Stitch control method of sewing machine

Abstract

The invention relates to the technical field of stitch control, in particular to a stitch control method of sewing machinery, which comprises the steps of coordinate system establishment and error modeling, discretization control quantity correction, error fusion and vibration detection, and optimized stitch output. The invention aims to improve the flexibility and accuracy of control stitches of sewing equipment and improve the quality and pattern diversity of sewn products. A double-system frame of a sewing machine transmission structure coordinate system and a cloth coordinate system is established, and the space positioning requirement of precise sewing equipment is met. Through real-time error calculation of an actual position and a theoretical position (such as through an encoder or visual feedback), a mathematical basis is provided for subsequent closed-loop control, and general design logic of a modern electromechanical system is met. By accurately establishing the conversion relation between the coordinate systems and calculating the error, the sewing process can be better controlled and adjusted, and the sewing precision and quality are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of stitch control, and in particular to a stitch control method for a sewing machine. Background Art

[0002] At present, as manual labor in industries such as clothing production and handbag making is gradually being automated, sewing machines, as the main sewing equipment, are being used more and more widely.

[0003] In order to adapt to the increasing demand for personalized customization, existing sewing machines need to have more flexible and precise stitch control methods to facilitate sewing operations with complex patterns and delicate craftsmanship.

[0004] While some sewing machines can achieve multi-directional swinging, they still rely on transmission components within the mechanical structure. This results in inaccurate and inconvenient control of the swing amplitude and stitch length. Traditional control methods struggle to meet the demands of high stitch lengths and high precision, especially when it comes to sewing machines.

[0005] In view of this, the present invention provides a stitch control method for a sewing machine. Summary of the Invention

[0006] The object of the present invention is to provide a stitch control method for a sewing machine in view of the deficiencies in the prior art.

[0007] In order to solve the above technical problems, the following technical solutions are adopted: A stitch control method for a sewing machine comprises the following steps: S1. Design the sewing transmission structure coordinate system and the cloth pattern coordinate system of the sewing work process, further establish the mapping transformation matrix of the sewing transmission structure coordinate system to the cloth pattern coordinate system, and calculate the system error of the mapping transformation; S2. Discrete the pattern stitch of the fabric into several stitching points, calculate the stitch offset of each stitching point, correct the current output torque of the servo controller of the sewing drive structure, and determine the theoretical error; S3. After adjusting the fabric pattern stitches based on the system error and theoretical error, conduct a test on the needle swing, needle placement, and vibration information of the sewing drive structure to calculate and adjust the overall error Δδ; S4. Iteratively optimize the pattern trace until the total error Δδ satisfies the threshold Δδ≤δ h , until the error requirements are met, and the adjusted pattern stitches are obtained for real-time fabric sewing processing to complete error control.

[0008] A further improvement based on the above technical solution is that the sewing transmission structure coordinate system is a coordinate system established based on the transmission structure of the sewing machine itself. The sewing transmission structure coordinate system takes the zero point of the encoder of the sewing transmission structure as the origin, and determines the direction of the coordinate axis according to the movement direction of the sewing transmission structure, which is used to describe the position of the sewing machine transmission components.

[0009] A further improvement based on the above technical solution is that the cloth pattern coordinate system is a coordinate system used by pattern drawing software, and the cloth pattern coordinate system is based on cloth and is used to determine the position of each point on the cloth so as to control the relative position of the needle and the cloth during the sewing process; Designing the cloth pattern coordinate system also includes acquiring a selected pattern to obtain a pattern stitch to be processed; and establishing a cloth pattern coordinate system based on the pattern stitch and the position of the pattern stitch on the cloth.

[0010] A further improvement based on the above technical solution is that the mathematical formula of the mapping transformation matrix is: , in is the mapping transformation matrix from the sewing transmission structure coordinate system to the cloth pattern coordinate system; R is the rotation transformation matrix, , exp is the exponential mapping, is the rotation angle, is the antisymmetric matrix with respect to the axis of rotation; p is the displacement vector, , which represents the translation relationship between the origin of the sewing transmission structure coordinate system and the origin of the cloth pattern coordinate system.

[0011] A further improvement based on the above technical solution is that the calculation of the system error includes: The position sensor installed on the sewing machine collects the sewing needle position coordinate value P of the sewing transmission structure in real time. actual At the same time, the theoretical position coordinate value P is obtained according to the sewing design requirements theory , calculate the system error δ1; δ1=P actual -P theory = (J·δq)+ εs Where J is the Jacobian matrix, δq is the joint space error, and εs is the sensor noise.

[0012] A further improvement based on the above technical solution is that S2 includes: S2-1. Feature extraction is performed on the pattern stitch to be processed to obtain stitch needle features, the stitch needle features are discretized into several stitch points, the stitch points are discretized and modeled to obtain parameterized pattern stitches: C(u)=[x(u),y(u)] Where C(u) is the curve function of the pattern stitch, which describes the stitch as a continuous mapping of the parameter u; x(u) and y(u) are the coordinate components of the curve in the cloth pattern coordinate system; u is a normalization parameter, and its range is limited to the interval [0,1]; S2-2. Obtain discrete point sets through curve curvature adaptive sampling: in, is the parameter value corresponding to the kth discrete point, is the parameter value corresponding to the k-1th discrete point; κ(u) is the curvature of the curve at parameter u, Δ max is the maximum step size threshold.

[0013] A further improvement based on the above technical solution is that S2 includes: S2-3. Calculate the moving distance of the swing and the needle distance for any needle point. According to the deviation of the swing and needle pitch, the stitch offset is formed. The swing - needle pitch dynamic correction is performed by the current output torque of the servo controller of the sewing transmission structure to determine the theoretical error δ2. A further improvement based on the above technical solution is that the specific process of dynamic correction of swing amplitude and needle distance is as follows: (a) Coordinate transformation error compensation: Let the current needle insertion point be P i =(x i ,y i ), the error correction process performs coordinate transformation error compensation: Where δx, δy are the trace offsets, J -1 is the inverse matrix of the Jacobian matrix, and εs is the sensor noise.

[0014] (b) Servo torque correction formula: used for the coupling control quantity of the rotary axis and the linear axis: where K p ,K d , are proportional control coefficient and differential control coefficient respectively, K f is the vibration coefficient, F vibrationis the vibration disturbance force measured by the laser vibrometer.

[0015] A further improvement based on the above technical solution is that the error fusion function Δδ of S3 is as follows: Where: δ1 is the system error; A j is the amplitude of the pendulum angular velocity spectrum, is the amplitude of the reference pendulum angular velocity spectrum; σ vib is the RMS value of vibration acceleration; α, β, and γ are weight coefficients, satisfying α+β+γ=1.

[0016] A further improvement based on the above technical solution is that the iterative optimization pattern stitch of S4 includes: S4-1. Optimize and minimize the objective function using the improved NSGA algorithm; F min =(Δδ,T execute ) Among them, T execute is the execution time of a single iteration, The constraints are as follows: g1:Δδ≤δ h , g2:τ max ≤τ limit , τ max is the maximum value of the servo torque, τ limit It is the servo torque limit value.

[0017] The above technical solution has the following beneficial effects: The present invention proposes a stitch control method for a sewing machine, aiming to improve the flexibility and accuracy of stitch control of the sewing equipment, and enhance the quality and pattern diversity of sewing products.

[0018] Multi-coordinate system collaborative control: This dual-system framework, combining the sewing machine's transmission coordinate system and the fabric coordinate system, meets the spatial positioning requirements of precision sewing equipment. Real-time error calculation between actual and theoretical positions (e.g., through encoders or visual feedback) provides a mathematical foundation for subsequent closed-loop control, consistent with the common design logic of modern mechatronic systems. Accurately establishing the transformation relationship between coordinate systems and calculating errors enables better control and adjustment of the sewing process, improving sewing precision and quality.

[0019] The scientific nature of the dynamic compensation mechanism: The proposed method of correcting the servo controller torque based on stitch offsets in step S2 embodies the concept of dynamic compensation. The servo system's closed-loop torque regulation effectively suppresses offsets caused by fabric deformation, mechanical transmission errors, and other factors. Its principle is similar to the feedforward-feedback compound control used in robot trajectory tracking. The integrated adjustment mechanism in step S3, combining system errors with theoretical errors, can adapt to uncertain interference during the machining process and is theoretically superior to single-source error compensation strategies.

[0020] The fusion of vibration information during the iterative optimization process in steps S3 and S4 is a key feature of this solution, surpassing the traditional optimization model based solely on position error. By incorporating vibration parameters (such as piezoelectric sensor or accelerometer data), a more comprehensive assessment of trace quality is possible, theoretically suppressing the accumulation of hidden errors caused by mechanical resonance.

[0021] The overall technical effects are as follows.

[0022] 1. Fully closed-loop control architecture: The closed-loop chain from error acquisition, dynamic correction to verification and optimization is complete, which is conducive to adaptive processing.

[0023] 2. Discretization compatibility: The pattern stitches of the fabric are discretized into several needle points. The discretization accuracy can be determined according to the complexity of the sewing stitches and the actual sewing requirements. For example, for simple straight stitches, the discrete interval can be larger; for complex curved stitches, the discrete interval needs to be smaller to ensure the accuracy of the stitches. It is compatible with a variety of complex patterns, improves the control accuracy of the fabric stitches during the sewing process, reduces stitch deviation, and improves sewing quality.

[0024] 3. Robustness potential: The process of achieving Δθ ≤ θh through iterative optimization implies the ability of system parameters to self-tune, and theoretically can cope with disturbances caused by different fabric materials. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The present invention will be further described below in conjunction with the accompanying drawings: Figure 1 It is an overall flow chart of the sewing stitch control method according to an embodiment of the present invention.

[0026] Figure 2 It is a schematic diagram of a mapping model between a sewing transmission coordinate system and a cloth pattern coordinate system according to an embodiment of the present invention.

[0027] Figure 3 This is a servo control logic block diagram for dynamic correction of swing amplitude and needle pitch according to an embodiment of the present invention.

[0028] Figure 4 It is a schematic diagram of the error fusion and optimization strategy of an embodiment of the present invention.

[0029] Figure 5Schematic diagram of curvature adaptive sampling of discrete needle insertion points according to an embodiment of the present invention. DETAILED DESCRIPTION

[0030] To make the objectives, technical solutions, and advantages of the present invention more clearly understood, the present invention is described in detail below with reference to the accompanying drawings and examples, which further improve upon the above-described technical solutions. However, it should be understood that the specific embodiments described herein are intended only to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessary confusion regarding the concepts of the present invention.

[0031] See Figure 1-5 A stitch control method for a sewing machine according to this embodiment includes the following steps: S1. Design the sewing machine transmission structure coordinate system and the fabric pattern coordinate system for the sewing process. The sewing machine transmission structure coordinate system is the machine coordinate system, denoted as the M system, which describes the position and motion of the sewing machine transmission structure. The fabric coordinate system is the workpiece coordinate system, denoted as the W system. This coordinate system, established with the fabric as the object, is used to determine the position of each point on the fabric, thereby accurately controlling the relative position of the needle and fabric during the sewing process.

[0032] Furthermore, a mapping transformation matrix from the sewing transmission structure coordinate system to the cloth pattern coordinate system is established, and a system error of the mapping transformation is calculated.

[0033] In this embodiment, the sewing transmission structure coordinate system is a coordinate system established based on the transmission structure of the sewing machine itself. Specifically, the sewing transmission structure coordinate system takes the zero point of the encoder of the sewing transmission structure as the origin, and determines the direction of the coordinate axis according to the movement direction of the sewing transmission structure, which is used to describe the position of the sewing machine transmission components.

[0034] In this embodiment, the cloth pattern coordinate system is a coordinate system used by pattern drawing software. The cloth pattern coordinate system takes cloth as the object and is used to determine the position of each point on the cloth so as to control the relative position of the needle and the cloth during the sewing process.

[0035] Designing the cloth pattern coordinate system also includes acquiring a selected pattern to obtain a pattern stitch to be processed; and establishing a cloth pattern coordinate system based on the pattern stitch and the position of the pattern stitch on the cloth.

[0036] In this embodiment, the mathematical formula of the mapping transformation matrix is: .

[0037] in is the mapping transformation matrix from the sewing transmission structure coordinate system to the cloth pattern coordinate system; R is the rotation transformation matrix, , exp is the exponential mapping, is the rotation angle, It is an antisymmetric matrix related to the rotation axis; this matrix is used to describe the transformation relationship of rotation in space, through which the direction information in the machine coordinate system can be converted to the cloth coordinate system.

[0038] p is the displacement vector, , which represents the translation relationship between the origin of the sewing transmission structure coordinate system and the origin of the cloth pattern coordinate system. That is, this vector can be used to translate the position information in the machine coordinate system to the fabric coordinate system.

[0039] In this embodiment, the calculation of the system error includes: The position sensor installed on the sewing machine collects the sewing needle position coordinate value P of the sewing transmission structure in real time. actual At the same time, the theoretical position coordinate value P is obtained according to the sewing design requirements theory , calculate the system error δ1; δ1=P actual -P theory = (J·δq)+ εs .

[0040] Where δ1 is the actual position P actual With the theoretical position P theory The error between .

[0041] J is the Jacobian matrix, which establishes the linear relationship between the joint space of the sewing transmission structure and the workpiece task space. Through it, the error δq in the joint space can be converted to the workpiece task space.

[0042] δq is the joint space error, which refers to the deviation between the actual position and theoretical position of each transmission shaft, upper needle and lower needle in the sewing transmission structure.

[0043] εs is the sensor noise. Since the sensor will inevitably generate noise during the measurement process, this noise will also affect the position error.

[0044] The complete definition of the system error δ1 is: in the sewing transmission structure coordinate system, the actual position coordinate P of the sewing needle measured in real time by the installed position sensor, visual feedback or other measuring equipment actual The position coordinate P required by theoretical design theory The difference is the system error value corrected by the linear combination of the Jacobian matrix J, the joint space error δq and the sensor noise εs.

[0045] Dimensional meaning: Assuming that the sewing machine has a three-axis motion mechanism, δl∈R 3Represents the position deviation vector in three-dimensional space. Specifically, a sewing machine with a general three-axis motion mechanism includes an upper shaft, a lower shaft, and a feed shaft. The upper shaft is the active shaft, and the lower shaft is driven by a transmission device such as a synchronous belt, gears, or a connecting rod. For example, in an industrial lockstitch sewing machine, the rotation of the upper shaft is connected to the synchronous pulley on the lower shaft through a synchronous pulley and a synchronous belt, driving the rotation of the lower shaft. The upper shaft is connected to the upper head of the sewing machine, driving the needle to move up and down, etc. The lower shaft is connected to the lower head of the sewing machine, driving the rotary shuttle and other components to rotate. The feed shaft is responsible for controlling the feed of the fabric, and works in conjunction with the upper and lower shafts to enable the sewing machine to accurately transport the fabric forward according to the set stitch length and speed during the sewing process to complete the sewing operation.

[0046] This parameter quantifies the coupling effect of mechanical transmission error (including gear clearance, connecting rod deformation, etc.) and sensor noise on trace accuracy, and is the core input quantity for subsequent error fusion control.

[0047] Establishing a dual-system framework for the sewing drive structure coordinate system and the fabric coordinate system meets the spatial positioning requirements of precision sewing equipment. Real-time error calculation between actual and theoretical positions (e.g., through encoder or visual feedback) provides the mathematical foundation for subsequent closed-loop control. This dual-coordinate mapping model and error calculation method are crucial for stitch control in sewing machines. By accurately establishing the transformation relationship between the coordinate systems and calculating the error, the sewing process can be better controlled and adjusted, improving sewing precision and quality.

[0048] S2. Discrete the pattern stitch of the fabric into several needle insertion points, calculate the stitch offset of each needle insertion point, correct the current output torque of the servo controller of the sewing transmission structure, and determine the theoretical error.

[0049] In this embodiment, S2 specifically includes: S2-1. Extract the features of the pattern stitch to be processed to obtain stitch stitch features, discretize the stitch stitch features into a number of stitch points, and model the discrete stitch points to obtain a parameterized pattern stitch. Specifically, for any pattern stitch, it can be parameterized as: C(u)=[x(u),y(u)] Where C(u) is the curve function of the pattern stitch, which describes the stitch as a continuous mapping of the parameter u.

[0050] x(u) and y(u) are the coordinate components of the curve in the cloth pattern coordinate system.

[0051] u is a normalization parameter, and its range is limited to the interval [0,1].

[0052] This parametric representation method provides a basis for the subsequent discretization of pattern stitches, and it can concisely describe the shape and position information of pattern stitches in mathematical form.

[0053] S2-2. Obtain discrete point sets through curve curvature adaptive sampling: in, is the parameter value corresponding to the kth discrete point, is the parameter value corresponding to the k-1th discrete point.

[0054] κ(u) is the curvature of the curve at parameter u, indicating the magnitude of the curvature of the curve at parameter u. The curvature reflects the degree of curvature of the curve. The greater the curvature, the more curved the curve; the smaller the curvature, the flatter the curve.

[0055] Δ max is the maximum step threshold, which limits the parameter u at each sampling k The maximum change.

[0056] According to the curve Adjust the curvature at the

[0057] When the curvature of the curve When the value of the denominator increases (i.e. the curve is more curved), the value of the denominator will increase, making - A smaller value means that more dense sampling will be performed where the curve is curved to better capture the shape of the curve; when the curvature of the curve is When is small, the value of the denominator is relatively small. - The value of will be larger, that is, the sampling points are relatively sparse where the curve is flat.

[0058] This discrete modeling method for needle insertion points, through parameterization of pattern stitches and adaptive sampling based on curve curvature, can reasonably distribute discrete points according to the shape characteristics of the stitches, providing more accurate basic data for subsequent sewing control, such as precise control of the needle insertion position, and helping to improve sewing precision and quality.

[0059] S2-3. For any needle point, calculate the moving distance of the swing amplitude and the moving distance of the needle pitch, form the stitch offset based on the deviation of the swing amplitude and needle pitch, and perform dynamic correction of the swing amplitude and needle pitch through the current output torque of the servo controller of the sewing transmission structure to determine the theoretical error δ2.

[0060] The theoretical error δ2 is generated by the following process: 1. Discretize the pattern stitches into needle points and calculate the swing deviation ΔS and stitch length deviation ΔD of each point 2. Inversely calculate the error using the servo torque correction formula: δ2 = Components: 1. Geometric offset: The deviation of stitches caused by nonlinear deformation of the fabric.

[0061] 2. Control residual: The residual error that is not fully compensated by the dynamic response of the servo system.

[0062] The difference between the systematic error δ1 and the theoretical error δ2 is shown in Table 1 below.

[0063] Table 1 In this embodiment, the specific process of performing the dynamic correction of the swing amplitude and needle distance is as follows: (a) Coordinate transformation error compensation: Let the current needle insertion point be P i =(x i ,y i ), the error correction process performs coordinate transformation error compensation: Where δx and δy are the trace offsets, that is, the deviations of the actual trace from the ideal trace in the x and y directions.

[0064] J is the Jacobian matrix mentioned above, which establishes the linear relationship between the joint space of the sewing transmission structure and the workpiece task space. -1 It is the inverse matrix of the Jacobian matrix, which can be used to convert the trace offset into the amounts ΔX and ΔY that need to be compensated in the coordinate space.

[0065] εs is the sensor noise. The sensor noise is subtracted in the calculation to more accurately obtain the coordinate change that needs to be compensated due to the trace offset, thereby compensating for the coordinate transformation error and improving the accuracy of the needle insertion point position.

[0066] (b) Servo torque correction formula: used for the coupling control quantity of the rotary axis and the linear axis: where K p , K d , are the proportional control coefficient and differential control coefficient respectively, which are similar to the parameters in PID control. A proportional control effect is directly generated according to the size of the coordinate compensation amount ΔX. The differential control effect is generated according to the rate of change of ΔX. The combination of the two can quickly respond and stably adjust the servo torque to cope with coordinate changes.

[0067] K fis the vibration coefficient, F vibration F is the vibration disturbance force measured by the laser vibrometer. vibration It reflects the mechanical vibration during the sewing process. This item incorporates the vibration factor into the correction of the servo torque. Because vibration affects the accuracy and quality of sewing, by adjusting the servo torque considering the vibration amount, the error caused by vibration can be better compensated, and more precise coupling control of the rotary axis and the linear axis can be achieved.

[0068] This dynamic correction method for swing amplitude and stitch length can effectively reduce system errors and improve the accuracy of swing amplitude and stitch length control during the sewing process by compensating for coordinate transformation errors and correcting servo torque based on multiple factors, thereby improving the quality of sewing products.

[0069] S3. After adjusting the pattern stitch of the fabric based on the system error and theoretical error, conduct a test on the needle swing, needle insertion, and vibration information of the sewing transmission structure to calculate and adjust the overall error Δδ.

[0070] In this embodiment, the error fusion function Δδ of S3 is as follows: Where: δ1 is the system error.

[0071] A j is the angular velocity spectrum amplitude of the pendulum needle, that is, the angular velocity amplitude of the pendulum needle at different frequencies, is the reference pendulum angular velocity spectrum amplitude, The difference between the amplitude of the pendulum angular velocity spectrum and the ideal value is measured.

[0072] σ vib It is the RMS value (Root-Mean-Square) of vibration acceleration, which is a statistic that measures the magnitude of vibration acceleration and reflects the intensity of vibration.

[0073] α, β, and γ are weight coefficients, satisfying α + β + γ = 1. These weight coefficients are used to adjust the importance of different error factors in the overall error and can be set according to actual conditions. For example, if the accuracy of the needle angular velocity has a significant impact on sewing quality, the value of β can be appropriately increased.

[0074] This vibration-error coupling model quantifies the impact of vibration on sewing errors by analyzing and processing vibration signals, and feeds the compensation amount back to the servo control loop, thereby achieving effective control of vibration errors during the sewing process.

[0075] S4. Iteratively optimize the pattern trace until the total error Δδ satisfies the threshold Δδ≤δ h, until the error requirements are met, and the adjusted pattern stitches are obtained for real-time fabric sewing processing to complete error control.

[0076] A further improvement based on the above technical solution is that the iterative optimization pattern stitch of S4 includes: S4-1. Optimize and minimize the objective function F using the improved NSGA algorithm (non-dominated sorting genetic algorithm) min ,Through this multi-objective optimization approach, we aim to optimize both the error and the execution time in the ,sewing process simultaneously and find the optimal balance between them.

[0077] F min =(Δδ,T execute ) Among them, T execute is the execution time of a single iteration, The constraints are as follows: g1:Δδ≤δ h , to ensure that the sewing quality meets certain standards.

[0078] g2:τ max ≤τ limit , τ max is the maximum value of the servo torque, τ limit is the servo torque limit. This constraint ensures that during the optimization process, the servo torque will not exceed its safe and effective operating range, avoiding damage to the sewing equipment or affecting the sewing effect.

[0079] Through the evolutionary optimization strategy, an intelligent optimization framework was constructed, which can dynamically optimize various parameters in the sewing process while considering multiple constraints to improve sewing quality and efficiency.

[0080] The technical effect verification data is shown in Table 2.

[0081] The above are only specific embodiments of the present invention, but the technical features of the present invention are not limited thereto. Any simple changes, equivalent substitutions, or modifications based on the present invention to solve substantially the same technical problems and achieve substantially the same technical effects are included within the scope of protection of the present invention.

Claims

1. A stitch control method for a sewing machine, characterized in that The following steps are involved: S1. Design the sewing transmission structure coordinate system and the cloth pattern coordinate system of the sewing work process, further establish the mapping transformation matrix of the sewing transmission structure coordinate system to the cloth pattern coordinate system, and calculate the system error of the mapping transformation; S2. Discretize the fabric pattern stitch into a number of stitching points, calculate the stitch offset for each stitching point, and correct the current output torque of the servo controller of the sewing drive mechanism to determine the theoretical error; S3. After adjusting the fabric pattern stitches based on the system error and theoretical error, conduct a test on the needle swing, needle placement, and vibration information of the sewing drive structure to calculate and adjust the overall error Δδ; S4. Iteratively optimize the pattern trace until the total error Δδ satisfies the threshold Δδ≤δ h , until the error requirements are met, and the adjusted pattern stitches are obtained for real-time fabric sewing processing to complete error control.

2. The stitch control method for a sewing machine according to claim 1, characterized in that: The sewing transmission structure coordinate system is a coordinate system established based on the transmission structure of the sewing machine itself. The sewing transmission structure coordinate system takes the zero point of the encoder of the sewing transmission structure as the origin, determines the direction of the coordinate axis according to the movement direction of the sewing transmission structure, and is used to describe the position of the sewing machine transmission components.

3. The stitch control method for a sewing machine according to claim 1, wherein: The cloth pattern coordinate system is the coordinate system used by the pattern drawing software. The cloth pattern coordinate system takes the cloth as the object and is used to determine the position of each point on the cloth so as to control the relative position of the needle and the cloth during the sewing process; Designing the cloth pattern coordinate system also includes acquiring a selected pattern to obtain a pattern stitch to be processed; and establishing a cloth pattern coordinate system based on the pattern stitch and the position of the pattern stitch on the cloth.

4. The stitch control method for a sewing machine according to claim 1, wherein: The mathematical formula of the mapping transformation matrix is: ， in is the mapping transformation matrix from the sewing transmission structure coordinate system to the cloth pattern coordinate system; R is the rotation transformation matrix, , exp is the exponential mapping, is the rotation angle, is the antisymmetric matrix with respect to the axis of rotation; p is the displacement vector, , which represents the translation relationship between the origin of the sewing transmission structure coordinate system and the origin of the cloth pattern coordinate system.

5. The stitch control method for a sewing machine according to claim 1, wherein: The calculation of the systematic error includes: The position sensor installed on the sewing machine collects the sewing needle position coordinate value P of the sewing transmission structure in real time. actual At the same time, the theoretical position coordinate value P is obtained according to the sewing design requirements theory , calculate the system error δ1; δ1=P actual -P theory = (J·δq)+ εs; Where J is the Jacobian matrix, δq is the joint space error, and εs is the sensor noise.

6. The stitch control method for a sewing machine according to claim 1, characterized in that: The S2 includes: S2-1. Feature extraction is performed on the pattern stitch to be processed to obtain stitch needle features, the stitch needle features are discretized into several stitch points, the stitch points are discretized and modeled to obtain parameterized pattern stitches: C(u)=[x(u),y(u)] Where C(u) is the curve function of the pattern stitch, which describes the stitch as a continuous mapping of the parameter u; x(u) and y(u) are the coordinate components of the curve in the cloth pattern coordinate system; u is a normalization parameter, and its range is limited to the interval [0,1]; S2-2. Obtain discrete point sets through curve curvature adaptive sampling: ； in, is the parameter value corresponding to the kth discrete point, is the parameter value corresponding to the k-1th discrete point; κ(u) is the curvature of the curve at parameter u, Δ max is the maximum step size threshold.

7. The stitch control method for a sewing machine according to claim 6, characterized in that: The S2 includes: S2-3. For any needle point, calculate the moving distance of the swing amplitude and the moving distance of the needle pitch, form the stitch offset based on the deviation of the swing amplitude and needle pitch, and perform dynamic correction of the swing amplitude and needle pitch through the current output torque of the servo controller of the sewing transmission structure to determine the theoretical error δ2.

8. The stitch control method for a sewing machine according to claim 7, characterized in that: The specific process of dynamic correction of swing amplitude and needle distance is as follows: (a) Coordinate transformation error compensation: Let the current needle insertion point be P i =(x i ,y i ), the error correction process performs coordinate transformation error compensation: ； Where δx, δy are the trace offsets, J -1 is the inverse matrix of the Jacobian matrix, εs is the sensor noise; (b) Servo torque correction formula: used for the coupling control quantity of the rotary axis and the linear axis: ； where K p ,K d , are proportional control coefficient and differential control coefficient respectively, K f is the vibration coefficient, F vibration is the vibration disturbance force measured by the laser vibrometer.

9. The stitch control method for a sewing machine according to claim 1, characterized in that: The error fusion function Δδ of S3 is as follows: ； Where: δ1 is the system error; A j is the amplitude of the pendulum angular velocity spectrum, is the amplitude of the reference pendulum angular velocity spectrum; σ vib is the RMS value of vibration acceleration; α, β, and γ are weight coefficients, satisfying α+β+γ=1.

10. The stitch control method for a sewing machine according to claim 1, characterized in that: The iterative optimization pattern stitch of S4 includes: S4-1. Optimize and minimize the objective function using the improved NSGA algorithm; F min =(Δδ,T execute ) Among them, T execute is the execution time of a single iteration, The constraints are as follows: g1:Δδ≤δ h , g2:τ max ≤τ limit , τ max is the maximum value of the servo torque, τ limit is the servo torque limit value.

Images