer fixed, and the operator or a simple robotic arm can remove the finished product. The positioning pins are lowered, the platform moves back to the loading station, and the system is ready for the next cycle.

[0029] Furthermore, the multi-axis linkage suture robotic arm submodule includes: The robotic arm sewing unit uses a six-axis articulated robotic arm, which is equipped with an integrated sewing machine head for sewing. The task parsing and control unit is used to receive and parse the first control command, decompose it into movement commands for each joint of the robotic arm suturing unit, so as to realize the suturing process; The safety protection unit, consisting of several torque sensors, is installed on the robotic arm sewing unit. When a foreign object or human body is detected, it promptly performs an emergency stop.

[0030] Furthermore, the wrist of the robotic arm suturing unit is equipped with a six-dimensional force sensor for real-time monitoring of the needle and thread tension during the suturing process; The central control module is further configured to receive feedback data from the six-dimensional force sensor and then dynamically adjust the sewing speed, stitch length, and presser foot pressure of the integrated sewing machine head to form a closed-loop control for the sewing tension, thereby actively suppressing material deformation caused by the sewing process itself.

[0031] Specifically, the core of this module is to convert the suture path instructions issued by the central control module into the actual movement of the robotic arm joints, and in the process, to achieve precise needle-thread synchronization and force-controlled suture. The suture path instruction issued by the central control module, namely the first control instruction, is a sequence of discrete suture path points, each path point... A six-dimensional pose vector represents the target pose of the sewing machine head end effector in the base coordinate system, denoted as:

[0032] in, , , These are represented as coordinate positions, , , This is expressed as an attitude parameter, i.e., the orientation of the nose relative to the base coordinate system; The task parsing and control unit will analyze each pose. By using inverse kinematics, the problem is transformed into the angles of the six joints of the robotic arm. In this process, let the forward motion function of the robotic arm be... , joint angle Mapped to the end effector pose, i.e., the pose of the integrated sewing machine head. , represented as:

[0033] The inverse kinematics solution is the inverse problem of solving the above equations, expressed as:

[0034] For a six-axis articulated arm, this equation typically has multiple solutions. The controller selects the optimal solution based on the shortest path principle, allowing each joint to move from its current position... Move to The rotation amount is minimized, and thus a path composed of a sequence of joint angles is obtained analytically; Then, trajectory planning is performed on the joint angle paths formed by the joint angle sequence. Fifth-order spline interpolation in joint space is used to generate smooth joint motion trajectories, and their first derivative (angular velocity) and second derivative (angular acceleration) are calculated. This ensures the smoothness and stability of the motion process while satisfying the constraints on joint velocity and acceleration, expressed as:

[0035] Simultaneously, needle-thread synchronization modeling is performed to further ensure that the sewing process meets actual production requirements. It is a typical hybrid system, including continuous robotic arm movement and discrete needle insertion, and is described using a finite state machine model: State 1 is insertion, where the needle tip pierces the fabric. At this time, the robotic arm's speed in three-dimensional space is forcibly constrained to near zero to ensure stable needle penetration without pulling on the fabric, represented as: ; State two is feeding, where the needle tip of the sewing machine head lifts up and leaves the fabric. At this time, the robotic arm quickly moves from the current point to the next point. This movement process is feeding. The switching between the two states is triggered by the angle signal of the main shaft encoder on the integrated sewing machine head, and its phase is synchronized with the movement trajectory of the robotic arm. In another embodiment, such as Figure 6 As shown, a six-dimensional force sensor is also installed on the wrist of the robotic arm to measure the force on the end of the sewing machine head, and its output is a six-dimensional force vector. , represented as:

[0036] Then, by coordinate transformation, the equivalent force at the needle tip can be obtained. ; Based on the equivalent force at the needle tip, a PID-based force closed-loop controller is used, with the suture speed adjustment as the main adjustment variable. Its control law is expressed as:

[0037] in, This represents the stitching speed command output at time t; This represents the set initial suturing speed; This represents the stitching force error. This represents the suture tension measured in real time at time t. It is the desired ideal suture tension; These are represented by the proportional, integral, and derivative gains of the PID controller, respectively; and the stitching tension is then dynamically adjusted based on the force measured in real time. The safety protection unit operates as an independent parallel process. The controller continuously reads torque sensors to monitor the load on each joint. Under normal circumstances, the load remains within a preset safety threshold, requiring no intervention. In abnormal situations, such as an abnormal collision, when the robotic arm collides with a person or foreign object, the joint load instantaneously exceeds the safety threshold. The controller triggers an emergency stop within milliseconds, immediately cutting off power and braking all servo motors to maximize the safety of personnel and equipment.

[0038] In Example 1 of the sewing process, for example, adaptive quilting of a three-dimensional bra cup, the bra cup is a complex three-dimensional curved surface. Traditional automated equipment has difficulty keeping the sewing head perpendicular to the normal of the curved surface, which can easily lead to skipped stitches, uneven stitches or material distortion. Upon receiving the dynamic sewing path from the central control module, the task receiving and parsing unit analyzes it and then controls the six-axis robotic arm to perform the sewing. The six-axis robotic arm not only moves the sewing machine head along a three-dimensional trajectory but also continuously adjusts its wrist posture to ensure that the machine head remains perpendicular to the cup's curved surface at any position. Force sensors detect an increasing trend in thread tension when sewing the area with equal curvature at the top of the cup. The CCM then instructs the sewing machine head to fine-tune the thread tensioner to maintain constant tension, thereby achieving high-quality, wrinkle-free, and uniformly stitched quilting on complex three-dimensional curved surfaces. In Example 2 of the sewing process, for example, when processing high-elasticity lace stitching, if the feeding speed and tension are not matched when stitching two pieces of high-elasticity lace fabric, it is very easy for the fabric to be overstretched or wrinkled, resulting in inaccurate dimensions after stitching. The task receiving and parsing unit receives the dynamic sewing path from the central control module. Simultaneously, the pre-calculation has already determined that the lace piece has a 10% pre-stretch, which is compensated for in the dynamic sewing path issued by the central control module. When sewing begins, the force sensor detects that the seam thread generates additional tensile force on the fabric. Using the feedback from the force sensor as the control target, the feeding speed of the robotic arm is dynamically adjusted to provide a constant, small tension that matches the elasticity of the fabric during the sewing process, supporting rather than pulling the fabric during sewing. The resulting seam is smooth and wrinkle-free, and its length matches the design dimensions, effectively suppressing secondary deformation caused by mechanical stretching.

[0039] Furthermore, the trimming execution submodule includes: The robotic arm cutting unit uses a six-axis articulated robotic arm, and an end effector for cutting is installed on the robotic arm; The task parsing and control unit is used to receive and parse the second control command, and decompose it into movement commands for each joint of the robotic arm cutting unit to realize the cutting process.

[0040] Furthermore, the wrist of the robotic arm cutting unit is equipped with a three-dimensional laser profilometer, which is used to perform a final scan of the surface of the semi-finished underwear before the cutting end actuator is activated, and transmits the scan results to the central control module. The central control module compares the scan results with the expected path at the micrometer level, and then performs the final path calibration before cutting.

[0041] Specifically, the task parsing and control unit receives the second control command from the central control module, which is the compensated actual clipping path. This path consists of a series of three-dimensional space points and their normal vectors, represented as:

[0042] in, Represented as the target position coordinates of the blade tip. It is represented as the unit normal vector of the surface at that point, used to determine the orientation of the end effector. During the parsing process, the task parsing and control unit needs to process each pose. Through inverse kinematics, the problem is transformed into the angles of the six joints of the robotic arm. Let the kinematic function be... This will further reduce the joint angle. Mapping to end-effector pose , represented as:

[0043] Solving the inverse problem of the above equations using inverse kinematics is expressed as:

[0044] In other embodiments, such as Figure 7 As shown, the wrist of the robotic arm cutting unit is equipped with a 3D laser profilometer, which scans and obtains the 3D point cloud of the current semi-finished product surface. By using the nearest point iterative ICP algorithm, point cloud registration is performed to find a rigid body transformation T that makes the 3D point cloud of the semi-finished product... Compared with the expected model The minimum error between them is expressed as:

[0045] If there is a deviation after registration Then, the current clipping path is fine-tuned, as shown below:

[0046] Then, based on the fine-tuned clipping path, the joint angles are recalculated and expressed as follows:

[0047] This further improves the accuracy of the fine-tuned cutting and positioning.

[0048] In other embodiments, the cutting end effector may employ contact cutting with a tool or non-contact cutting, such as laser cutting; When using contact cutting, such as ultrasonic scalpels, force sensors can be installed on the robotic arm to sense the cutting force. An impedance control model can then be used to maintain a constant contact force. The core of the control law is to establish a dynamic model of the end effector.

[0049] in , represented as the current position x and the desired position x. Deviation; It is the set inertia, damping, and stiffness matrix; It is the external force measured by the force sensor; The controller adjusts the actual position x, allowing for a certain positional deviation e, to adjust the contact force. converges to the expected value This allows the ultrasonic scalpel to adaptively conform to undulating surfaces, ensuring consistent cutting depth.

[0050] An underwear processing method based on the above-mentioned automated sewing and cutting management system for underwear processing includes: Step S1, Underwear raw material feeding: The underwear raw material is placed on the material conveying and positioning sub-module by manual or machine personnel. The underwear raw material is fixed and conveyed by the flexible coarse positioning unit and the multi-zone vacuum adsorption fine positioning unit on the material conveying and positioning sub-module. Step S2, planning the sewing path of underwear raw materials: The machine vision acquisition submodule takes pictures of the underwear raw materials laid out before sewing and transmits the pre-processed image data to the central control module. The adaptive path planning unit in the central control module calculates the initial deformation based on the built-in integrated algorithm. Based on the initial deformation, the pre-stored standard sewing path is corrected to generate a dynamic sewing path, and then the first control command of the sewing process is generated and transmitted. Step S3: Underwear raw materials are sewn into semi-finished underwear: The multi-axis linkage sewing robot sub-module parses and executes the first control command transmitted by the central control module, thereby sewing the underwear raw materials into semi-finished underwear. Step S4: Underwear semi-finished product cutting path planning: The machine vision acquisition submodule captures images of the sewn underwear semi-finished product and transmits the pre-processed image data to the central control module. The adaptive path planning unit in the central control module calculates the quadratic deformation variable generated by the sewing process based on the built-in integrated algorithm. Based on the quadratic deformation variable, it performs elastic mapping and rearrangement of the pre-stored standard cutting path to generate the actual cutting path, and then generates and transmits the second control command for the cutting process. Step S5: Cutting the semi-finished underwear into finished underwear: The cutting execution submodule parses and executes the second control instruction transmitted by the central control module, thereby cutting the semi-finished underwear into finished underwear. Step S6, Finished Underwear Conveying: The material conveying and positioning submodule conveys the processed underwear to the subsequent processes.

[0051] In the specific implementation process, this embodiment takes the processing of a one-piece seamless sports bra as an example. The process requirements recorded in the central control module include that the main body of the product is formed in one piece from a high-elasticity knitted fabric, and that it needs to be sewn on a three-dimensional curved surface and the final contour cut. It has high requirements for precision and consistency. The specific steps include: Underwear material loading and fixing: The operator lays the cut sports bra knitted fabric flat on the material placement platform. The electromagnetic matrix pins on the platform rise according to the control instructions of the central control module, forming a physical boundary that matches the bra outline, and roughly positioning the underwear material; multi-zone vacuum adsorption is activated, only activating the suction holes in the area covered by the fabric, and evenly and flatly adsorbing and fixing the high elastic fabric, eliminating all wrinkles. Initial visual scan and initial deformation calculation: The material placement platform moves to the pre-sewing station, i.e., the first visual station, under servo drive. The sensor triggers the multispectral industrial camera to capture high-definition images. The ring shadowless light source provides optimal illumination. The image preprocessing and output unit extracts the actual coordinates of the alignment mark points on the fabric and sends them to the central control module. The adaptive path planning unit in the central control module compares the actual coordinates with the theoretical coordinates in the digital twin and calculates the initial deformation of the fabric caused by weaving and laying through the ICP algorithm. Sewing path planning and adaptive sewing: The adaptive path planning unit performs reverse scaling compensation on the standard three-dimensional quilting path based on the calculated initial deformation, generating a dynamic sewing path. The path data is then fitted into a first control command, which is subsequently sent to the multi-axis linkage sewing robot arm submodule. The multi-axis linkage sewing robot arm submodule parses the first control command, and the six-axis sewing robot arm, carrying the integrated sewing head, moves according to the new path. During sewing, the six-dimensional force sensor on the wrist monitors the thread tension in real time, and the central control module dynamically fine-tunes the sewing speed and presser foot pressure to ensure uniform stitches and suppress deformation. After sewing is completed, the robot arm resets, and the platform carries the sewn underwear semi-finished product to the next workstation. Secondary visual scanning and secondary deformation calculation: The material placement platform moves to the sewn workstation, i.e., the second visual workstation, under the servo drive. The sensor triggers the multispectral industrial camera to capture high-definition images. The image preprocessing and output unit extracts the actual coordinates of the alignment mark points on the fabric and sends them to the central control module. The adaptive path planning unit in the central control module compares the actual coordinates with the theoretical coordinates in the digital twin and calculates the secondary deformation of the fabric caused by sewing through the ICP algorithm. Cutting Path Planning and Adaptive Cutting: The adaptive path planning unit generates a dynamic cutting path based on the calculated quadratic deformation, then fits the path data into a second control command, which is then sent to the cutting execution submodule. After parsing the command, the cutting execution submodule uses a 3D laser profilometer to perform high-speed scanning on the sewn semi-finished product to obtain its three-dimensional point cloud data. The point cloud data is registered and compared with the ideal three-dimensional model in the digital twin. For example, if the algorithm detects a 1.5mm concave deformation in the center area of ​​the top of the bra cup due to the tension of the suture line, the data is transmitted to the adaptive path planning unit. The thin plate spline TPS algorithm is used to interpolate the discrete deformation data into a continuous deformation field and perform non-uniform compensation on the initial cutting path, thereby generating an optimized actual cutting path which is sent to the six-axis cutting robot arm. Before cutting, the 3D laser profilometer on the wrist of the robot arm performs a micron-level scan again to perform the final path calibration. The cutting robot arm controls the ultrasonic scalpel to cut along the compensated three-dimensional path; the ultrasonic scalpel vibrates at high frequency, instantly melting and sealing the fabric, resulting in a smooth cut without burrs. Finished lingerie cutting: After cutting, the vacuum suction is turned off, and the operator or collaborative robot takes away the perfectly formed sports bra and puts it into the finished product area; the material placement platform is reset, the electromagnetic pins are reconfigured according to the pattern of the next production order, and the system is ready to start the next cycle.

[0052] This invention discloses an automated sewing and cutting management system for underwear processing. During the underwear processing, through two visual inspections before and after sewing and real-time path compensation, deformation errors caused by material elasticity and processing stress are further eliminated, improving the size qualification rate of underwear products. Simultaneously, it achieves seamless integration and automation of sewing and cutting processes, reducing manual intervention and semi-finished product handling time, further improving production efficiency. The central control module is equipped with a digital twin module, which can store underwear product data according to actual applications, enabling seamless switching of product models without changing physical molds. This allows for rapid changeover of small batches and multiple varieties, greatly enhancing the company's ability to respond to market changes. It digitizes and intelligently transforms complex sewing and cutting processes, reducing the skill requirements for operators and solving the problem of insufficient intelligence and automation in the garment manufacturing industry.

[0053] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. An automated sewing and cutting management system for underwear processing, characterized in that, include: The central control module has a built-in digital twin model, which is used to store several standard pattern data for producing underwear, and to receive, process and control the coordinated execution of each sub-module. The material conveying and positioning submodule is used to fix and convey raw underwear materials and semi-finished underwear products; The machine vision acquisition submodule is used to acquire and transmit image data of the raw underwear material before sewing and the semi-finished underwear product after sewing. The multi-axis linkage sewing robot arm submodule is used to perform sewing operations on the underwear raw materials on the material conveying and positioning submodule; The cutting execution submodule is used to perform cutting operations on the semi-finished underwear products on the material conveying and positioning submodule. The central control module is configured as follows: Based on the image data of the underwear raw material before sewing collected by the machine vision acquisition submodule, the initial deformation of the sheet material is calculated, and then a dynamic sewing path adapted to the initial deformation is generated, and a first control command is generated for the sewing robotic arm module to execute. Based on the image data of the sewn underwear semi-finished product acquired by the machine vision acquisition submodule, the secondary deformation caused by the sewing operation is calculated, and then an actual cutting path to compensate for the secondary deformation is generated, and a second control command is generated for the cutting execution module to execute.

2. The automated sewing and cutting management system for underwear processing according to claim 1, characterized in that, The central control module includes: The main control and scheduling unit is responsible for coordinating the timing control and task scheduling of the control system workflow. According to the preset process formula, it triggers the various actions of visual scanning, robotic arm movement, sewing, and cutting in an orderly manner to ensure that all sub-modules work together in the correct order. Digital twin units are used to store standardized model data of underwear styles to be produced, which can be retrieved during production; The adaptive path planning unit, based on a built-in integrated algorithm, is used to analyze the image data acquired by the machine vision acquisition submodule, and then generate the stitching path or cutting path for the corresponding process.

3. The automated sewing and cutting management system for underwear processing according to claim 1, characterized in that, The machine vision acquisition submodule includes: The industrial camera imaging unit includes several industrial cameras that can switch between different wavelength light sources to enhance the ability to identify feature points of different colors, textures and transparent underwear materials. They are set up at the sewing station and the cutting station respectively to collect image data of underwear materials before sewing and image data of underwear semi-finished products after sewing in real time. The trigger shooting unit includes several pressure sensors, which are respectively set on the material conveying and positioning submodule to identify the entry of underwear raw materials or underwear semi-finished products, so as to trigger the real-time data acquisition of the industrial camera shooting unit. The imaging lighting unit uses a ring-shaped LED shadowless light source and is respectively set at the sewing station and the cutting station to enhance the imaging effect of the industrial camera imaging unit; The image preprocessing and output unit is used to preprocess the image data acquired by the industrial camera capturing unit using an image preprocessing method, and then use feature extraction and recognition algorithms to convert the preprocessed image data into engineering data and output it to the central control module.

4. The automated sewing and cutting management system for underwear processing according to claim 1, characterized in that, The material conveying and positioning submodule includes: Material placement platform, used for placing lingerie raw materials and semi-finished lingerie products; The flexible coarse positioning unit, composed of electromagnetic matrix pins, is set on the material placement platform and is used to dynamically form a first positioning area for coarse positioning based on the shape and size of the underwear raw materials and underwear semi-finished products. The multi-zone vacuum adsorption precision positioning unit consists of several vacuum chamber partitions for vacuum adsorption of underwear raw materials and underwear semi-finished products, which are respectively set on the material placement platform. Each vacuum chamber partition is independently controlled to dynamically travel to the second positioning zone for precision positioning, so as to adapt to underwear raw materials and underwear semi-finished products of different sizes and shapes. The servo drive and transmission unit is responsible for moving the material placement platform so that the underwear raw materials and underwear semi-finished products can pass through the sewing and cutting processes in sequence.

5. The automated sewing and cutting management system for underwear processing according to claim 1, characterized in that, The multi-axis linkage suture robotic arm submodule includes: The robotic arm sewing unit uses a six-axis articulated robotic arm, which is equipped with an integrated sewing machine head for sewing. The task parsing and control unit is used to receive and parse the first control command, decompose it into movement commands for each joint of the robotic arm suturing unit, so as to realize the suturing process; The safety protection unit, consisting of several torque sensors, is installed on the robotic arm sewing unit. When a foreign object or human body is detected, it promptly performs an emergency stop.

6. The automated sewing and cutting management system for underwear processing according to claim 1, characterized in that, The trimming execution submodule includes: The robotic arm cutting unit uses a six-axis articulated robotic arm, and an end effector for cutting is installed on the robotic arm; The task parsing and control unit is used to receive and parse the second control command, and decompose it into movement commands for each joint of the robotic arm cutting unit to realize the cutting process.

7. An automated sewing and cutting management system for underwear processing according to claim 5, characterized in that, The wrist of the robotic arm suturing unit is equipped with a six-dimensional force sensor for real-time monitoring of the needle and thread tension during the suturing process. The central control module is further configured to receive feedback data from the six-dimensional force sensor and then dynamically adjust the sewing speed, stitch length, and presser foot pressure of the integrated sewing machine head to form a closed-loop control for the sewing tension, thereby actively suppressing material deformation caused by the sewing process itself.

8. An automated sewing and cutting management system for underwear processing according to claim 6, characterized in that, The wrist of the robotic arm cutting unit is equipped with a three-dimensional laser profilometer, which is used to perform a final scan of the surface of the semi-finished underwear before the cutting end effector is activated, and transmits the scan results to the central control module. The central control module compares the scan results with the expected path at the micrometer level, and then performs the final path calibration before cutting.

9. A method for processing underwear based on an automated sewing and cutting management system for underwear processing according to any one of claims 1-8, characterized in that, include: Step S1, Underwear raw material feeding: The underwear raw material is placed on the material conveying and positioning sub-module by manual or machine personnel. The underwear raw material is fixed and conveyed by the flexible coarse positioning unit and the multi-zone vacuum adsorption fine positioning unit on the material conveying and positioning sub-module. Step S2, planning the sewing path of underwear raw materials: The machine vision acquisition submodule takes pictures of the underwear raw materials laid out before sewing and transmits the pre-processed image data to the central control module. The adaptive path planning unit in the central control module calculates the initial deformation based on the built-in integrated algorithm. Based on the initial deformation, the pre-stored standard sewing path is corrected to generate a dynamic sewing path, and then the first control command of the sewing process is generated and transmitted. Step S3: Underwear raw materials are sewn into semi-finished underwear: The multi-axis linkage sewing robot sub-module parses and executes the first control command transmitted by the central control module, thereby sewing the underwear raw materials into semi-finished underwear. Step S4: Underwear semi-finished product cutting path planning: The machine vision acquisition submodule captures images of the sewn underwear semi-finished product and transmits the pre-processed image data to the central control module. The adaptive path planning unit in the central control module calculates the quadratic deformation variable generated by the sewing process based on the built-in integrated algorithm. Based on the quadratic deformation variable, it performs elastic mapping and rearrangement of the pre-stored standard cutting path to generate the actual cutting path, and then generates and transmits the second control command for the cutting process. Step S5: Cutting the semi-finished underwear into finished underwear: The cutting execution submodule parses and executes the second control instruction transmitted by the central control module, thereby cutting the semi-finished underwear into finished underwear. Step S6, Finished Underwear Conveying: The material conveying and positioning submodule conveys the processed underwear to the subsequent processes.

Images