A method and system for controlling garment cutting

Abstract

The present application belongs to the technical field of garment cutting, and relates to a garment cutting control method and system. By obtaining fabric feature information and determining cutting parameters accordingly, a pre-performance mechanism based on a digital model of a cutting head motion system is introduced, which can predict in advance mechanical vibration and / or contour deviation and other interference information caused by cutting parameter switching when executing a cutting instruction sequence. For these predicted interference information, corresponding compensation instructions are generated, including vibration suppression instructions and / or contour geometry correction instructions, and these compensation instructions are embedded into the cutting instruction sequence to form a second cutting instruction sequence to control the cutting head execution. Through forward-looking pre-performance and compensation by the digital model, the mechanical impact and positioning error caused by parameter switching are actively offset while ensuring that the cutting head can adapt to the local physical properties of the fabric.

Description

Technical Field

[0001] This application relates to the field of garment cutting technology, and more specifically, to a garment cutting control method and system. Background Technology

[0002] In automated garment production, automatic cutting tables play a crucial role. Through fabric laying, vacuum adsorption, and loading of the pattern, the XY motion system drives the cutting head to complete precise cutting at high speed. The goal is to balance accuracy, edge quality, and speed. However, fabrics are not uniform: denim has slight differences in thickness, density, and yarn tension, and elastic knitted fabrics also have varying densities after being laid out. If fixed parameters (cutting frequency, cutting depth, speed, or laser power) are used, thick areas may not be cut through, resulting in frayed edges and requiring rework at reduced speed; while sparse areas are prone to overcutting and damaging the fabric, creating a dilemma between efficiency and quality.

[0003] Existing technology incorporates pre-scanning: using non-contact sensors to quickly acquire the fabric's 3D / texture, generating a "feature map" aligned with the layout coordinates. The control system compares the path with the map in real time, adjusting parameters in advance to ensure a clean cut without damage. However, this introduces a new challenge: the electromechanical system of the cutting head has mass and inertia, requiring constant acceleration and deceleration to match frequently changing parameters. Especially with small cut pieces with complex contours and drastically changing features, continuous impact induces vibration and hysteresis, causing the actual trajectory to deviate from the command, resulting in "overcutting" or "undercutting." Thus, the system faces a dilemma: maintaining edge quality sacrifices geometric accuracy, while maintaining geometric accuracy makes it difficult to adapt to material variations.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] This application discloses a garment cutting control method and system, which aims to solve the problems of inconsistent cutting quality caused by fabric unevenness in existing automated garment production, as well as the mechanical vibration and contour deviation generated by the cutting head motion system when frequently adjusting cutting parameters to adapt to fabric changes.

[0006] The technical solution of this application is as follows:

[0007] In a first aspect, this application discloses a garment cutting control method, which includes: acquiring fabric feature information and determining cutting parameters based on the fabric feature information; the fabric feature information includes fabric thickness and / or density; the cutting parameters include tool vibration frequency, cutting depth, and forward speed; acquiring a cutting path and determining a corresponding set of cutting parameters based on the cutting path, and generating a first cutting instruction sequence; performing a pre-simulation of the movement of the cutting head on the cutting path based on a digital model of the cutting head motion system, so as to predict the interference information generated by the cutting head when executing the cutting instruction sequence; the interference information includes mechanical vibration and / or contour deviation caused by switching cutting parameters when executing the first cutting instruction sequence; generating a compensation instruction to cancel the interference information based on the predicted interference information; the compensation instruction includes a vibration suppression instruction to suppress mechanical vibration and / or a contour geometry correction instruction to correct contour deviation; embedding the compensation instruction into the cutting instruction sequence to generate a second cutting instruction sequence, and controlling the cutting head to execute the second cutting instruction sequence.

[0008] Furthermore, acquiring fabric feature information includes: performing non-contact scanning of the fabric to obtain three-dimensional topographic data of the fabric surface and generating a fabric feature map; determining the three-dimensional height and / or distance information of each point on the fabric surface based on the three-dimensional topographic data; and converting the three-dimensional height and / or distance information into quantitative values ​​representing fabric feature information.

[0009] Based on this, the cutting path is obtained and a set of corresponding cutting parameters is determined according to the cutting path to generate the first cutting instruction sequence, including: aligning the fabric feature map with the garment sample layout in the coordinate system; querying the local physical feature information of the fabric corresponding to each point of the cutting path based on the fabric feature map; the local physical feature information of the fabric includes the local fabric thickness and / or density; determining the cutting parameters of each point according to the local physical feature information of the fabric corresponding to each point and the preset cutting parameter lookup table to generate the first cutting instruction sequence.

[0010] In some preferred embodiments, the digital model of the cutting head motion system is a model of the components of the cutting head using finite element analysis software; the digital model outputs the response of the cutting head under different driving forces;

[0011] Based on the predicted interference information, a compensation instruction is generated to counteract the interference, including:

[0012] The micro-vibration signal of the cutting head is monitored in real time, and the frequency, phase and amplitude of the micro-vibration signal are determined by real-time spectrum analysis. Based on the micro-vibration signal analysis results, an anti-vibration waveform with opposite phase, same frequency and matching amplitude to the detected vibration waveform is dynamically generated. The anti-vibration waveform is superimposed on the cutting command sequence of the cutting head.

[0013] Furthermore, based on the predicted interference information, a compensation command is generated to counteract the interference information. This also includes: monitoring the micro-vibration signal of the cutting head in real time, and monitoring the instantaneous current signal of the servo motor of the cutting head; comparing the micro-vibration signal and the instantaneous current signal with the expected motion data predicted by the digital model to determine the additional transient micro-disturbance caused by the micro-material of the fabric; and dynamically generating a reaction force command based on the additional transient micro-disturbance.

[0014] Preferably, embedding the compensation instruction into the trimming instruction sequence to generate the second trimming instruction sequence further includes: integrating the reaction force instruction, vibration suppression instruction and / or contour geometry correction instruction into the trimming instruction sequence to form the second trimming instruction sequence.

[0015] Based on the above, the micro-vibration signal and instantaneous current signal are compared with the expected motion data predicted by the digital model to determine the additional transient micro-perturbations caused by the fabric micro-materials. This includes: comparing the micro-vibration signal and instantaneous current signal with the expected motion data predicted by the digital model to obtain the residual signal; and performing multi-band decomposition on the residual signal to obtain multiple frequency sub-band signals.

[0016] Within each frequency sub-band signal, feature extraction is performed on the frequency sub-band signal to obtain the vibration amplitude, phase, and frequency center of each frequency sub-band; based on the vibration amplitude, phase, and frequency center of each frequency sub-band, additional transient micro-perturbations are identified.

[0017] Furthermore, based on the additional transient micro-perturbation, a reaction force command is dynamically generated, including: dynamically generating a composite reaction force command based on the waveform characteristics of the additional transient micro-perturbation; after executing the cutting command sequence that applies the composite reaction force command, continuously monitoring the residual vibration of the cutting head and acquiring the residual vibration signal; and adjusting the composite reaction force command based on the residual vibration signal.

[0018] More specifically, adjusting the composite reaction force command based on the residual vibration signal includes: predicting the evolution trend of the residual vibration within a short time window based on the instantaneous frequency and amplitude variation trend of the residual vibration signal; and dynamically adjusting the amplitude and / or phase of the composite reaction force command based on the evolution trend.

[0019] Secondly, this application also discloses a garment cutting control system, which includes: an acquisition module for acquiring fabric feature information and determining cutting parameters based on the fabric feature information; the fabric feature information includes fabric thickness and / or density; the cutting parameters include blade vibration frequency, cutting depth, and forward speed; a first generation module for acquiring a cutting path and determining a corresponding set of cutting parameters based on the cutting path, and generating a first cutting instruction sequence; a pre-simulation module for pre-simulating the movement of the cutting head on the cutting path based on a digital model of the cutting head motion system, so as to predict the interference information generated by the cutting head when executing the cutting instruction sequence; the interference information includes mechanical vibration and / or contour deviation caused by switching cutting parameters when executing the first cutting instruction sequence; a second generation module for generating compensation instructions to cancel the interference information based on the predicted interference information; the compensation instructions include vibration suppression instructions to suppress mechanical vibration and / or contour geometry correction instructions to correct contour deviation; and a control module for embedding the compensation instructions into the cutting instruction sequence to generate a second cutting instruction sequence, and controlling the cutting head to execute the second cutting instruction sequence.

[0020] Beneficial effects

[0021] The garment cutting control method disclosed in this application effectively addresses the inherent unevenness of the fabric by acquiring fabric characteristic information and determining cutting parameters accordingly, thus avoiding cutting quality problems caused by fixed parameters. More importantly, this method introduces a pre-simulation mechanism based on a digital model of the cutting head motion system, which can predict in advance the interference information such as mechanical vibration and / or contour deviation caused by the switching of cutting parameters during the execution of the cutting instruction sequence. For these predicted interference information, this method can generate corresponding compensation instructions, including vibration suppression instructions and / or contour geometry correction instructions, and embed these compensation instructions into the cutting instruction sequence to form a second cutting instruction sequence to control the execution of the cutting head.

[0022] Through the above technical solution, this application overcomes the problems of mechanical vibration and contour deviation that may occur in the cutting head motion system when frequently adjusting cutting parameters to adapt to changes in fabric, as seen in existing technologies. Instead of passively adjusting during the cutting process, it uses a digital model for proactive pre-simulation and compensation, thereby ensuring the cutting head can adapt to the local physical characteristics of the fabric while actively offsetting the mechanical impact and positioning errors caused by parameter switching. This effectively solves the dilemma in existing technologies of "sacrificing contour accuracy to ensure edge quality, or failing to fully adapt to material changes to ensure contour accuracy," significantly improving the overall accuracy, edge quality, and production efficiency of garment cutting. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating the steps of the clothing cutting control method disclosed in the embodiments of the present invention;

[0025] Figure 2 This is a schematic diagram of the clothing cutting control system disclosed in an embodiment of the present invention. Detailed Implementation

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which these embodiments belong; the terminology used herein and in the specification of the application is for the purpose of describing particular embodiments only and is not intended to limit these embodiments; the terms "comprising" and "having," and any variations thereof, in the specification of these embodiments and the foregoing drawings, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification of these embodiments and the foregoing drawings are used to distinguish different objects, not to describe a particular order.

[0027] The implementation details of the technical solution in this embodiment are described in detail below:

[0028] In traditional automated garment production, automatic cutting beds acquire fabric feature information through a pre-scanning process and adjust cutting parameters accordingly to adapt to fabric unevenness, thereby improving cutting quality. However, this frequent parameter adjustment, especially the rapid changes in the cutting head's movement speed, can subject the cutting head's motion system to continuous impacts, leading to mechanical vibration and positioning errors, which in turn affect the geometric accuracy of the cut contour. This parameter adjustment behavior, introduced to address material issues, may ironically become the "culprit" that damages cutting geometric accuracy, putting the system in a dilemma.

[0029] In response, this application proposes a method for controlling garment cutting, such as... Figure 1 As shown, the method includes:

[0030] S101, acquire fabric feature information and determine cutting parameters based on the fabric feature information; the fabric feature information includes fabric thickness and / or density; the cutting parameters include tool vibration frequency, cutting depth and forward speed;

[0031] S102, Obtain the cutting path and determine a set of cutting parameters corresponding to the cutting path to generate a first cutting instruction sequence;

[0032] S103, Based on the digital model of the cutting head motion system, the movement of the cutting head on the cutting path is pre-simulated to predict the interference information generated by the cutting head when executing the cutting instruction sequence; the interference information includes mechanical vibration and / or contour deviation caused by switching the cutting parameters when executing the first cutting instruction sequence.

[0033] S104, Based on the predicted interference information, generate a compensation instruction to counteract the interference information; the compensation instruction includes a vibration suppression instruction to suppress mechanical vibration and / or a contour geometry correction instruction to correct the contour deviation;

[0034] S105, the compensation instruction is embedded into the cutting instruction sequence to generate a second cutting instruction sequence, and the cutting head is controlled to execute the second cutting instruction sequence.

[0035] This embodiment uses a digital model of the cutting head motion system for pre-simulation, which can predict interference information such as mechanical vibration and contour deviation caused by the switching of cutting parameters in advance, and generate compensation instructions accordingly. These instructions are then embedded into the cutting instruction sequence, thereby effectively counteracting these interferences during the actual cutting process, ensuring cutting accuracy and edge quality, while also taking into account production efficiency.

[0036] The garment cutting control method proposed in this embodiment aims to solve the cutting quality problems caused by fabric unevenness in automated garment production, as well as the resulting mechanical vibration and contour deviation problems of the cutting head motion system. "Fabric characteristic information" refers to data describing the physical properties of the fabric, such as its thickness, density, elasticity, and texture. This information is crucial for determining appropriate cutting parameters. "Cutting parameters" are the settings that control the working state of the cutting head, such as the vibration frequency of the cutter, the cutting depth, and the forward speed. These parameters directly affect the cutting effect and efficiency. "Cutting path" refers to the predetermined trajectory of the cutting head moving on the fabric, usually determined by the garment pattern layout. "Cutting instruction sequence" consists of a series of control instructions used to guide the cutting head along the cutting path and perform the cutting operation. "Digital model of the cutting head motion system" is a mathematical model of the cutting head and its drive system, capable of simulating the dynamic response of the cutting head under different working conditions, thereby predicting its motion behavior and potential disturbances. "Interference information" refers to unexpected effects, such as mechanical vibration and contour deviation, that occur during the cutting process due to changes in the system's dynamic response or fabric characteristics. "Compensation instructions" are control instructions generated based on the predicted interference information and used to counteract or mitigate these interferences, including "vibration suppression instructions" and "contour geometry correction instructions".

[0037] In practical implementation, the garment cutting control method first needs to acquire fabric characteristic information and determine cutting parameters based on this information. Fabric characteristic information may include fabric thickness and / or density. For example, this can be obtained through manual measurement or by scanning the fabric using a non-contact sensor. As a preferred embodiment, the operator can manually input the average thickness and density values ​​of the fabric. As another preferred embodiment, a laser rangefinder or ultrasonic sensor can be used to perform local scanning of the fabric to obtain thickness data for different areas. Based on the acquired fabric characteristic information, corresponding cutting parameters can be determined, such as the blade vibration frequency, cutting depth, and forward speed. For example, for thicker fabrics, a higher blade vibration frequency and a larger cutting depth can be set; for thinner fabrics, the blade vibration frequency and cutting depth can be reduced.

[0038] Subsequently, the cutting path is acquired, and a corresponding set of cutting parameters is determined based on the cutting path to generate the first cutting instruction sequence. The cutting path is usually provided by the garment pattern layout. As a preferred implementation, the cutting path can be directly imported from CAD / CAM software. As another preferred implementation, the cutting path can be extracted from the paper pattern using image recognition technology. Based on the fabric feature information corresponding to each point on the cutting path, a preset cutting parameter lookup table can be consulted to determine a set of optimal cutting parameters for each point or line segment on the cutting path. For example, when the cutting head is about to enter a thicker area of ​​the fabric, the system will adjust the cutting depth and vibration frequency of the blade in advance. These parameters, combined with the cutting path information, generate the first cutting instruction sequence.

[0039] Next, based on the digital model of the cutting head motion system, the motion of the cutting head along the cutting path is pre-simulated to predict the interference information generated by the cutting head when executing the first cutting command sequence. The digital model of the cutting head motion system can be a simulation model based on physical principles. For example, by modeling the mechanical structure, motor characteristics, and transmission mechanism of the cutting head, its dynamic response under different driving forces can be simulated. As a preferred implementation, the digital model can be a simplified kinematic model, considering only the velocity and acceleration of the cutting head. As another preferred implementation, the digital model can be a more complex dynamic model, considering factors such as the system's inertia, damping, and stiffness. By inputting the first cutting command sequence into the digital model for pre-simulation, the mechanical vibration and / or contour deviation that may occur during the actual execution of the cutting head can be predicted. For example, when the cutting parameters are frequently switched, especially when the velocity changes drastically, the digital model can predict the transient vibration amplitude and frequency that the cutting head may generate, and the resulting deviation between the actual cutting trajectory and the ideal cutting path.

[0040] Based on the predicted interference information, compensation instructions are generated to counteract the interference. These compensation instructions may include vibration suppression instructions to suppress mechanical vibration and / or contour geometry correction instructions to correct contour deviations. As a preferred embodiment, if it is predicted that the cutting head will generate significant mechanical vibration in a certain area, an inverse vibration signal can be generated as a vibration suppression instruction, and the vibration can be counteracted by controlling the cutting head's drive system to apply a reaction force. As another preferred embodiment, if it is predicted that a deviation in the cutting contour will occur, a geometry correction instruction can be generated to fine-tune the cutting path to compensate for the expected deviation. For example, if it is predicted that the cutting head will "overcut" at a corner, the cutting path at that corner can be adjusted to shift it slightly inward, thereby obtaining an accurate contour during actual cutting.

[0041] Finally, the compensation instruction is embedded into the cutting instruction sequence to generate a second cutting instruction sequence, which controls the cutting head to execute the second cutting instruction sequence. As a preferred embodiment, the compensation instruction can be directly superimposed on the first cutting instruction sequence to form a new control instruction. As another preferred embodiment, the compensation instruction can be used as an independent control signal, executed in parallel with the first cutting instruction sequence to jointly control the movement of the cutting head. By executing the second cutting instruction sequence containing the compensation instruction, the movement of the cutting head will be precisely controlled during the actual cutting process, and the expected mechanical vibration and contour deviation will be effectively suppressed or corrected, thereby ensuring the accuracy and quality of the cutting.

[0042] The garment cutting control method proposed in this application pre-simulates the movement of the cutting head using a digital model, enabling it to predict interference such as mechanical vibration and contour deviation caused by switching cutting parameters. Specifically, the method first acquires fabric feature information and determines cutting parameters, then generates a first cutting instruction sequence based on the cutting path. Based on this, the digital model is used to pre-simulate the movement of the cutting head, accurately predicting the mechanical vibration and contour deviation that may occur when executing the first cutting instruction sequence. Once these interferences are identified, the system generates corresponding compensation instructions, such as vibration suppression instructions and contour geometry correction instructions. These compensation instructions are then embedded into the original cutting instruction sequence to form a second cutting instruction sequence, which is ultimately used to control the cutting head for actual cutting. The entire process forms a closed-loop prediction-compensation mechanism, enabling the cutting head to effectively offset the negative impacts of material unevenness and frequent parameter switching during high-speed, high-precision cutting, ensuring the quality of the cut edges and the geometric accuracy of the contour.

[0043] Compared with existing technologies, the garment cutting control method of this application has significant advantages. While existing technologies address the cutting quality issues caused by fabric unevenness through pre-scanning and real-time adjustment of cutting parameters, they neglect the mechanical vibrations and positioning errors generated by frequent parameter adjustments on the cutting head motion system. This leads to a dilemma: while pursuing cutting quality, the geometric accuracy of the cutting contour is sacrificed. The core innovation of this application lies in introducing a digital model of the cutting head motion system for pre-simulation, building upon existing technologies. By simulating the movement of the cutting head along the cutting path, interference information such as mechanical vibrations and contour deviations caused by cutting parameter switching can be predicted in advance and accurately. This "predictive" capability allows the system to generate targeted compensation instructions, such as vibration suppression instructions and contour geometry correction instructions, before actual cutting occurs. These compensation instructions are seamlessly embedded into the cutting instruction sequence, thereby actively canceling or correcting anticipated interferences during the actual cutting process. Therefore, this application not only effectively adapts to fabric unevenness and ensures the quality of the cut edges, but more importantly, it solves the problems of mechanical vibration and contour deviation caused by parameter adjustments in existing technologies, significantly improving the geometric accuracy of the cut contour. This predictive-compensation mechanism enables the cutting system to operate at high speed while maintaining high precision and high quality, breaking through the dilemma between efficiency and precision in existing technologies and achieving a major advancement in automated garment cutting technology.

[0044] In some embodiments described above in this application, a method is proposed to obtain fabric feature information and determine cutting parameters based on the fabric feature information. Specifically, obtaining the fabric feature information includes:

[0045] The fabric is scanned non-contactly to obtain three-dimensional topographic data of the fabric surface and generate a fabric feature map; the three-dimensional height and / or distance information of each point on the fabric surface is determined based on the three-dimensional topographic data; and the three-dimensional height and / or distance information is converted into quantitative values ​​representing fabric feature information.

[0046] Specifically, non-contact scanning refers to the rapid and accurate acquisition of geometric information about a fabric surface using devices such as optical sensors, laser scanners, or structured light projectors without physical contact. This method avoids any physical damage or deformation to the fabric, making it particularly suitable for delicate or fragile fabrics. The acquired three-dimensional topographic data is point cloud data or mesh data of the fabric surface in three-dimensional space, which comprehensively reflects the microscopic features of the fabric surface, such as undulations, texture, and local thickness variations.

[0047] The fabric feature map is constructed based on the aforementioned 3D topographic data. It visually presents the 3D height and / or distance information of the fabric surface and accurately maps it to the actual position of the fabric. This map may include geometric attributes such as the 3D coordinates, normal vectors, and curvature of each point on the fabric surface, as well as physical attributes such as local thickness and density obtained through further processing. The 3D height and / or distance information may, for example, refer to the height value of the fabric surface relative to a reference plane, or the spatial distance between any two points on the fabric surface. This information is converted into quantitative values ​​representing fabric feature information. For example, by statistically analyzing local height changes, the local thickness of the fabric can be obtained; by calculating the density of the surface texture, the local density of the fabric can be obtained. These quantitative values ​​are crucial for subsequently determining cutting parameters.

[0048] This embodiment employs non-contact scanning technology to accurately and non-destructively acquire three-dimensional topographic data of the fabric surface. This allows for the generation of a detailed fabric feature map, containing three-dimensional height and / or distance information for each point on the fabric surface. This three-dimensional information is further converted into quantified values, accurately characterizing key features of the fabric such as local thickness and density. This refined method of acquiring fabric feature information provides a more reliable and accurate data foundation for subsequently determining cutting parameters based on these features, ensuring that the cutting parameters better adapt to the actual physical properties of the fabric, thereby improving cutting accuracy and effectiveness.

[0049] The above technical solution achieves non-contact, high-precision acquisition of fabric feature information, avoiding fabric damage or deformation that may occur with traditional contact measurements. By generating a fabric feature map and converting it into quantifiable values, a more comprehensive and detailed understanding of the fabric's local physical properties, such as local thickness and density, can be obtained. This precise fabric feature information makes the determination of subsequent cutting parameters more targeted and accurate, thereby effectively improving the precision and quality of garment cutting, reducing cutting errors caused by differences in fabric characteristics, and extending the service life of cutting tools.

[0050] This embodiment further proposes the following steps for obtaining the cutting path and determining a corresponding set of cutting parameters based on the cutting path to generate the first cutting instruction sequence:

[0051] Align the fabric feature map with the garment sample layout in the coordinate system;

[0052] Based on the fabric feature map, query the local physical feature information of the fabric corresponding to each point of the cutting path; the local physical feature information of the fabric includes the fabric thickness and / or density in the local area;

[0053] Based on the local physical characteristics of the fabric corresponding to each point and the preset cutting parameter lookup table, the cutting parameters of each point are determined to generate the first cutting instruction sequence.

[0054] Specifically, aligning the fabric feature map with the garment pattern layout in a coordinate system refers to using image processing or sensor positioning technology to precisely match the fabric feature map obtained through non-contact scanning with the garment pattern layout to be cut in a unified two-dimensional or three-dimensional coordinate system. The purpose is to ensure that every point on the cutting path accurately corresponds to its corresponding position on the fabric feature map, thereby obtaining precise local physical feature information of the fabric at that location.

[0055] The process of querying the local physical features of the fabric at each point along the cutting path based on the fabric feature map can be understood as the system scanning point-by-point or segment-by-segment along the predetermined cutting path and searching for physical attributes such as fabric thickness and / or density in the local area corresponding to the current cutting point on the fabric feature map. In practical applications, the local physical features of the fabric specifically refer to the fabric thickness and / or density of each tiny region along the cutting path. For example, interpolation algorithms or nearest neighbor algorithms can be used to extract these local feature values ​​from the fabric feature map, with the aim of providing refined input data for subsequent cutting parameter determination.

[0056] Furthermore, based on the local physical characteristics of the fabric at each point and a pre-defined cutting parameter lookup table, the cutting parameters for each point are determined. This means that the system matches the local fabric thickness and / or density at each cutting point against a pre-established cutting parameter lookup table. This lookup table typically contains the correspondence between different fabric characteristics (such as thickness and density) and optimal cutting parameters (such as tool vibration frequency, cutting depth, and forward speed). In this way, the most suitable cutting parameters can be dynamically determined for each point on the cutting path, thereby generating a refined first cutting instruction sequence.

[0057] This embodiment precisely aligns the fabric feature map with the garment pattern layout, ensuring that every point on the cutting path accurately maps to the actual physical location of the fabric. Therefore, the system can precisely query the local physical characteristics of the fabric corresponding to each point on the cutting path based on the fabric feature map, such as the fabric thickness and / or density in local areas. It is precisely because of this acquisition of detailed local fabric feature data that the system can dynamically determine the most suitable cutting parameters for each point on the cutting path based on this local information and a pre-set cutting parameter lookup table. This point-by-point or segmental parameter determination method effectively solves the problem of cutting parameter mismatch caused by the uneven local characteristics of the fabric in traditional methods, thus ensuring the best cutting effect throughout the entire cutting path.

[0058] In some preferred embodiments, a specific example is given below. Suppose there is a piece of fabric used to make clothing, with slight differences in thickness in different areas. For example, the edges of the fabric may be slightly thicker due to hemming or compaction, while the center area is relatively uniform. First, a fabric feature map is generated using non-contact scanning, which records detailed three-dimensional height information of each point on the fabric surface and converts it into quantified thickness data. Then, this fabric feature map is precisely aligned with the layout of the garment sample to be cut in a unified coordinate system. As the cutting head moves along the preset cutting path, the system queries the local fabric thickness information corresponding to the current cutting point on the fabric feature map in real time. For example, when the cutting head moves to a thicker area at the fabric edge, the system matches a set of corresponding cutting parameters from a preset cutting parameter lookup table based on the larger thickness value found. For example, it may be necessary to increase the blade vibration frequency and / or the cutting depth. When the cutting head moves to a thinner area in the center of the fabric, it is matched with another set of cutting parameters, such as appropriately reducing the blade vibration frequency and / or the cutting depth. In this way, each point on the cutting path can obtain cutting parameters that match its local fabric characteristics, thereby generating a series of finely adjusted first cutting instructions to ensure optimal cutting results throughout the entire cutting process.

[0059] This embodiment further proposes the above-mentioned garment cutting control method, wherein the digital model of the cutting head motion system is a model of the components of the cutting head using finite element analysis software; the digital model outputs the response of the cutting head under different driving forces; based on the predicted interference information, compensation instructions for canceling the interference information are generated, including: real-time monitoring of the micro-vibration signal of the cutting head, and real-time spectrum analysis of the micro-vibration signal to determine the frequency, phase, and amplitude of the micro-vibration signal; based on the micro-vibration signal analysis results, dynamically generating an anti-vibration waveform that is opposite in phase, has the same frequency, and matches the amplitude of the detected vibration waveform; and superimposing the anti-vibration waveform onto the cutting instruction sequence of the cutting head.

[0060] Specifically, the digital model of the cutting head motion system refers to a virtual model constructed by using finite element analysis (FEA) software to meticulously model each component of the cutting head. This digital model can simulate the dynamic response of the cutting head under different driving forces (e.g., driving force generated by a servo motor, interaction force between the cutter and the fabric, etc.), including its displacement, velocity, acceleration, and possible vibration modes. The digital model outputs the expected motion data and vibration characteristics of the cutting head under specific driving forces, providing a precise theoretical basis for subsequent interference prediction and compensation command generation.

[0061] The real-time monitoring of the micro-vibration signal of the cutting head refers to the continuous acquisition of minute, high-frequency mechanical vibration data generated by the cutting head during operation using high-precision sensors (such as accelerometers, laser displacement sensors, etc.) installed on or near the cutting head. Real-time spectrum analysis of the micro-vibration signal involves converting the acquired time-domain vibration signal into a frequency-domain signal using digital signal processing techniques such as Fourier transform, thereby accurately determining the frequency components, amplitude, and phase information of the current micro-vibration signal.

[0062] Furthermore, based on the analysis results of the micro-vibration signals, an anti-vibration waveform is dynamically generated that is opposite in phase, has the same frequency, and matches the amplitude of the detected vibration waveform. This means that the system can calculate and generate a compensation waveform that is exactly the same in frequency and amplitude as the current interfering vibration waveform, but precisely opposite in phase, based on the vibration characteristics acquired in real time. This anti-vibration waveform aims to cancel out the original mechanical vibration through active intervention. Dynamic generation emphasizes that the system can quickly adjust the parameters of the anti-vibration waveform according to the real-time changes in the vibration signal to adapt to constantly changing working conditions.

[0063] Therefore, the anti-vibration waveform is superimposed onto the cutting command sequence of the cutting head. This means that the generated anti-vibration waveform is fused with the original cutting path command in the form of control commands. These superimposed commands drive the actuator of the cutting head (e.g., a servo motor) to generate a force or movement opposite to the interfering vibration while performing the predetermined cutting action, thereby effectively suppressing mechanical vibration.

[0064] The solution proposed in this application constructs a high-precision digital model of the cutting head motion system, enabling more accurate prediction of the cutting head's dynamic behavior under different working conditions, thus providing a solid foundation for vibration prediction. Furthermore, by real-time monitoring of the micro-vibration signals of the cutting head and performing spectral analysis, the system can accurately capture the frequency, phase, and amplitude of the actual vibrations. Based on this real-time data, an anti-vibration waveform with the opposite phase, the same frequency, and matching amplitude to the detected vibration waveform is dynamically generated, ensuring the accuracy and timeliness of the compensation commands. Therefore, by superimposing the anti-vibration waveform onto the cutting command sequence, the cutting head can actively generate a reaction force to counteract its own vibration while performing the cutting task, thereby effectively suppressing mechanical vibration.

[0065] As a specific implementation, suppose that when cutting a high-density fabric along a complex curve, the cutting head needs to frequently switch its forward speed and cutting depth. First, key components of the cutting head, such as the cutter, cutter holder, and drive motor, are modeled using finite element analysis software to obtain a digital model that accurately reflects its dynamic response. This model can predict the natural frequencies and vibration modes that the cutting head may generate under a specific driving force. During the actual cutting process, a miniature accelerometer mounted on the cutting head collects its micro-vibration signals in real time. These signals are sent to a signal processing unit for real-time spectrum analysis, quickly identifying the frequency, phase, and amplitude of the current vibration signal. For example, if significant vibration of the cutting head is detected at a specific frequency, the system immediately generates a counter-vibration waveform with the same frequency and amplitude but with an exact opposite phase, based on that frequency, phase, and amplitude. Subsequently, this counter-vibration waveform is converted into an electrical signal and superimposed on the servo motor commands controlling the cutting head's movement. While executing the cutting path commands, the servo motor also generates a corresponding reaction force based on the superimposed counter-vibration waveform, thereby actively counteracting the mechanical vibration of the cutting head. In this way, even under high-speed, high-dynamic cutting conditions, the cutting head can maintain stable operation, ensuring the accuracy of the cutting path and avoiding burrs or contour distortion caused by vibration.

[0066] This embodiment further proposes the above-mentioned method of generating a compensation instruction to cancel out the interference information based on the predicted interference information, and also includes:

[0067] While monitoring the micro-vibration signal of the cutting head in real time, the instantaneous current signal of the servo motor of the cutting head is also monitored.

[0068] The micro-vibration signal and instantaneous current signal are compared with the expected motion data predicted by the digital model to determine the additional transient micro-perturbations caused by the fabric micro-materials.

[0069] Based on the additional transient micro-perturbations, reaction force commands are dynamically generated.

[0070] Specifically, while monitoring the micro-vibration signal of the cutting head in real time, the system also monitors the instantaneous current signal of the servo motor of the cutting head. The micro-vibration signal primarily reflects the mechanical vibration state of the cutting head during its movement, while the instantaneous current signal of the servo motor provides real-time information about the motor load, torque output, and interaction force with the fabric. Transient micro-disturbances caused by the microscopic material properties of the fabric directly affect the contact force between the cutting blade and the fabric, which is then reflected in the instantaneous current changes of the servo motor. By simultaneously acquiring these two signals, the dynamic response of the cutting head under complex working conditions can be captured more comprehensively and precisely.

[0071] Furthermore, the acquired micro-vibration signals and instantaneous current signals are compared with the expected motion data predicted by the digital model of the cutting head motion system. The digital model provides expected data such as the cutting head's trajectory, speed, acceleration, and corresponding motor current under ideal or preset working conditions. By performing difference analysis between the actually monitored signals and these expected data, additional transient micro-perturbations caused by the fabric's micro-material properties can be effectively separated. For example, when the cutting head encounters a region with high local density in the fabric, the instantaneous current of the servo motor may experience a brief peak, and the micro-vibration signal may also undergo changes in specific frequency or amplitude. These deviations from the steady motion data predicted by the digital model are identified as additional transient micro-perturbations caused by the fabric's micro-material properties.

[0072] Therefore, once these additional transient micro-perturbations are identified, the system dynamically generates reaction force commands based on their characteristics. These reaction force commands are designed to counteract the instantaneous effects of the fabric's microstructure on the cutting head's movement by applying a precise counterforce or torque to the cutting head. For example, if a transient leftward resistance is detected from the fabric against the cutter, the system immediately generates a rightward reaction force command, precisely controlling the cutting head via a servo motor to keep it on the predetermined cutting path, thereby correcting potential contour deviations.

[0073] This application's solution, by introducing the monitoring of the instantaneous current signal of the cutting head servo motor and comprehensively comparing it with micro-vibration signals and expected motion data predicted by digital models, can gain a deeper understanding of the dynamic interaction between the cutting head and the fabric. It is precisely because the additional transient micro-perturbations caused by the fabric's microscopic material properties can be accurately identified and quantified that the system can dynamically generate and apply reaction force commands for these specific perturbations. This proactive reaction force compensation mechanism goes beyond simple mechanical vibration suppression, directly acting to counteract the instantaneous influence of the fabric material itself on the cutting process, thereby ensuring that the cutting head can more accurately follow the preset cutting path and effectively avoid contour deviations caused by changes in the local properties of the fabric.

[0074] Through the above technical solution, this application can significantly improve the accuracy and stability of garment cutting. Especially when handling fabrics with complex or non-uniform microscopic material properties, this solution can effectively identify and counteract transient micro-disturbances caused by the fabric itself, thereby avoiding subtle contour deviations that may occur in traditional methods. As a result, the cut garment parts have higher dimensional consistency and edge quality, reducing subsequent finishing work and improving production efficiency and product qualification rate. This refined compensation capability for microscopic material disturbances in the fabric enables the cutting system to exhibit stronger adaptability and robustness when dealing with diverse fabrics.

[0075] In some preferred embodiments, it is assumed that the cutting head is cutting a piece of cotton or linen fabric with localized fiber entanglement or uneven density at high speed. When the cutting blade contacts an area of ​​fiber entanglement in the fabric, the resistance of the fabric to the blade increases instantaneously. This may not only cause slight mechanical vibrations in the cutting head, but more importantly, the instantaneous current of the cutting head's servo motor will rise rapidly in order to maintain the preset feed speed and cutting depth. At this time, the system will simultaneously monitor the micro-vibration signal of the cutting head and the instantaneous current signal of the servo motor.

[0076] Specifically, the system compares these real-time monitored signals with the expected motion data predicted by the digital model of the cutting head motion system during cutting on a uniform fabric. The comparison reveals significant deviations between the actual instantaneous current and micro-vibration signals and the expected data in fiber entanglement areas. These deviations are identified as additional transient micro-perturbations caused by the fabric's microstructure. For example, the system might identify a drag perturbation with an extremely short duration but high amplitude.

[0077] Based on the identified disturbance characteristics, the system dynamically generates a precise reaction force command. This command may manifest as an instantaneous thrust, opposite in direction and matching in magnitude to the fabric resistance, applied to the cutting head by a servo motor within a very short time. For example, if the fabric resistance attempts to push the cutter away from the cutting path to the right, the reaction force command will cause the servo motor to generate an instantaneous force to the left, thereby precisely maintaining the cutter on the predetermined cutting path. In this way, even if the fabric itself has microscopic defects, the cutting head can maintain a high-precision cutting trajectory, ensuring the contour accuracy of the final cut piece.

[0078] In this embodiment, a step is further proposed to embed the above-mentioned compensation instructions into the cutting instruction sequence to generate a second cutting instruction sequence, specifically including: integrating the reaction force instructions, vibration suppression instructions and / or contour geometry correction instructions into the cutting instruction sequence to form a second cutting instruction sequence.

[0079] Specifically, "integration" refers to the systematic fusion and coordination of different types of compensation commands, including vibration suppression commands and / or contour geometry correction commands generated from anticipated interference information, as well as reaction force commands generated from additional transient micro-disturbances caused by the fabric's micro-materials. This integration is not simply a superposition of commands, but rather considers the timing, amplitude matching, and mutual influence of each command to ensure they work synergistically during the cutting head's execution, jointly offsetting or mitigating various interferences. Specifically, the vibration suppression command aims to offset the mechanical vibrations caused by the cutting head switching cutting parameters when executing the first cutting command sequence; the contour geometry correction command corrects the resulting contour deviations; and the reaction force command provides immediate, localized compensation for additional transient micro-disturbances caused by the fabric's micro-materials. Through this integration, a second cutting command sequence containing all necessary compensation information is ultimately formed, which can more comprehensively and accurately guide the movement of the cutting head.

[0080] This application's solution integrates multiple compensation commands, overcoming the limitations of a single compensation command and improving the overall compensation effect. Specifically, when the cutting head performs a cutting task, its movement is disturbed by various factors, including mechanical vibration, contour deviation, and transient micro-disturbances caused by the fabric's microstructure. Vibration suppression commands and contour geometry correction commands primarily compensate for anticipated, macroscopic disturbances, while reaction force commands focus on real-time, microscopic disturbances. By integrating these compensation commands at different levels, the system can simultaneously handle macroscopic and microscopic disturbances within a unified second cutting command sequence. This integration allows the cutting head to not only pre-counteract known vibrations and deviations but also respond in real-time to unpredictable micro-disturbances caused by the fabric material, thereby ensuring the smoothness of the cutting head's movement and the accuracy of the cutting path.

[0081] In some preferred embodiments, the cutting control system first generates a first cutting instruction sequence based on fabric characteristic information and the cutting path. Subsequently, the pre-simulation module pre-simulates the cutting head motion based on a digital model of the cutting head motion system, predicting potential mechanical vibrations and contour deviations caused by cutting parameter switching, and generating corresponding vibration suppression and contour geometry correction instructions. Simultaneously, the system monitors the micro-vibration signals of the cutting head and the instantaneous current signals of the servo motor in real time, comparing them with the expected motion data predicted by the digital model to identify additional transient micro-disturbances caused by the fabric's microstructure, and dynamically generates reaction force instructions accordingly. Finally, the control module integrates these compensation instructions from different sources—vibration suppression, contour geometry correction, and reaction force instructions—for time synchronization and amplitude coordination. For example, at a specific cutting point or segment, if mechanical vibration, contour deviation, and micro-disturbances are predicted to coexist, the system precisely aligns the corresponding vibration suppression, contour correction, and reaction force signals on the time axis, and superimposes or fuses them according to their respective amplitude and phase relationships to form a comprehensive control signal. This comprehensive control signal, as part of the second cutting instruction sequence, is sent to the cutting head to guide it to complete the cutting in a smoother and more precise manner, thereby effectively counteracting various interferences and ensuring cutting quality.

[0082] In this embodiment, the steps of comparing the micro-vibration signal, instantaneous current signal, and expected motion data predicted by the digital model to determine the additional transient micro-perturbations caused by the fabric micromaterials are further proposed, including:

[0083] The micro-vibration signal and the instantaneous current signal are compared with the expected motion data predicted by the digital model to obtain the residual signal;

[0084] The residual signal is decomposed into multiple frequency sub-band signals.

[0085] Within each frequency sub-band signal, feature extraction is performed on the frequency sub-band signal to obtain the vibration amplitude, phase, and frequency center of each frequency sub-band;

[0086] The additional transient micro-perturbations are identified based on the vibration amplitude, phase, and frequency center of each frequency sub-band.

[0087] Specifically, the micro-vibration signal and the instantaneous current signal are compared with the expected motion data predicted by the digital model, respectively. The purpose is to separate the deviation between the actual motion and the ideal motion of the cutting head. The residual signal can be understood as the difference between the actual measured value and the model prediction value, which includes unexpected dynamic responses caused by various factors such as the microstructure of the fabric and the nonlinearity of the mechanical system. By comparing them separately, the disturbance components from different sources can be identified more clearly.

[0088] The purpose of multi-band decomposition of the residual signal is to break down the complex broadband disturbance signal into multiple narrowband signals, facilitating independent analysis of disturbances within different frequency ranges. For example, wavelet transform, Fourier transform, or filter banks can be used to decompose the residual signal into multiple frequency sub-bands, such as low-frequency, mid-frequency, and high-frequency sub-bands. Each frequency sub-band represents the disturbance energy distribution within a specific frequency range. In practical applications, feature extraction is performed on each frequency sub-band signal to quantify the key attributes of the micro-perturbations within each sub-band. Specifically, the vibration amplitude, phase, and frequency center of each frequency sub-band can be extracted. The vibration amplitude reflects the intensity of the disturbance, the phase information reveals the starting point and propagation characteristics of the disturbance, and the frequency center indicates the main frequency components of the disturbance within that sub-band. These features can be extracted using methods such as spectral analysis and envelope analysis.

[0089] Therefore, based on the vibration amplitude, phase, and frequency center of each frequency sub-band, the additional transient micro-perturbations are identified. The purpose is to accurately distinguish and locate specific perturbations caused by the microstructure of the fabric based on quantified feature information. For example, by setting thresholds, pattern recognition algorithms, or machine learning models, signal patterns with specific combinations of frequency, amplitude, and phase can be identified as transient perturbations caused by uneven microstructure of the fabric (such as fiber entanglement or local thickness variations), thereby distinguishing them from the inherent vibration of the system or other external disturbances.

[0090] This application's solution compares the micro-vibration signals generated by the actual movement of the cutting head and the instantaneous current signals of the servo motor with the expected motion data predicted by the digital model. This allows for the initial separation of the deviation between the actual and ideal motion, forming a residual signal containing various disturbance information. Furthermore, this residual signal undergoes multi-band decomposition, breaking down the complex broadband micro-disturbance signal into multiple more easily analyzed frequency sub-band signals, enabling refined processing of disturbances within different frequency ranges. Extracting key features such as vibration amplitude, phase, and frequency center within each frequency sub-band allows for a comprehensive and accurate characterization of the dynamic properties of the micro-disturbance. Therefore, based on this refined feature information, additional transient micro-disturbances caused by the fabric's microstructure can be accurately identified, avoiding potential inaccuracies or omissions in traditional methods. This provides a more reliable basis for subsequently generating precise reaction force commands.

[0091] The above technical solution enables more refined and accurate identification of additional transient micro-perturbations caused by the microstructure of fabric. Compared to methods that only perform simple comparisons, this solution, by introducing multi-band decomposition and feature extraction, can deeply analyze key parameters such as the frequency, phase, and amplitude of micro-perturbations, thereby significantly improving the accuracy and robustness of micro-perturbation identification. This provides a more reliable basis for the subsequent dynamic generation of reaction force commands, ensuring the effectiveness of compensation commands, ultimately helping to improve the precision and quality of garment cutting, reduce material waste, and extend the service life of the cutting head.

[0092] In some preferred embodiments, a specific example is illustrated below. Suppose that when a cutting head is cutting a piece of fabric with locally uneven fiber density, its micro-vibration signals and the instantaneous current signals of the servo motor will show deviations from the expected motion data predicted by the digital model. Specifically, firstly, the real-time acquired micro-vibration signals and instantaneous current signals are compared point-by-point with the ideal motion trajectory data predicted by the digital model, and the instantaneous differences between the two are calculated, thus obtaining a comprehensive residual signal. This residual signal contains information on all unexpected motion and mechanical responses. Subsequently, the residual signal is subjected to multi-band decomposition processing such as wavelet transform or Fourier transform, decomposing it into multiple frequency sub-band signals such as low-frequency, mid-frequency, and high-frequency signals to facilitate the analysis of disturbance components in different frequency ranges. Within each frequency sub-band, its vibration amplitude, dominant frequency, and phase information are further extracted, for example, through peak detection and spectral analysis. Ultimately, based on the detailed features of each extracted frequency subband, the system can accurately identify which additional transient micro-perturbations are caused by the fabric's microstructure (such as fiber entanglement and local thickness changes) and distinguish them from other types of disturbances (such as the inherent vibration of the mechanical system), laying the foundation for the subsequent generation of accurate reaction force commands.

[0093] This embodiment further proposes a scheme for continuously monitoring and adjusting the dynamically generated reaction force commands in order to more accurately suppress vibrations during the cutting process.

[0094] This embodiment proposes a step for dynamically generating a reaction force command based on the aforementioned additional transient micro-perturbation, including: dynamically generating a composite reaction force command based on the waveform characteristics of the additional transient micro-perturbation; after executing a cutting command sequence that applies the composite reaction force command, continuously monitoring the residual vibration of the cutting head and acquiring the residual vibration signal; and adjusting the composite reaction force command based on the residual vibration signal.

[0095] Specifically, dynamically generating composite reaction force commands based on the waveform characteristics of the additional transient micro-perturbations means that after identifying the additional transient micro-perturbations caused by the fabric's micro-materials, the system analyzes the specific waveform characteristics of these perturbations, such as their frequency components, amplitude, phase relationship, and duration. Based on these detailed waveform characteristics, one or more precisely matched reaction force commands can be generated. These commands can be designed to have a phase opposite to the perturbation waveform, a matching frequency, and an amplitude, in order to maximize the cancellation of the perturbation when applied. The "composite reaction force command" can be understood as a comprehensive control command that may contain multiple frequency or amplitude components, designed to cope with complex and variable micro-perturbations.

[0096] Furthermore, after executing the cutting command sequence that applies the combined reaction force, continuously monitoring the residual vibration of the cutting head and acquiring the residual vibration signal means that after the cutting head executes the cutting command sequence containing preliminary compensation instructions, the system does not stop monitoring the motion state of the cutting head. Instead, it uses high-precision sensors (such as accelerometers or laser displacement sensors) to continuously detect the minute vibrations of the cutting head during the actual cutting process. These vibration data are acquired and processed in real time to form a "residual vibration signal," which reflects the vibration components of the cutting head that still exist and have not been completely eliminated under the preliminary compensation. The purpose of continuous monitoring is to capture any new or incompletely suppressed dynamic disturbances that may occur during the cutting process.

[0097] Therefore, adjusting the composite reaction force command based on the residual vibration signal means inputting the acquired residual vibration signal as feedback information into the control system. The system analyzes the characteristics of the residual vibration signal, such as its frequency, amplitude, and phase, and compares it with the expected ideal motion state. Based on this comparison, the system can dynamically correct or optimize the previously generated composite reaction force command. This adjustment can be performed in real time to ensure that the movement of the cutting head remains in an optimal state, thereby further improving cutting accuracy and stability. For example, if the residual vibration signal shows that vibration at a specific frequency still exists, the system can enhance the reaction force component for that frequency or adjust its phase to achieve more effective cancellation.

[0098] The proposed solution establishes a closed-loop feedback control mechanism by continuously monitoring the residual vibration of the cutting head after the initial application of the compound reaction force command, and dynamically adjusting the compound reaction force command based on the acquired residual vibration signal. Specifically, when the cutting head executes a cutting command sequence containing an initial compensation command, if residual vibrations remain that are not completely eliminated, these residual vibrations are captured by sensors and converted into residual vibration signals. These signals are then used to analyze the deviation between the actual and ideal motion of the cutting head. Based on this deviation, the system can accurately identify the deficiencies of the current compensation scheme and accordingly correct the compound reaction force command in real time. This adaptive adjustment mechanism enables the system to cope with various dynamic changes and uncertainties that may occur during the cutting process, such as minor differences in local fabric properties, tool wear, or changes in ambient temperature, ensuring that the cutting head's motion trajectory remains highly accurate, thereby effectively suppressing vibration and correcting contour deviations.

[0099] Through the above technical solution, this application can significantly improve the accuracy and stability of garment cutting. Compared with solutions that rely solely on prediction and one-time compensation, the introduction of residual vibration monitoring and dynamic adjustment mechanisms gives the system adaptive capabilities, enabling it to correct compensation commands in real time and effectively counteract unforeseen or dynamically changing interferences during the cutting process. As a result, the actual movement trajectory of the cutting head can more accurately conform to the preset path, significantly reducing cutting quality problems caused by mechanical vibration and contour deviations. This advantage is particularly pronounced when handling complex fabrics or high-precision cutting tasks. This method of continuously optimizing compensation commands ensures the robustness and reliability of the cutting process, ultimately improving the overall efficiency and finished product quality of garment cutting.

[0100] In some preferred embodiments, a specific example is illustrated below. Assume that during the cutting process, after the cutting head executes the initial compensation command, an accelerometer detects a weak residual vibration at a specific frequency (e.g., 50Hz). The system analyzes this residual vibration signal in real time to determine its amplitude and phase. For example, if the residual vibration signal shows a vibration with an amplitude of 0.1g and a phase of 0 degrees at 50Hz, the control system dynamically generates a reaction force component with an amplitude of 0.1g and a phase of 180 degrees at 50Hz and adds it to the current composite reaction force command. After applying the adjusted composite reaction force command, the system continues to monitor the vibration of the cutting head. If the residual vibration further decreases, the adjustment is effective; if residual vibration persists, the system repeats the monitoring and adjustment process until the residual vibration is reduced below an acceptable threshold. This iterative feedback adjustment ensures that the cutting head remains in an optimal vibration suppression state throughout the entire cutting process.

[0101] In some embodiments described above, this application proposes a scheme to adjust the composite reaction force command based on the residual vibration signal to further suppress the vibration generated by the cutting head when executing the cutting command sequence. However, in practical applications, the characteristics of the residual vibration signal may dynamically change when the cutting head is cutting at high speed or along complex paths; for example, the frequency and amplitude may evolve rapidly over time. If adjustment is made solely based on the instantaneous residual vibration signal, the response of the compensation command may be delayed, making it difficult to effectively cope with rapidly changing vibration modes, thereby affecting cutting accuracy and stability. Therefore, this application further proposes a more refined and forward-looking adjustment strategy, which dynamically adjusts the composite reaction force command by predicting the evolution trend of the residual vibration signal.

[0102] Specifically, the above-mentioned command to adjust the composite reaction force based on the residual vibration signal includes:

[0103] Based on the instantaneous frequency and amplitude variation trend of the residual vibration signal, the evolution trend of the residual vibration within a short time window is predicted; based on the evolution trend, the amplitude and / or phase of the composite reaction force command are dynamically adjusted.

[0104] The "instantaneous frequency" refers to the frequency value of the signal at a given moment, which can be obtained through time-frequency analysis methods such as Short-Time Fourier Transform (STFT), Wavelet Transform, or Hilbert-Huang Transform (HHT). The "amplitude variation trend" indicates the rate and direction of change of the residual vibration signal's amplitude over time, such as whether it gradually increases, decreases, or remains stable. This information collectively describes the dynamic characteristics of the residual vibration signal. Predicting the evolution trend of the residual vibration within a short time window means using historical data and the current instantaneous frequency and amplitude variation trend, through a predictive model (e.g., based on Kalman filtering, Autoregressive Moving Average (ARMA), neural networks, or support vector machines), to estimate how the frequency and amplitude of the residual vibration might change over a future period. This "short time window" is typically set to a length sufficient for effective intervention before significant changes in vibration characteristics occur, such as tens to hundreds of milliseconds. Dynamically adjusting the amplitude and / or phase of the composite reaction force command based on the evolution trend means correcting the amplitude and / or phase of the composite reaction force command in real time based on the predicted frequency and amplitude changes of the residual vibration over a short future period. For example, if it is predicted that the amplitude of the residual vibration will increase, the amplitude of the composite reaction force command can be increased in advance; if it is predicted that the frequency will drift, the phase of the command can be adjusted accordingly to ensure that the reaction force command always maintains the best cancellation relationship with the residual vibration.

[0105] This application's solution effectively solves the response lag problem that may result from adjusting solely based on instantaneous residual vibration signals by introducing a prediction mechanism for the evolution trend of residual vibration signals. Specifically, when the cutting head executes the cutting command sequence, the continuously monitored residual vibration signals are analyzed in real time, extracting their instantaneous frequency and amplitude change trends. These dynamic characteristics are input into the prediction model, enabling the prediction of possible frequency drift or amplitude increases / decreases in residual vibration within the next short time window. This forward-looking predictive capability allows the adjustment of the composite reaction force command to move beyond a passive response to current vibration, enabling proactive and advance correction. For example, if an increase in residual vibration is predicted, the system can pre-increase the amplitude of the reaction force command, ensuring timely compensation when the vibration actually intensifies, thus avoiding instantaneous vibration aggravation due to response delay. Similarly, predictive phase adjustment ensures that the reaction force command and residual vibration maintain optimal anti-phase cancellation, thereby maximizing vibration suppression.

[0106] In some preferred embodiments, this application is implemented as follows: First, after continuously monitoring the residual vibration signal of the cutting head, the Hilbert-Huang Transform (HHT) can be used to perform time-frequency analysis on the signal to extract its instantaneous frequency and instantaneous amplitude in real time. By continuously sampling and analyzing these instantaneous parameters, the rate of change of instantaneous frequency and the rate of change of amplitude can be calculated, thereby obtaining the instantaneous frequency and amplitude change trends of the residual vibration signal. Second, these instantaneous frequency and amplitude change trends are used as inputs and fed into a pre-trained prediction model. This prediction model can be a neural network model based on a Long Short-Term Memory (LSTM) network, which can predict the frequency and amplitude of residual vibration within a short time window of 50 to 100 milliseconds by learning historical vibration data and corresponding evolution trends. For example, the model may predict that in the next 80 milliseconds, the center frequency of residual vibration will drift from 50 Hz to 52 Hz, and the amplitude will increase by 10%. Finally, based on the predicted evolution trend, the compound reaction force command is dynamically adjusted. Specifically, if a frequency drift to 52Hz is predicted, the frequency of the composite reaction force command will be pre-adjusted to 52Hz; if an amplitude increase of 10% is predicted, the amplitude of the composite reaction force command will also increase by 10% accordingly. This proactive adjustment ensures that the compensation command can precisely match when the residual vibration actually changes, thereby achieving continuous and efficient suppression of vibration. In this way, even under rapidly changing cutting conditions, the cutting head can maintain extremely high stability, ensuring cutting quality.

[0107] This application also discloses a garment cutting control system, such as... Figure 2 As shown, the system includes:

[0108] The acquisition module 201 is used to acquire fabric feature information and determine cutting parameters based on the fabric feature information; the fabric feature information includes fabric thickness and / or density; the cutting parameters include tool vibration frequency, cutting depth and forward speed;

[0109] The first generation module 202 is used to obtain the cutting path and determine a set of cutting parameters corresponding to the cutting path to generate a first cutting instruction sequence;

[0110] The pre-simulation module 203 is used to pre-simulate the movement of the cutting head on the cutting path based on the digital model of the cutting head motion system, so as to predict the interference information generated by the cutting head when executing the cutting instruction sequence; the interference information includes mechanical vibration and / or contour deviation caused by switching the cutting parameters when executing the first cutting instruction sequence.

[0111] The second generation module 204 is used to generate a compensation instruction to counteract the interference information based on the predicted interference information; the compensation instruction includes a vibration suppression instruction to suppress mechanical vibration and / or a contour geometry correction instruction to correct the contour deviation;

[0112] Control module 205 is used to embed the compensation instruction into the cutting instruction sequence to generate a second cutting instruction sequence, and to control the cutting head to execute the second cutting instruction sequence.

[0113] Therefore, the system of this application can not only effectively adapt to the unevenness of the fabric and ensure the quality of the cut edges, but more importantly, it significantly improves the geometric accuracy of the cut contour by actively predicting and canceling interference. This integrated predictive-compensation control system enables garment cutting to achieve both high precision and high quality while operating at high speed, breaking through the dilemma between efficiency and accuracy in existing technologies and realizing a major advancement in automated garment cutting technology.

[0114] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A garment cutting control method characterized by, The method comprises: acquiring fabric characteristic information and determining cutting parameters according to the fabric characteristic information; the fabric characteristic information includes fabric thickness and / or density; the cutting parameters include tool vibration frequency, cutting depth and forward speed; acquiring a cutting path and determining a corresponding set of cutting parameters according to the cutting path to generate a first cutting instruction sequence; based on a digital model of a cutting head motion system, pre-playing the motion of the cutting head on the cutting path to predict interference information generated by the cutting head when executing the cutting instruction sequence; the interference information includes mechanical vibration and / or profile deviation generated when switching the cutting parameters when executing the first cutting instruction sequence; according to the predicted interference information, generating compensation instructions for offsetting the interference information; the compensation instructions include vibration suppression instructions for suppressing mechanical vibration and / or profile geometry correction instructions for correcting the profile deviation; embedding the compensation instructions into the cutting instruction sequence to generate a second cutting instruction sequence, and controlling the cutting head to execute the second cutting instruction sequence.

2. The garment cutting control method according to claim 1, characterized by, The acquisition of fabric characteristic information comprises: non-contact scanning of the fabric to acquire three-dimensional topographic data of the fabric surface and generate a fabric characteristic map; determining the three-dimensional height and / or distance information of each point on the fabric surface according to the three-dimensional topographic data; converting the three-dimensional height and / or distance information into quantized values representing fabric characteristic information.

3. The garment cutting control method according to claim 2, wherein, The acquisition of the cutting path and the determination of a corresponding set of cutting parameters according to the cutting path to generate a first cutting instruction sequence comprises: aligning the fabric characteristic map with the garment pattern layout in the coordinate system; based on the fabric characteristic map, querying the fabric local physical characteristic information corresponding to each point of the cutting path; the fabric local physical characteristic information includes local area fabric thickness and / or density; determining the cutting parameters of each point according to the fabric local physical characteristic information corresponding to each point and the preset cutting parameter lookup table to generate a first cutting instruction sequence.

4. The garment cutting control method according to claim 1, characterized by, The digital model of the cutting head motion system is a modeling of the components of the cutting head through finite element analysis software; the digital model outputs the response of the cutting head under different driving forces; According to the predicted interference information, generating compensation instructions for offsetting the interference information, comprises: real-time monitoring of the micro-vibration signal of the cutting head and real-time frequency spectrum analysis of the micro-vibration signal to determine the frequency, phase and amplitude of the micro-vibration signal; based on the micro-vibration signal analysis result, dynamically generating an anti-vibration waveform opposite in phase, same in frequency and matched in amplitude to the detected vibration waveform; superimposing the anti-vibration waveform into the cutting instruction sequence of the cutting head.

5. The garment cutting control method according to claim 4, wherein According to the predicted interference information, generating compensation instructions for offsetting the interference information, further comprises: real-time monitoring of the micro-vibration signal of the cutting head while monitoring the instantaneous current signal of the servo motor of the cutting head; comparing the micro-vibration signal, the instantaneous current signal and the expected motion data predicted by the digital model to determine the additional transient micro-disturbance caused by the micro-material of the fabric; According to the additional transient micro-disturbance, a counterforce instruction is dynamically generated.

6. The garment cutting control method according to claim 5, wherein The embedding of the compensation instruction into the cutting instruction sequence to generate the second cutting instruction sequence further includes integrating the counterforce instruction, and a vibration suppression instruction and / or a contour geometry correction instruction into the cutting instruction sequence to form the second cutting instruction sequence.

7. The garment cutting control method according to claim 5, wherein The comparison of the micro-vibration signal and the transient current signal with the expected motion data predicted by the digital model to determine the additional transient micro-disturbance caused by the micro-material of the cloth includes: The comparison of the micro-vibration signal and the transient current signal with the expected motion data predicted by the digital model to obtain a residual signal; The multi-band decomposition of the residual signal to obtain a plurality of frequency sub-band signals; The feature extraction of the frequency sub-band signal in each of the frequency sub-band signals to obtain the vibration amplitude, phase and frequency center of each frequency sub-band; According to the vibration amplitude, phase and frequency center of each frequency sub-band, the additional transient micro-disturbance is identified.

8. The garment cutting control method according to claim 7, wherein, The dynamic generation of the counterforce instruction according to the additional transient micro-disturbance includes: According to the waveform characteristics of the additional transient micro-disturbance, a composite counterforce instruction is dynamically generated; After executing the cutting instruction sequence for applying the composite counterforce instruction, the residual vibration of the cutting head is continuously monitored and a residual vibration signal is obtained; The adjustment of the composite counterforce instruction according to the residual vibration signal includes:

9. The garment cutting control method according to claim 8, wherein, According to the instantaneous frequency and amplitude change trend of the residual vibration signal, the evolution trend of the residual vibration in a short time window is predicted; According to the evolution trend, the amplitude and / or phase of the composite counterforce instruction are dynamically adjusted. The system includes:

10. A garment cutting control system characterized by, The acquisition module is configured to acquire cloth characteristic information and determine cutting parameters according to the cloth characteristic information; the cloth characteristic information includes cloth thickness and / or density; and the cutting parameters include tool vibration frequency, cutting depth and forward speed. The first generation module is configured to acquire a cutting path and determine a corresponding set of cutting parameters according to the cutting path to generate a first cutting instruction sequence. The pre-performance module is configured to pre-perform the motion of the cutting head on the cutting path based on a digital model of the cutting head motion system to pre-judge interference information generated by the cutting head when executing the cutting instruction sequence; the interference information includes mechanical vibration and / or contour deviation generated when the cutting parameters are switched when executing the first cutting instruction sequence. The second generation module is configured to generate compensation instructions for offsetting the interference information according to the pre-judged interference information; the compensation instructions include a vibration suppression instruction for suppressing mechanical vibration and / or a contour geometry correction instruction for correcting the contour deviation. The control module is configured to embed the compensation instructions into the cutting instruction sequence to generate a second cutting instruction sequence and control the cutting head to execute the second cutting instruction sequence. ​

Images