Cable reel flexible production equipment control method based on industrial internet

Abstract

The present application relates to the technical field of industrial internet, in particular to a cable reel flexible production equipment control method based on industrial internet. The present application takes the spindle encoder pulse extreme point as the interrupt source to trigger the acquisition of the wire winding cut point physical three-dimensional coordinates; the wire guide theoretical pulse coordinates are converted into theoretical three-dimensional coordinates, the two coordinate space deviations are calculated to extract the wire guide lag pose deviation matrix and calculate the maximum geometric opening angle; according to the maximum geometric opening angle and the target tension, the transverse component is calculated and the fluctuation amount is extracted; the elastic deformation elongation is calculated by combining the free length of the suspended section and the physical property parameters of the wire, which is converted into a position leading compensation pulse amount to control the wire guide servo motor to execute the leading compensation; the running electrical signals are collected within the compensation period, the process equivalent components are filtered out to extract the harmonic characteristic residual error; the residual error is calculated for stability to extract the energy change rate index, and when the index exceeds the limit, the take-up reel servo motor is controlled to output a reverse damping torque.

Description

Technical Field

[0001] This invention relates to the field of industrial internet technology, specifically to a control method for flexible cable reel production equipment based on industrial internet. Background Technology

[0002] In the complete industrial chain of wires, cables, and metal wires, there are typically upstream cable manufacturing processes and downstream finished cable reel rewinding processes. During the finished cable reel production stage, the cable reel, as the load-bearing substrate, is subject to precise linkage control during its high-speed winding and unwinding processes, which directly determines the final product's tightness and flatness. With the shift in manufacturing models towards flexible production with multiple product varieties and high-frequency changes, higher demands are placed on the dynamic response accuracy and adaptability to various operating conditions of flexible cable reel production equipment.

[0003] Current cable winding and unwinding equipment control schemes primarily rely on independent stand-alone programmable logic controllers (PLCs) combined with proportional-integral-derivative (PI-DE) algorithms, supplemented by mechanical tension adjustment devices such as counterweight cable storage racks. Cable unwinding control often employs fixed program dead time for lead or lag compensation during pole reversal; tension control typically uses variable-parameter PID algorithms based on roll diameter calculations for steady-state error feedback adjustment. Furthermore, while some systems incorporate Industrial Internet architecture, this is mainly limited to work order issuance and macro-level status monitoring at the upper-level process system level, failing to achieve integration with the underlying multi-axis servo motor high-frequency control algorithms.

[0004] Under high-speed and unsteady-state conditions, the aforementioned existing technologies exhibit significant control deficiencies. In the winding dimension, fixed dead-zone compensation cannot adapt to the elastic deformation hysteresis effect of different wire specifications under high-speed motion. A dynamic, nonlinear spatial phase difference exists between the mechanical displacement signal of the winding guide roller and the actual physical landing point of the wire on the reel, easily leading to shape distortions such as "roller stretching" or "not reaching the edge" at the flange reversal pole. In the tension dimension, the PID model based on steady-state feedback is prone to control overshoot and low-frequency oscillations when faced with sudden equipment starts / stops or large nonlinear changes in rotational inertia due to increased reel diameter, resulting in plastic deformation or wire breakage. Furthermore, when the winding reaches the reversal pole, the sudden change in the spatial geometric angle of the wire alters the tension vector distribution at the actual winding tangent point; and the abrupt change in winding radius caused by minor distortions in the winding shape instantly generates sudden changes in rotational inertia and reverse tensile stress. The existing control architecture decouples the cabling and tension as isolated closed loops, which makes it very easy for the two control loops to interfere with each other at the moment of pole commutation.

[0005] In summary, how to solve the problem of system coordinated control instability and adjustment lag caused by the strong nonlinear spatiotemporal coupling between the spatial displacement of multiple actuators and the dynamic mechanical state of the wire in non-steady-state flexible production equipment is a technical problem that urgently needs to be solved in this field.

[0006] To address this, a control method for flexible cable reel production equipment based on the Industrial Internet is proposed. Summary of the Invention

[0007] The purpose of this invention is to provide a control method for flexible cable reel production equipment based on the Industrial Internet. This invention uses the extreme pulse point of the spindle encoder as an interrupt source to trigger and obtain the physical three-dimensional coordinates of the wire winding tangent point; it converts the theoretical pulse coordinates of the cable guide wheel into theoretical three-dimensional coordinates, calculates the spatial deviation between the two coordinates to extract the cable hysteresis pose deviation matrix, and calculates the maximum geometric angle; it calculates the lateral component force based on the maximum geometric angle and the target tension and extracts its fluctuation; it combines the free length of the suspended section with the physical property parameters of the wire to calculate the elastic deformation elongation, converts it into a position advance compensation pulse quantity to control the cable guide wheel servo motor to perform advance compensation; it collects the operating electrical signal during the compensation period, filters out the process equivalent components, and extracts the harmonic characteristic residuals; it performs stability calculations on the residuals to extract the energy change rate index, and controls the take-up reel servo motor to output reverse damping torque when the limit is exceeded. This invention solves the problems of system collaborative control instability and adjustment lag in flexible cable reel production equipment under unsteady conditions. To achieve the above objectives, the present invention provides the following technical solution: The control method for flexible cable reel production equipment based on the Industrial Internet includes: Using the pulse extreme point of the encoder on the take-up reel as the interrupt source, the interruption is triggered when the wire guide wheel runs to the preset angle range at both ends of the cable reel flange, and the physical three-dimensional coordinates of the wire winding cutting point are obtained. The theoretical pulse coordinates of the cable guide wheel are mapped to theoretical three-dimensional coordinates. The physical three-dimensional coordinates are transformed to the same coordinate system through rigid body transformation and then subtracted to extract the cable hysteresis pose deviation matrix. The maximum geometric angle formed by the cable guide wheel and the winding tangent point is calculated. The lateral force along the cable reel axis is calculated based on the maximum geometric angle and the target tension, and the fluctuation of the lateral force is extracted. Based on the fluctuation amount, the free length of the suspended section of the wire and the physical property parameters, the elastic deformation elongation of the wire is calculated and converted into a position advance compensation pulse amount, and the servo motor of the wire guide wheel is controlled to perform advance compensation. The system continuously collects the operating electrical signals of the take-up reel servo motor, filters out the equivalent electrical signal components generated by production stage commands, extracts harmonic characteristic residuals, and extracts the energy change rate index through stability calculation. When the energy change rate index exceeds the stability threshold, a reverse damping torque command is injected into the underlying servo driver to drive the take-up reel servo motor to output reverse damping torque.

[0008] Preferably, acquiring the physical three-dimensional coordinates of the wire winding cutting point specifically includes: configuring a non-contact three-dimensional optical sensor as a physical coordinate acquisition unit; when the interrupt source is triggered, the physical coordinate acquisition unit switches to the active working state and performs local field-of-view scanning within a preset angle range at both ends of the cable reel flange; acquiring three-dimensional spatial contour feature data of the contact point between the edge of the wire winding trajectory and the inner wall of the flange; processing the three-dimensional spatial contour feature data using an edge reconstruction algorithm to calculate the physical three-dimensional coordinates of the wire winding cutting point in the basic coordinate system of the physical coordinate acquisition unit.

[0009] Preferably, extracting the hysteresis pose deviation matrix for wiring specifically includes: constructing a rigid body transformation matrix for aligning the physical coordinate acquisition unit and the wiring mechanism; performing spatial translation and rotation transformations on the physical three-dimensional coordinates using the rigid body transformation matrix to unify the physical three-dimensional coordinates to the mechanical transmission coordinate system of the wiring mechanism; subtracting the translated and rotated physical three-dimensional coordinates from the theoretical three-dimensional coordinates to obtain the linear deviation components and angular deviation components in the three axes of the spatial rectangular coordinate system, and combining them to form the hysteresis pose deviation matrix for wiring that characterizes the direction and amplitude of spatial geometric distortion.

[0010] Preferably, the maximum geometric angle formed by the cable guide wheel and the winding tangent point is calculated, and the lateral component of the force along the cable reel axis is calculated based on the maximum geometric angle and the target tension. Specifically, this includes: extracting the axial position deviation component and the radial position deviation component from the cable hysteresis pose deviation matrix; calculating the arctangent trigonometric function values ​​of the axial position deviation component and the radial position deviation component to obtain the dynamic angle value between the cable guide wheel axis and the cable centerline; using the flange geometric boundary parameters of the cable reel at the cable reversal pole as boundary constraints to correct the dynamic angle value and determine the maximum geometric angle formed by the line connecting the cable guide wheel and the winding tangent point at the moment of flange edge reversal; obtaining the target tension issued by the system, and using a sine function to calculate the component of the target tension in the direction of the maximum geometric angle, which is defined as the lateral component of the force along the cable reel axis.

[0011] Preferably, calculating the elastic deformation elongation of the wire specifically includes: continuously monitoring the calculated lateral force value within a preset time window, extracting the difference between its maximum and minimum values ​​as the fluctuation of the lateral force; retrieving the elastic modulus and cross-sectional area parameters of the current wire, and obtaining the spatial straight-line distance between the center point of the current wire guide wheel and the winding tangent point as the free length of the suspended section; using the free length of the suspended section and the maximum geometric angle, vectorizing and converting the fluctuation of the lateral force into the additional tensile tension fluctuation along the wire axis; dividing the additional tensile tension fluctuation by the cross-sectional area to obtain the tensile stress variable, performing linear strain calculation in combination with the elastic modulus, and integrating to obtain the elastic deformation elongation of the wire along the longitudinal direction.

[0012] Preferably, the position advance compensation pulse quantity is converted into a position advance compensation pulse quantity to control the cable guide wheel servo motor to perform advance compensation. Specifically, this includes: obtaining the inherent pulse resolution per revolution of the cable guide wheel servo motor and the pitch parameter of the cable lead screw; multiplying the elastic deformation elongation by the sine value of the maximum geometric angle and projecting it into an equivalent displacement deviation along the axial direction of the cable reel; dividing the equivalent displacement deviation by the pitch parameter and multiplying it by the inherent pulse resolution per revolution to calculate the absolute value of the target pulse; performing phase advance prediction on the absolute value of the target pulse based on a preset advance time constant to generate a position advance compensation pulse quantity with time prediction characteristics, and injecting it into the bottom position loop controller of the cable guide wheel servo motor to control the servo motor to perform an equivalent advance displacement in advance before mechanical switching, thereby offsetting the winding position lag caused by the elastic deformation of the wire.

[0013] Preferably, the extraction of harmonic characteristic residuals specifically includes: acquiring the three-phase stator current signal and back electromotive force signal of the take-up reel servo motor controller as the initial operating electrical signal; receiving production stage instructions issued by the upper-level process system of the Industrial Internet; inputting the macroscopic target torque and speed corresponding to the production stage instructions into a preset nonlinear electromagnetic state observer, and calculating the equivalent electrical signal components including macroscopic load step and inherent nonlinear harmonics inside the motor; performing high-frequency differential comparison between the initial operating electrical signal and the equivalent electrical signal components, stripping the background electrical signal, and extracting the harmonic characteristic residuals.

[0014] Preferably, the stability calculation of the harmonic characteristic residuals specifically includes: constructing a Lyapunov numerical analysis function to quantitatively represent the energy convergence state of the accumulator disk rotation dynamics system; inputting the extracted harmonic characteristic residuals as state variables into the Lyapunov numerical analysis function in real time, and obtaining the energy change rate index by calculating the first-order time derivative of the state variables; if the energy change rate index is greater than zero for a consecutive preset number of servo sampling periods, an over-limit warning is triggered; comparing the cumulative number of control loops with the energy change rate index being greater than zero with the root mean square value of the amplitude of the harmonic characteristic residuals, when either the cumulative number of control loops or the root mean square value of the amplitude exceeds the corresponding preset tolerance upper limit, it is determined that the energy change rate index exceeds the set stability threshold.

[0015] Preferably, generating a damping torque command in the opposite direction and injecting it into the underlying servo driver to control the take-up reel servo motor to output the reverse damping torque specifically includes the following steps: extracting the dominant harmonic frequency, instantaneous positive and negative signs, and amplitude of the harmonic characteristic residual when the energy change rate index exceeds the stability threshold; if the dominant harmonic frequency exceeds the safe phase margin bandwidth, mapping the amplitude to the compensation current amplitude based on the motor torque constant, and generating an initial command in combination with the opposite direction of the instantaneous positive and negative signs; after the initial command is limited by a preset maximum allowable current threshold, generating an anti-disturbance quadrature-axis current control command; directly injecting the anti-disturbance quadrature-axis current control command into the current loop control register of the underlying servo driver to drive the take-up reel servo motor to generate an electromagnetic damping torque that tracks the harmonic characteristic residual fluctuations and is in the opposite direction, absorbing the excitation energy transmitted to the axial direction of the take-up reel by the wire reel commutation impact through the electromagnetic damping torque, thereby achieving physical suppression of nonlinear harmonic chattering.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention establishes a mapping mechanism between hardware interrupt source triggering and the hysteresis spatial phase tensor. It transforms the spatial three-dimensional coordinates acquired on-site and calculates the dynamic geometric angle of the commutation pole. Based on this, the target tension is orthogonally decomposed into a lateral component, which is then converted into a position advance compensation pulse for the cable guide wheel servo motor. This design changes the closed-loop adjustment mode in traditional control systems that isolates spatial cable routing from mechanical tension. Before the physical impact of high-speed commutation occurs, the small-inertia cable routing mechanism actively absorbs the landing point deviation caused by elastic deformation through spatial displacement, effectively suppressing the morphological distortion that easily occurs when the wire reverses at the flange edge.

[0017] 2. This invention performs data acquisition and intervention on the take-up reel servo motor within a specific time period during which the guide wheel servo motor performs position advance compensation. By coordinating within this highly volatile spatiotemporal window, the advance fine-tuning of spatial position and the micro-torque compensation of the underlying driver achieve rigid locking and coordinated response on the time axis. This strong temporal coupling control avoids speed misalignment caused by response speed differences during unsteady transition phases such as speed changes and commutations in multi-axis servo systems, ensuring process stability in continuous production.

[0018] 3. After acquiring the operating electrical signal, this invention removes the equivalent electrical signal component generated by production stage commands, extracts the harmonic characteristic residual caused by external excitation, and performs stability determination based on the energy change rate index. Once an exceedance is detected, it bypasses the delay of the upper-layer bus communication cycle and directly injects a damping torque command in the opposite direction into the lower-layer servo driver, achieving instant damping at the current loop level. This physical suppression mechanism, which directly eliminates harmonic chattering at the lower layer, prevents micro-vibrations from spreading to the macro-control loop, reducing the risk of system oscillation overshoot and wire damage. Attached Figure Description

[0019] Figure 1 The present invention provides a control method for flexible cable reel production equipment based on the Industrial Internet. Figure 2 This is a hardware and software interaction architecture diagram of the device control system provided in an embodiment of the present invention; Figure 3 This is a data flow graph of spatial-mechanical cross-domain mapping and feedforward compensation provided for embodiments of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see Figures 1 to 3 This invention provides a control method for flexible cable reel production equipment based on the Industrial Internet, and the technical solution is as follows: The control method for flexible cable reel production equipment based on the Industrial Internet includes: Using the pulse extreme point of the encoder on the take-up reel as the interrupt source, the interruption is triggered when the wire guide wheel runs to the preset angle range at both ends of the cable reel flange, and the physical three-dimensional coordinates of the wire winding cutting point are obtained. The theoretical pulse coordinates of the cable guide wheel are mapped to theoretical three-dimensional coordinates. The physical three-dimensional coordinates are transformed to the same coordinate system through rigid body transformation and then subtracted to extract the cable hysteresis pose deviation matrix. The maximum geometric angle formed by the cable guide wheel and the winding tangent point is calculated. The lateral force along the cable reel axis is calculated based on the maximum geometric angle and the target tension, and the fluctuation of the lateral force is extracted. Based on the fluctuation amount, the free length of the suspended section of the wire and the physical property parameters, the elastic deformation elongation of the wire is calculated and converted into a position advance compensation pulse amount, and the servo motor of the wire guide wheel is controlled to perform advance compensation. The system continuously collects the operating electrical signals of the take-up reel servo motor, filters out the equivalent electrical signal components generated by production stage commands, extracts harmonic characteristic residuals, and extracts the energy change rate index through stability calculation. When the energy change rate index exceeds the stability threshold, a reverse damping torque command is injected into the underlying servo driver to drive the take-up reel servo motor to output reverse damping torque.

[0022] Example 1: This embodiment proposes a control method for flexible cable reel production equipment based on the Industrial Internet. Its typical application scenario is a high-speed take-up and unwinding production line for flexible cables (such as special optical fiber composite cables or micro-fine copper wires) with multiple product types and high-frequency production changes. In this context, the production line operates at high speed and in a non-steady-state condition. The spatial geometrical abrupt change of the guide roller at the flange reversal point is strongly coupled with the dynamic tension fluctuation of the take-up reel, easily causing wire accumulation or breakage at the extreme point. Traditional independent closed-loop systems relying on a single logic controller suffer from lag in regulation and cannot maintain process stability. The method in this embodiment is implemented through the following specific steps.

[0023] As one embodiment of the present invention, refer to Figure 1 A control method for flexible cable reel production equipment based on the Industrial Internet, referring to Figure 2 The hardware and software interaction architecture diagram of the equipment control system is shown in the reference diagram. Figure 3 Spatial-mechanical cross-domain mapping and feedforward compensation data flow graph.

[0024] All embodiments of this invention control the "multi-axis linkage mechanism for winding and coiling cables" (i.e., flexible cable reel production equipment) in the stage of precision winding and coiling of cables. By controlling the position advance compensation of the servo motor for the winding and coiling and injecting the bottom current loop damping of the servo motor for the winding reel, the physical behavior of the flexible wires being wound and arranged in a high-precision manner on the reel is controlled. The aim is to produce "finished cable reels" with flat winding and uniform tension. The overall solution does not involve any physical or chemical processing technology of the wire body in the preceding process.

[0025] The controller reads the pulse signal fed back by the absolute encoder of the take-up reel spindle servo motor in real time. In the cable laying control architecture, there is a strict electronic gear or mechanical speed ratio mapping relationship between the axial displacement of the cable guide roller and the number of rotations of the take-up reel spindle. Through initial stroke calibration, the pulse value of the take-up reel spindle encoder corresponding to the extreme physical position when the cable guide roller approaches the inner wall of the flange on both sides of the cable reel is set as the pulse extreme point. To reduce the ineffective computing power overhead of the control system, this pulse extreme point is configured as the low-level hardware interrupt source of the controller. Since the elastic deformation of the wire at the flange edge does not only occur at the extreme point instantaneously, but exists in a specific physical transition region, a preset angle range is set around the above-mentioned pulse extreme point. For example, the preset angle range can be set as the spindle rotation angle corresponding to the spatial span of 3 to 5 wire diameter widths offset inward from the inner boundary of the flange by the cable guide roller. When the cable guide roller runs to this preset angle range, the pulse value output by the encoder falls into the trigger range, and the low-level hardware circuit directly generates an interrupt level signal to trigger the interrupt source, thereby waking up the subsequent spatial coordinate acquisition action.

[0026] Further, acquiring the physical three-dimensional coordinates of the wire winding cutting point specifically includes: configuring a non-contact three-dimensional optical sensor as a physical coordinate acquisition unit; when the interrupt source is triggered, the physical coordinate acquisition unit switches to the active working state and performs local field-of-view scanning within a preset angle range at both ends of the cable reel flange; acquiring the three-dimensional spatial contour feature data of the contact point between the edge of the wire winding trajectory and the inner wall of the flange; processing the three-dimensional spatial contour feature data using an edge reconstruction algorithm to calculate the physical three-dimensional coordinates of the wire winding cutting point in the basic coordinate system of the physical coordinate acquisition unit.

[0027] Specifically, a non-contact 3D optical sensor (such as a binocular machine vision sensor or a laser array displacement sensor) is configured above or to the side of the cable winding mechanism as a physical coordinate acquisition unit. In the non-reversing, conventional cable winding section, this physical coordinate acquisition unit is in a dormant state to conserve computing power. When the underlying controller receives a pulse value from the encoder of the take-up reel spindle reaching a set extreme point—that is, when the cable guide wheel has moved to a preset angle range from the flange edges on both sides of the cable reel—it directly triggers the physical coordinate acquisition unit to switch to active operation via a hardware interrupt pin. After activation, the sensor performs a high-frequency local field-of-view scan within the preset angle range, acquiring 3D spatial contour feature data (such as image edge features or spatial depth point cloud data) of the point where the outermost winding trajectory edge of the cable and the inner wall of the cable reel flange will come into contact and collide. Subsequently, the edge computing node calls an edge reconstruction algorithm based on least squares fitting to process the point cloud data, removing noise caused by dust or lighting conditions, and calculating the absolute spatial coordinates of the cable at the winding tangent point in real time. These coordinates are the actual position parameters in a basic coordinate system established with the optical center of the lens of the physical coordinate acquisition unit as the origin.

[0028] This invention changes the conventional machine vision's continuous monitoring mode by introducing a hardware interrupt mechanism and local scanning within a preset angle range. This on-demand activation mechanism reduces the invalid data processing load on edge computing nodes, while ensuring that physical spatial location data can be acquired in areas with high incidence of morphological distortion (flange poles), providing reliable data source support for subsequent spatial deviation calculations.

[0029] Further, the hysteresis pose deviation matrix of the wiring mechanism is extracted, specifically including: constructing a rigid body transformation matrix for aligning the physical coordinate acquisition unit and the wiring mechanism; performing spatial translation and rotation transformations on the physical three-dimensional coordinates using the rigid body transformation matrix to unify the physical three-dimensional coordinates to the mechanical transmission coordinate system of the wiring mechanism; subtracting the translated and rotated physical three-dimensional coordinates from the theoretical three-dimensional coordinates to obtain the linear deviation components and angular deviation components in the three axes of the spatial rectangular coordinate system, and combining them to form the hysteresis pose deviation matrix of the wiring mechanism, which characterizes the direction and amplitude of spatial geometric distortion.

[0030] Specifically, the controller internally constructs a rigid body transformation matrix to align the coordinate system of the physical coordinate acquisition unit with the coordinate system of the wiring mechanism. The translation and rotation parameters of this matrix are three-dimensional spatial relative position constants calculated through multi-point calibration during the initial installation and calibration phase of the equipment. Using this rigid body transformation matrix, spatial translation and rotation transformation operations are performed on the physical three-dimensional coordinates to eliminate spatial perspective errors caused by sensor installation angles and positions, strictly unifying them under the mechanical transmission coordinate system of the wiring mechanism. Simultaneously, the encoder feedback pulse value of the wiring guide wheel servo motor in the current sampling period is read and substituted into the geometric mapping equation of the mechanical transmission chain of the wiring mechanism.

[0031] The calculation logic of the geometric mapping equation of the mechanical transmission chain is as follows: divide the absolute value of the encoder feedback pulse by the pulse resolution per revolution of the servo motor, and then divide by the reduction ratio of the reducer to obtain the theoretical number of revolutions of the ball screw; then multiply the theoretical number of revolutions by the lead parameter of the ball screw to calculate the single-dimensional linear displacement of the guide wheel in the transmission axis; finally, combined with the initial spatial calibration origin of the equipment, use the single-dimensional linear displacement as the translation increment of the corresponding coordinate axis to calculate the theoretical three-dimensional coordinates of the current spatial center point of the guide wheel in the mechanical transmission coordinate system. Due to the flexible hysteresis of the wire under high-speed motion and possible mechanical installation gaps, the transformed physical three-dimensional coordinates and the theoretical three-dimensional coordinates do not coincide in space. Subtract the two in the X, Y, and Z directions of the spatial rectangular coordinate system to obtain their independent linear deviation components; and combine the vector dot product operation to obtain the angular deviation component. Finally, the linear deviation components and angular deviation components are combined according to the matrix structure to form a lag pose deviation matrix for the winding. This matrix fully represents the three-dimensional direction and magnitude of the geometric distortion of the winding landing point under the current working condition.

[0032] The hysteresis pose deviation matrix is ​​represented as a 4x4 homogeneous transformation block matrix structure. The first three rows and three columns of the matrix form a third-order antisymmetric rotation matrix consisting of pitch, yaw, and roll angle deviations; the fourth column of the first three rows is a three-dimensional translation column vector consisting of linear deviation components of the X, Y, and Z axes; the fourth row is set as a constant row vector [0,0,0,1]. This standard fourth-order homogeneous matrix allows for the allocation of pose variables in the memory of the underlying controller, representing the absolute pose distortion in three-dimensional space.

[0033] This invention enables the comparison between theoretical control commands and actual physical morphology within the same mathematical dimension. By constructing a rigid body transformation matrix and extracting a second-order spatial phase tensor, multi-source spatial errors are structurally reduced in dimension, eliminating the interference of static installation tolerances and enabling the quantification of pure time-delay distortions in high-speed dynamic cabling processes.

[0034] Further, the maximum geometric angle formed by the cable guide wheel and the winding tangent point is calculated. Based on the maximum geometric angle and the target tension, the lateral force along the cable reel axis is calculated. Specifically, this includes: extracting the axial position deviation component and the radial position deviation component from the cable hysteresis pose deviation matrix; calculating the arctangent trigonometric function values ​​of the axial position deviation component and the radial position deviation component to obtain the dynamic angle value between the cable guide wheel axis and the cable centerline; using the flange geometric boundary parameters of the cable reel at the cable reversal pole as boundary constraints, and correcting the dynamic angle value through extreme value comparison logic: that is, when the calculated dynamic angle value is greater than the sum of the inner tilt angle and the width tolerance conversion angle of the flange, the dynamic angle value is forcibly limited to the physical maximum critical angle allowed by the geometric boundary parameters, thereby determining the maximum geometric angle formed by the line connecting the cable guide wheel and the winding tangent point at the moment of flange edge reversal; obtaining the target tension issued by the system, and using a sine function to calculate the component of the target tension in the direction of the maximum geometric angle, which is defined as the lateral force along the cable reel axis.

[0035] Specifically, the cable hysteresis pose deviation matrix is ​​used as input data. The axial position deviation component parallel to the cable reel's axis and the radial position deviation component perpendicular to the axis are extracted. The edge computing node uses the axial position deviation component as the opposite edge value and the radial position deviation component as the adjacent edge value, performing arctangent trigonometric function calculations to obtain the dynamic angle between the cable guide wheel's central axis and the cable's centerline. To ensure safety control, the flange geometric boundary parameters (such as the flange's inner inclination angle and width tolerance) at the cable reel's reversal pole are retrieved as boundary constraints. The calculated dynamic angle value is compared and corrected with the flange boundary constraints to determine the maximum geometric angle formed in space by the line connecting the cable guide wheel and the winding tangent point at the instant the cable guide wheel reaches the flange edge to perform mechanical reversal. Based on this, the target tension scalar set for the current production process and issued by the upper-level process system of the Industrial Internet is obtained through the Industrial Internet interface. By multiplying the target tension scalar by the sine trigonometric function value of the maximum geometric angle, orthogonal decomposition of the spatial mechanics vector is performed to calculate the component of the target tension scalar in the direction of the maximum geometric angle. This component is then explicitly defined as the lateral force acting on the wire along the axial direction of the cable reel, attempting to pull the wire off course.

[0036] This invention breaks through the limitations of traditional tension control and cable control by transforming spatial geometric distortions into mechanical angles and performing orthogonal decomposition of the target tension. It quantitatively analyzes the influence of spatial position changes on the force on the cable, realizes bidirectional coupling prediction of cable geometry and tension disturbance, and lays a physical and mechanical foundation for subsequent cross-control domain feedforward compensation.

[0037] The calculation of the lateral force along the cable reel axis based on the maximum geometric angle and the target tension issued by the system further includes the following steps: introducing tension wave velocity variable parameters using a preset string vibration differential model, and continuously acquiring morphological data of the cable under high-speed operation through the physical coordinate acquisition unit for multiple frames; fitting a microscopic vibration envelope surface characterizing the transient oscillation waveform of the cable based on the multiple frames of morphological data, and substituting the spatial geometric amplitude of the microscopic vibration envelope surface as a boundary excitation condition into the string vibration differential model to calculate and solve for the high-frequency dynamic angular displacement compensation amount; using the high-frequency dynamic angular displacement compensation amount to dynamically correct the calculated maximum geometric angle in real time, and outputting the transient lateral force containing dynamic angle fluctuations.

[0038] The preset string vibration differential model is a second-order partial differential physical model based on d'Alembert's wave theory. In the edge computing node, this model uses the lateral vibration displacement of the wire as the dependent variable and the geometric coordinates of the wire along the spatial axis and time as independent variables. Internally, the model calculates the mass per unit length of the wire based on the retrieved wire material density and cross-sectional area. At each extreme interruption trigger, the current process target tension is read, and the tension wave velocity variable parameter under the current operating condition is dynamically calculated using the square root of the ratio of the target tension to the mass per unit length. During the numerical solution process, the edge computing node uses the maximum geometric amplitude of the microscopic vibration envelope surface extracted by the visual sensor as a Dirichlet boundary condition, substitutes it into the partial differential equation, and uses the finite difference method for high-frequency discretization. Thus, the complex continuous vibration spatial waveform is reduced in dimension and transformed into a quantifiable transient high-frequency dynamic angular displacement compensation quantity within a servo control cycle, ensuring that the dynamic angle correction has a solid material dynamics basis.

[0039] Specifically, in the step of calculating the lateral component of the force along the cable reel axis based on the maximum geometric angle and the target tension issued by the system, dynamic compensation is provided for the string vibration caused by the high-speed axial movement of the flexible wire and flange collision. A first-order and higher-order string vibration differential model of the wire based on the wave equation is pre-constructed, and a tension wave velocity variable parameter including the current process target tension scalar and the wire unit volume density parameter is introduced. After the aforementioned hardware interruption source is triggered, a binocular machine vision sensor or a laser array displacement sensor continuously captures multiple frames of physical spatial morphology data of the wire in the reversing transition zone at a high sampling rate matching or higher than the refresh rate of the underlying servo position ring (e.g., industrial-grade high frame rate of 1000 frames / second to 4000 frames / second). Edge computing nodes extract the maximum swing boundary of the wire's geometric central axis from the multi-frame data through temporal image stacking and feature point trajectory tracking, thereby fitting a microscopic vibration envelope characterizing the high-frequency dynamic oscillation waveform of the wire. Subsequently, the edge computing node substitutes the spatial geometric amplitude of the microscopic vibration envelope as the boundary excitation condition into the string vibration differential model to calculate the transient dynamic angular displacement compensation caused by the high-speed movement of the wire. Finally, this transient dynamic angular displacement compensation is directly superimposed on the maximum geometric angle calculated in the previous steps for real-time dynamic correction, outputting a transient lateral force including the dynamic angle fluctuation, which serves as the mechanical input for subsequent elastic deformation calculations.

[0040] This invention upgrades the traditional static geometric angle calculation to a dynamic angle model that incorporates the dynamic wave characteristics of wire by introducing string vibration wave velocity parameters and micro-vibration envelope surfaces. It solves the problem of uneven transient force caused by string vibration in high-speed axial reciprocating motion of wires and flange collisions, enabling the decomposed lateral force component to reflect the force jumps of the physical entity. This provides a dynamically characteristic mechanical excitation source for subsequent position advance compensation and further suppresses morphological distortion at the poles.

[0041] Further, the calculation of the elastic deformation elongation of the wire includes: continuously monitoring the calculated lateral force value within a preset time window, extracting the difference between its maximum and minimum values ​​as the fluctuation of the lateral force; retrieving the elastic modulus and cross-sectional area parameters of the current production wire, and obtaining the spatial straight-line distance between the center point of the current wire guide wheel and the winding tangent point as the free length of the suspended section; using the free length of the suspended section and the maximum geometric angle, vectorizing and converting the fluctuation of the lateral force into the additional tensile tension fluctuation along the wire axis; dividing the additional tensile tension fluctuation by the cross-sectional area to obtain the tensile stress variable, combining it with the elastic modulus to perform linear strain calculation, and integrating to obtain the elastic deformation elongation of the wire along the longitudinal direction.

[0042] The underlying implementation mechanism for continuous monitoring within a preset time window is as follows: the edge computing node allocates a circular sliding buffer queue in memory based on a first-in-first-out (FIFO) data structure. The physical duration of the preset time window is rigorously calibrated based on the dynamic inertial response time of acceleration and deceleration during mechanical reversal of the cable guide wheel, typically ranging from 5 to 15 milliseconds. This window covers dozens of underlying servo control cycles. When the equipment reaches the preset angle range of the flange, the real-time calculated lateral force values ​​are continuously pushed into the circular sliding buffer queue using a high-frequency clock equivalent to the refresh frequency of the underlying position ring. Within each clock interrupt cycle, the microprocessor traverses all discrete lateral force values ​​stored within the time window, executes an extreme value optimization algorithm to extract the maximum peak and minimum valley values ​​within the current window, and defines the absolute difference between the two as the lateral force fluctuation caused by the transient reversal impact. As new sampled data is pushed in, the oldest data in the queue is automatically discarded.

[0043] Specifically, the calculated lateral force values ​​are continuously monitored within a preset time window, and their fluctuations are extracted through differential calculations. Simultaneously, the physical property parameters of the current batch of production wire are retrieved online via the upper-level process system interface or directly from the local database of the edge computing node. These parameters include at least the elastic modulus (characterizing the material's resistance to elastic deformation), cross-sectional area, and Poisson's ratio. Next, the spatial straight-line distance between the center point of the wire guide wheel and the winding tangent point of the take-up reel is obtained at the current moment through sensor readings or geometric derivation, and this distance is assigned as the free length of the suspended section of the wire. The edge computing node has a built-in material mechanics deformation decoupling model based on Hooke's Law and fundamental theories of material mechanics. The fluctuations of the lateral force, the free length of the suspended section of the wire, and the physical property parameters of the wire are input into this decoupling model as independent variables. The model first uses the maximum geometric angle to divide the fluctuation of the lateral component force by the sine of the maximum geometric angle to calculate the additional tensile tension fluctuation along the wire axis. Then, the additional tensile tension fluctuation is divided by the cross-sectional area of ​​the wire to obtain the longitudinal tensile stress variable, and the linear strain value is calculated by combining the elastic modulus. Finally, the linear strain value is multiplied by the current free length of the suspended section of the wire to calculate the elastic deformation elongation of the wire along the longitudinal direction. This value represents the tensile distortion distance at a purely physical level.

[0044] This invention transforms abstract mechanical fluctuation parameters into intuitive physical dimensional deviations. By introducing the inherent material mechanical properties and physical geometric constraints of the wire itself, the control system possesses the ability to adapt to different wire characteristics, ensuring differentiated quantitative assessment of the stress deformation of cables made of different materials during flexible switching between multiple product types.

[0045] Further, the position advance compensation pulse quantity is converted into a position advance compensation pulse quantity, which controls the servo motor of the cable guide wheel to perform advance compensation. Specifically, this includes: obtaining the inherent pulse resolution per revolution of the servo motor of the cable guide wheel and the pitch parameter of the cable lead screw; multiplying the elastic deformation elongation by the sine value of the maximum geometric angle and projecting it into an equivalent displacement deviation along the axial direction of the cable reel; dividing the equivalent displacement deviation by the pitch parameter and multiplying it by the inherent pulse resolution per revolution to calculate the absolute value of the target pulse; performing phase advance prediction on the absolute value of the target pulse based on a preset advance time constant to generate a position advance compensation pulse quantity with time prediction characteristics, and injecting it into the bottom position loop controller of the servo motor of the cable guide wheel, controlling the servo motor to perform an equivalent advance displacement in advance before mechanical switching, thereby offsetting the winding position lag caused by the elastic deformation of the wire.

[0046] Specifically, the inherent pulse resolution per revolution of the cable guide wheel servo motor (e.g., 10,000 pulses / revolution) and the pitch parameter of the cable screw are first obtained from the underlying driver configuration file. Mathematical conversion logic is then executed, multiplying the previously calculated elastic deformation elongation by the sine trigonometric function value of the maximum geometric angle, and converting its spatial projection into an equivalent displacement deviation along the cable reel axis. Subsequently, this equivalent displacement deviation is divided by the pitch parameter and multiplied by the inherent pulse resolution per revolution to calculate the absolute value of the target pulse required to correct the landing point offset on the mechanical structure. To eliminate the hysteresis of the mechanical transmission response, a preset lead time constant is introduced. This lead time constant is the equivalent system lag time obtained through the initial motion step response calibration test of the equipment. Specifically, the lead time constant is set as the sum of the dead time of the industrial internet bus communication, the delay time of the internal position loop operation of the servo driver, and the time equivalent corresponding to the idle backlash of the cable mechanical transmission chain (e.g., the reducer and the ball screw). Based on this lead time constant, the absolute value of the target pulse is subjected to phase lead prediction processing (such as a first-order differential feedforward algorithm) to generate a position lead compensation pulse quantity with time prediction characteristics. Subsequently, this position lead compensation pulse quantity is directly and in parallel injected into the underlying position loop controller register of the cable guide wheel servo motor in the form of pulse superposition. In this way, the cable guide wheel servo motor is controlled to perform an advance displacement equivalent to the elastic deformation elongation before the actual physical mechanical commutation action occurs, actively adjusting the guide wheel position to compensate for the winding position lag caused by the elastic deformation of the wire.

[0047] This invention changes the traditional lag mode of passively waiting for errors to occur before performing PID closed-loop regulation. By converting deformation into underlying lead pulse compensation, mechanical position pre-adjustment is completed before spatial deformation occurs at the commutation pole, effectively suppressing wire accumulation and edge collapse defects in the commutation dead zone and improving the flatness of cable winding.

[0048] Furthermore, the harmonic characteristic residuals are extracted, specifically including: acquiring the three-phase stator current signal and back electromotive force signal of the take-up reel servo motor controller as the initial operating electrical signal; receiving the production stage instructions issued by the upper-level process system of the Industrial Internet; inputting the macroscopic target torque and speed corresponding to the production stage instructions into a preset nonlinear electromagnetic state observer, and calculating the equivalent electrical signal components including macroscopic load step and inherent nonlinear harmonics inside the motor; performing high-frequency differential comparison between the initial operating electrical signal and the equivalent electrical signal components, stripping the background electrical signal, and extracting the harmonic characteristic residuals.

[0049] Specifically, to overcome signal distortion caused by magnetic circuit saturation and nonlinear electromagnetic coupling during motor operation, residuals caused by external mechanical excitation are extracted, and a unified nonlinear full-order state observer architecture is constructed. Signal processing is performed synchronously within the time period of the lead-in compensation of the cable guide wheel servo motor. Three-phase stator current signals and back electromotive force signals are acquired in real time as initial operating signals through current sensors and voltage conditioning circuits deployed in the take-up reel servo motor drive circuit.

[0050] The nonlinear full-order state observer is built within the underlying driver. Its state equations are input variables: real-time quadrature and direct-axis stator voltage commands after Clark and Park coordinate transformations, and production stage commands issued by the system. The state variables are set as quadrature and direct-axis stator current and rotor angular velocity. The core mathematical model of this observer is based on linear electromagnetic equations derived from the motor's nominal inductance and resistance, with feedforward compensation for the magnetic circuit saturation characteristic function and cogging torque mapping matrix. Through this composite model, the observer calculates in real-time the estimated value of the quadrature and direct-axis stator current, which simultaneously includes macroscopic load steps and the motor's inherent nonlinear electromagnetic fluctuations. This estimated value is defined as the equivalent electrical signal component.

[0051] Subsequently, the actual initial operating electrical signal (i.e., the actual AC and DC axis current) is compared with the estimated stator current using high-frequency differential comparison to calculate the current residual. This current residual is multiplied by a preset feedback gain matrix and used as a correction term in the closed-loop feedback to the observer's state equation to ensure state convergence. Since the equivalent electrical signal component has already absorbed the electromagnetic nonlinearity of the motor body within the model, the high-frequency non-periodic current residual outside the observer's convergence bandwidth and not absorbed by the composite model represents the harmonic characteristic residual caused by external micro-mechanical excitation (such as commutation impact of cabling).

[0052] This invention effectively separates macroscopic process changes from microscopic mechanical disturbances within complex electromagnetic signals. By eliminating equivalent electrical signal components, it ensures that the system does not misinterpret normal process acceleration / deceleration as fault disturbances, thereby extracting the harmonic characteristic residuals that truly reflect the system's unsteady-state fluctuations and enhancing the anti-interference and decoupling capabilities of the underlying data processing.

[0053] Further, stability calculations are performed on the harmonic characteristic residuals, specifically including: constructing a Lyapunov numerical analysis function to quantitatively represent the energy convergence state of the rotating disc dynamics system; inputting the extracted harmonic characteristic residuals as state variables into the Lyapunov numerical analysis function in real time, and obtaining the energy change rate index by calculating the first-order time derivative of the state variables; if the energy change rate index is greater than zero for a consecutive preset number of servo sampling periods, an over-limit warning is triggered; comparing the cumulative number of control loops with the energy change rate index being greater than zero with the root mean square value of the amplitude of the harmonic characteristic residuals, when either the cumulative number of control loops or the root mean square value of the amplitude exceeds the corresponding preset tolerance upper limit, it is determined that the energy change rate index exceeds the set stability threshold.

[0054] The stability threshold employs a combined criterion, consisting of the upper tolerance limit corresponding to the cumulative number of control cycles where the energy change rate index is consistently greater than zero, and the upper tolerance limit corresponding to the root mean square value of the harmonic characteristic residual amplitude. To determine the upper tolerance limit of the root mean square amplitude value corresponding to the physical load-bearing capacity of the wire, the edge computing node retrieves the tensile yield strength stress of the current wire. Allowable working stress The cross-sectional area A, elastic modulus E, and current free length L of the suspended segment, where the allowable working stress is... The process parameters corresponding to the current wire specifications are preset and do not exceed the tensile yield stress. According to the linear elastic relationship, the maximum elastic potential energy of the wire before it reaches its yield limit is... The elastic potential energy allowed by the current process Let N be the upper limit of tolerance for the cumulative number of control loops, and the servo sampling period be... Then As the safety response time corresponding to the stability determination, and in accordance with Determine the allowable rate of elastic potential energy accumulation per unit time for the wire; simultaneously obtain the allowable electromagnetic dissipation power of the take-up reel servo motor. ,Will and The smaller value in the range is denoted as the allowable disturbance power. This is used to determine the upper limit of tolerance for the root mean square value of the harmonic characteristic residual.

[0055] Specifically, a Lyapunov numerical analysis function was constructed for the take-up reel rotation dynamics system to quantitatively characterize its energy convergence or divergence state. The harmonic characteristic residuals were used as state variables and input into this Lyapunov numerical analysis function in real time. Edge computing nodes used mathematical calculus modules to calculate the first-order time derivative of this state variable, thereby obtaining a specific value representing the current energy change rate index. If the obtained energy change rate index is greater than zero for a consecutive preset number (e.g., 3 to 5 consecutive servo sampling periods), it indicates that the excitation energy in the mechanical system is accumulating and not being effectively dissipated, triggering the underlying over-limit warning state. To prevent misjudgment due to occasional interference, the cumulative number of control cycles with the energy change rate index greater than zero (i.e., the number of discrete adjustment steps in which the index remains greater than zero) and the root mean square value of the harmonic characteristic residual amplitude during this period were further compared. Only when either the cumulative number of control cycles or the root mean square value of the amplitude exceeds the corresponding preset tolerance upper limit in the edge computing node is it officially determined that the energy change rate index has exceeded the set stability threshold, confirming that the system has entered an over-limit chattering state that is about to lead to tension loss of control or wire breakage.

[0056] The preset tolerance upper limit for the cumulative number of control cycles is determined based on the ratio of the inherent resonant response period of the mechanical transmission chain of the wiring mechanism to the sampling period of the underlying servo position loop. The physical purpose of setting this threshold is to filter out random white noise from the sensors and occasional single-cycle communication interference, ensuring that the system identifies real, continuous mechanical jitter. Under a typical servo control cycle in the hundreds of microseconds range (e.g., 250 microseconds), the tolerance upper limit for the cumulative number of control cycles is typically set between 5 and 15 cycles.

[0057] The preset tolerance upper limit corresponding to the root mean square value of the amplitude is based on the allowable disturbance power. Confirmed. Since the harmonic characteristic residual is the quadrature-axis current residual after stripping the equivalent electrical signal components from the production stage, the torque constant of the take-up reel servo motor is obtained. and the spindle angular velocity during the current winding operation. Based on the relationship between electromagnetic torque and quadrature-axis current, the allowable disturbance power is converted into the allowable residual current. ; the allowable residual current The smaller of the values ​​in the rated output current of the take-up reel servo motor is determined as the preset tolerance upper limit corresponding to the root mean square value of the harmonic characteristic residual amplitude. Therefore, when the current wire specifications, the free length of the suspended section, or the winding operation status change, the preset tolerance upper limit can be re-determined based on the corresponding physical parameters.

[0058] By comparing the dual independent thresholds for the time dimension (number of cycles) and the energy dimension (residual amplitude), the control system can accurately define the temporal persistence and physical destructiveness of the excitation disturbance, ensuring that the timing of the underlying reverse damping torque injection is precise and effective.

[0059] The Lyapunov numerical analysis function is constructed using a combination of a quadratic form and a mechanical dissipation term. Let r(k) be the harmonic characteristic residual obtained in the k-th servo sampling period, and let the servo sampling period be... The first-order time difference of the residual is then... And construct a state variable vector Let the maximum permissible harmonic characteristic residual amplitude of the system be... The maximum permissible rate of change of harmonic characteristic residuals is Construct a positive definite weight matrix The rated mechanical power of the take-up reel servo motor. With servo sampling period The product of and is used as the reference energy. ,Right now Let the mechanical friction dissipation power determined in the j-th sampling period be... Then the Lyapunov numerical analysis function in the kth sampling period is according to Calculate; the energy change rate index of adjacent sampling periods according to Calculation. When When, it indicates that the equivalent disturbance energy corresponding to the harmonic characteristic residual after deducting inherent mechanical dissipation is still increasing; when When the value is zero, it indicates that the equivalent disturbance energy has not continued to accumulate. Therefore, based on r(k) obtained from the continuous servo sampling cycle, the energy change rate index G(k) can be calculated cycle by cycle, and the subsequent cumulative control cycle count and root mean square value of the amplitude can be determined.

[0060] This invention introduces the theory of nonlinear system stability determination. By calculating the energy change rate index and combining it with multi-dimensional tolerance upper limit determination, it can make forward-looking identifications before microscopic disturbances evolve into macroscopic damage, which not only ensures the sensitivity of identification, but also effectively avoids false alarms in the production environment.

[0061] Further, the mechanical friction dissipation power is determined. Specifically, this includes: collecting the current actual winding radius of the take-up reel, the actual output speed of the take-up servo motor, and the preload parameters of the mechanical bearings, and substituting them into a preset mechanical friction dynamics model to determine the transient mechanical friction work generated by the friction of rotating parts, bearing rolling resistance, and air viscous resistance in the current sampling period; dividing the transient mechanical friction work by the corresponding servo sampling period to obtain the mechanical friction dissipation power for that sampling period. And when calculating the energy change rate index, it is subtracted as a negative dissipation term, so that the calculated energy change rate index characterizes the disturbance energy change state after deducting the inherent mechanical dissipation.

[0062] Specifically, in the step of constructing the Lyapunov numerical analysis function for quantitatively representing the energy convergence state of the take-up reel rotation dynamics system, to avoid the judgment error caused by the pure mathematical idealization model ignoring the real damping loss of the physical entity in the industrial field, this embodiment introduces a nonlinear friction dissipation compensation mechanism. The edge computing node structurally expands this Lyapunov numerical analysis function at the bottom layer of the control algorithm, and explicitly introduces a mechanical friction dissipation term that varies with the current real-time winding radius of the take-up reel and the spindle angular velocity during the energy change rate calculation. During equipment operation, the controller reads the current actual physical winding radius of the take-up reel and the actual output speed of the take-up servo motor in real time through the data bus, and retrieves the static parameters of the mechanical bearing preload force fed back by the pressure sensor preset in the take-up bearing end seat during the equipment assembly stage. Using the above-mentioned actual winding radius, actual output speed, and mechanical bearing preload force parameters as independent variables, and substituting them into the dynamic physical equation containing a multi-level friction coefficient matrix, the transient mechanical friction work done by the system under the current transient operating condition due to the combined friction of multi-axis mechanical rotating parts, bearing rolling resistance, and air viscous resistance is calculated online. Subsequently, the work done by the transient mechanical friction is converted into an equivalent energy dissipation rate, which is used as a negative damping term to supplement and cancel out the first-order time derivative of the Lyapunov function, and then the energy change rate index obtained by the solution is corrected.

[0063] This invention supplements the calculation of energy change rate with a mechanical friction dissipation term based on physical entity parameters, making the stability determination model fully consistent with the laws of thermodynamics and conservation of momentum. By treating friction loss as a negative damping term for energy cancellation, it eliminates the technological drawback of misjudging inherent mechanical damping losses as system energy dissipation, improves the accuracy of stability threshold determination, and avoids false alarms or over-control behavior under unsteady-state extreme conditions.

[0064] Furthermore, generating a damping torque command in the opposite direction and injecting it into the underlying servo driver to control the take-up reel servo motor to output the reverse damping torque specifically includes the following steps: extracting the dominant harmonic frequency, instantaneous positive and negative signs, and amplitude of the harmonic characteristic residual when the energy change rate index exceeds the stability threshold; if the dominant harmonic frequency exceeds the safe phase margin bandwidth, mapping the amplitude to the compensation current amplitude based on the motor torque constant, and generating an initial command in combination with the opposite direction of the instantaneous positive and negative signs; after the initial command is limited by a preset maximum allowable current threshold, generating an anti-disturbance quadrature-axis current control command; directly injecting the anti-disturbance quadrature-axis current control command into the current loop control register of the underlying servo driver to drive the take-up reel servo motor to generate an electromagnetic damping torque that tracks the harmonic characteristic residual fluctuations and is in the opposite direction, absorbing the excitation energy transmitted to the axial direction of the take-up reel by the wire reel commutation impact through the electromagnetic damping torque, thereby achieving physical suppression of nonlinear harmonic chattering.

[0065] Specifically, to prevent high-frequency harmonic commands from exceeding the physical bandwidth of the servo motor, which could lead to positive feedback instability or even hardware overcurrent damage, this embodiment imposes physical constraints on the underlying control boundary. First, the dominant harmonic frequency, instantaneous sign, and amplitude of the harmonic characteristic residual at the instant the energy change rate index exceeds the stability threshold are extracted. The dominant harmonic frequency is then compared with the current loop diametrical plot model embedded in the underlying servo driver to verify whether it is within the set safe phase margin bandwidth. If the dominant harmonic frequency exceeds the safe phase margin bandwidth, a protection mechanism is triggered, prohibiting the injection of anti-interference current during the current cycle.

[0066] For frequency bands within the safe bandwidth, a digital phase-advancing compensator is invoked to perform phase-advancing correction on the harmonic characteristic residuals. Simultaneously, based on the motor torque constant, the corrected residual amplitude is linearly mapped to the compensation current amplitude, and combined with the opposite direction of the instantaneous positive and negative signs, an initial anti-disturbance current command is generated. Specifically, firstly, based on the Lyapunov energy dissipation requirements, the target damping torque amplitude required to suppress the current chattering is determined. Subsequently, the edge computing node retrieves the inherent electromagnetic physical parameters of the take-up reel servo motor itself from the configuration parameters of the underlying servo driver, namely the motor torque constant. Based on the linear coupling equation between the electromagnetic torque and quadrature-axis current of the permanent magnet synchronous motor, the target damping torque amplitude is divided by the motor torque constant to accurately map the quadrature-axis compensation current amplitude required to counteract the disturbance. Finally, combining the phase-advancing correction amount processed in the aforementioned steps with the quadrature-axis compensation current amplitude, an anti-disturbance current control command with phase compensation and a determined current ampere amplitude is generated in the current loop control register of the underlying servo driver.

[0067] Subsequently, the initial anti-disturbance current command is compared with the preset maximum allowable current threshold of the driver. The preset maximum allowable current threshold of the driver is calculated by the edge computing node based on the limits of the underlying hardware. Its value is equal to the absolute physical peak current tolerance of the take-up reel servo motor and the driver IGBT module, minus the macroscopic quadrature-axis current reference value necessary to maintain normal production process (i.e., target tension and macroscopic speed). After saturation limiting of the dynamic safety margin, the final anti-disturbance quadrature-axis current control command is output. Then, the final anti-disturbance quadrature-axis current control command is directly overwritten in the current loop control register of the underlying servo driver. Under the premise of satisfying the Nyquist stability criterion and not triggering hardware overcurrent protection, the take-up reel servo motor generates an electromagnetic damping torque that effectively tracks mechanical fluctuations and is in the opposite direction. This safely absorbs the excitation energy transmitted from the wire reel commutation impact to the axial direction of the take-up reel, and achieves physical suppression of nonlinear harmonic chattering.

[0068] This invention constructs a physical-level anti-interference closed loop that directly acts on the motor current loop, eliminating the delay between multi-layer network communication and upper-layer computation. By injecting reverse electromagnetic damping torque in real time, it can eliminate high-frequency dangerous vibrations at the underlying physical actuator with a response level far exceeding that of traditional PID regulation, thereby improving the dynamic stability of the entire flexible cable production equipment.

[0069] Example 2 This embodiment illustrates how cross-domain collaborative control, combining "spatial domain physical deformation prediction" and "electromagnetic domain underlying electromechanical disturbance rejection," addresses the spatial distortion and dynamic tension instability of the wire landing point caused by the extreme reversal at both ends of the flange during the high-speed precision winding and coiling of flexible wires. The specific implementation process is as follows: First, the real-time perception and position feedforward compensation process in the spatial domain is executed. The pulse extreme point of the encoder on the take-up reel is used as the interrupt source for the underlying hardware. When the cable guide wheel moves to a preset angle range close to the flanges at both ends of the cable reel, the pulse extreme point immediately triggers an interrupt, and the control system instantly activates the non-contact three-dimensional optical sensor (such as a binocular machine vision sensor or a laser array displacement sensor) configured at the workstation. In the activated state, the sensor performs a high-frequency local field-of-view scan of the preset range at both ends of the flange, capturing the three-dimensional spatial contour feature data of the contact and collision point between the outermost winding trajectory edge of the cable and the inner wall of the flange, and calculates the physical three-dimensional coordinates of the tangent point online through an edge reconstruction algorithm.

[0070] Simultaneously, the control system acquires the current theoretical pulse coordinates of the wire guide wheel in real time and converts them into theoretical three-dimensional coordinates in the mechanical transmission coordinate system through the transmission chain mapping equation. By constructing a rigid body transformation matrix that aligns the physical coordinate acquisition unit and the wire guiding mechanism, spatial translation and rotation transformations are performed on the acquired physical three-dimensional coordinates to unify them into a unified mechanical transmission coordinate system. Subsequently, the transformed physical three-dimensional coordinates are subtracted from the theoretical three-dimensional coordinates to extract the wire guiding hysteresis pose deviation matrix containing three axial linear deviation components and angular deviation components. The axial and radial position deviation components in this tensor are further extracted, and the dynamic angle between the wire guide wheel axis and the wire centerline is calculated using the arctangent trigonometric function. The angle value is then corrected using the flange geometric boundary parameters at the reversal pole as constraints, thereby locking the maximum geometric angle formed between the wire guide wheel and the winding tangent point.

[0071] Next, mechanical deformation decoupling and feedforward pulse injection are performed. The control system uses the calculated sine of the maximum geometric angle and the target tension currently issued by the system to calculate the lateral component force along the cable reel axis. Within a preset time window, the value of this lateral component force is continuously monitored, and the difference between its maximum and minimum values ​​is extracted as the fluctuation of the lateral component force. Simultaneously, the inherent physical property parameters of the current production material (including elastic modulus and cross-sectional area parameters) are retrieved, and the spatial straight-line distance between the center point of the guide wheel and the winding tangent point is obtained in real time as the current free length of the suspended section of the wire. Based on the material mechanics deformation decoupling model of Hooke's law, the fluctuation of the lateral component force and the free length of the suspended section are combined and vectorized into the additional tensile tension fluctuation along the wire axis. This is then divided by the cross-sectional area to obtain the tensile stress variable. Through linear strain calculation and integral calculation, the elastic deformation elongation of the wire along the longitudinal direction is finally solved.

[0072] After obtaining the elastic deformation elongation, the control system acquires the inherent pulse resolution per revolution of the cable guide wheel servo motor and the pitch parameter of the cable guide screw. The elastic deformation elongation is multiplied by the sine of the maximum geometric angle, projected, and converted into an equivalent displacement deviation along the cable reel axis. This deviation is then divided by the pitch parameter and multiplied by the inherent pulse resolution per revolution to calculate the absolute value of the target pulse used to correct the landing point offset. To overcome the inherent physical response delay of the electromechanical actuator, a preset lead time constant is introduced. The absolute value of the target pulse is then subjected to phase lead prediction processing to generate a position lead compensation pulse quantity with time prediction characteristics. This pulse quantity is directly injected into the underlying position loop controller of the cable guide wheel servo motor in the form of pulse superposition. This drives the cable guide wheel servo motor to perform a lead displacement equivalent to the elastic deformation elongation before the actual mechanical commutation action occurs, thereby offsetting the landing point lag in the spatial domain.

[0073] During the time period of position advance compensation in the spatial domain, the control system simultaneously initiates the underlying anti-disturbance control process in the electromagnetic domain. Because the significant mechanical impact of the wire-laying mechanism during advance displacement compensation and mechanical reversal at its extreme positions is dynamically transmitted to the take-up spindle via the flexible wire, it can cause electromechanical instability on the take-up side. Therefore, a current sensor and voltage conditioning circuit are used to collect the three-phase stator current and back electromotive force signals of the take-up reel servo motor controller in real-time at high frequency, serving as the initial operating electrical signals. Simultaneously, production stage commands (such as acceleration steps, deceleration steps, or wire-laying reversal commands) issued by the upper-level process system of the Industrial Internet are received, and the corresponding macroscopic target torque and speed are input to a preset nonlinear electromagnetic state observer. This observer dynamically simulates the magnetic circuit saturation and inherent nonlinear harmonics inside the motor, and calculates the background electrical signal components generated by the proactive process adjustments, i.e., the equivalent electrical signal components. A filtering algorithm based on dynamic differential reduction or adaptive noise cancellation is adopted to perform high-frequency differential comparison between the initial operating electrical signal and the equivalent electrical signal component, thereby completely eliminating the influence of macroscopic load step and accurately extracting the harmonic characteristic residuals caused by external microscopic excitation.

[0074] Finally, precise stability assessment and energy dissipation control are performed on the harmonic characteristic residual. A Lyapunov numerical analysis function is constructed to quantitatively characterize the energy convergence or divergence state of the entire winding system. The extracted harmonic characteristic residual is input into this function to calculate its first-order time derivative, and the energy change rate index is obtained. When this index is continuously greater than zero, an instability warning is triggered. The time span of the index being continuously greater than zero and the root mean square value of the harmonic characteristic residual amplitude are further compared. When either of them exceeds the preset tolerance upper limit, the energy change rate index is determined to have exceeded the stability threshold. Once the limit is exceeded, the sign of the harmonic characteristic residual, the trend of the rate of change, and its dominant harmonic frequency are instantly captured at the moment of the determination. To prevent the risk of phase lag caused by the bandwidth limitation of the servo motor current loop, the dominant harmonic frequency is compared with the current loop bandwidth of the underlying driver in real time, and a digital lead phase compensator is called to perform phase lead correction on the frequency band with the risk of lag. Finally, in the current loop control register of the underlying servo driver, an anti-disturbance current control command with phase compensation and a sign completely opposite to the residual is generated. This command directly drives the take-up reel servo motor to output a set of dynamic electromagnetic damping torques with opposite directions and magnitudes that track harmonic fluctuations in real time. This absorbs and dissipates the micro-excitation energy transmitted to the take-up spindle by mechanical commutation on-site, thereby achieving high-dimensional stable control of the take-up process in the electromagnetic domain.

[0075] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A control method for flexible cable reel production equipment based on the Industrial Internet, characterized in that, include: Using the pulse extreme point of the encoder on the take-up reel as the interrupt source, the interruption is triggered when the wire guide wheel runs to the preset angle range at both ends of the cable reel flange, and the physical three-dimensional coordinates of the wire winding cutting point are obtained. The theoretical pulse coordinates of the cable guide wheel are mapped to theoretical three-dimensional coordinates. The physical three-dimensional coordinates are transformed to the same coordinate system through rigid body transformation and then subtracted to extract the cable hysteresis pose deviation matrix. The maximum geometric angle formed by the cable guide wheel and the winding tangent point is calculated. The lateral force along the cable reel axis is calculated based on the maximum geometric angle and the target tension, and the fluctuation of the lateral force is extracted. Based on the fluctuation amount, the free length of the suspended section of the wire and physical property parameters, the elastic deformation elongation of the wire is calculated and converted into the position advance compensation pulse amount, and the servo motor of the wire guide wheel is controlled to perform advance compensation. The system continuously collects the operating electrical signals of the take-up reel servo motor, filters out the equivalent electrical signal components generated by production stage commands, extracts harmonic characteristic residuals, and extracts the energy change rate index through stability calculation. When the energy change rate index exceeds the stability threshold, a reverse damping torque command is injected into the underlying servo driver to drive the take-up reel servo motor to output reverse damping torque.

2. The control method for flexible cable reel production equipment based on the Industrial Internet according to claim 1, characterized in that, The process of acquiring the physical three-dimensional coordinates of the wire winding tangent point includes: configuring a non-contact three-dimensional optical sensor as a physical coordinate acquisition unit; when an interrupt source is triggered, the physical coordinate acquisition unit switches to the active working state and performs a local field-of-view scan within a preset angle range at both ends of the cable reel flange; acquiring the three-dimensional spatial contour feature data of the contact point between the edge of the wire winding trajectory and the inner wall of the flange; and using an edge reconstruction algorithm to process the three-dimensional spatial contour feature data to calculate the physical three-dimensional coordinates of the wire winding tangent point in the basic coordinate system of the physical coordinate acquisition unit.

3. The control method for flexible cable reel production equipment based on the Industrial Internet according to claim 1, characterized in that, Extracting the hysteresis pose deviation matrix for cable laying specifically includes: constructing a rigid body transformation matrix for aligning the physical coordinate acquisition unit and the cable laying mechanism; performing spatial translation and rotation transformations on the physical three-dimensional coordinates using the rigid body transformation matrix to unify the physical three-dimensional coordinates to the mechanical transmission coordinate system of the cable laying mechanism; subtracting the translated and rotated physical three-dimensional coordinates from the theoretical three-dimensional coordinates to obtain the linear deviation components and angular deviation components in the three axes of the spatial rectangular coordinate system, and combining them to form the hysteresis pose deviation matrix for cable laying that characterizes the direction and amplitude of spatial geometric distortion.

4. The control method for flexible cable reel production equipment based on the Industrial Internet according to claim 1, characterized in that, The maximum geometric angle formed by the cable guide wheel and the winding tangent point is calculated. Based on the maximum geometric angle and the target tension, the lateral force along the cable reel axis is calculated. Specifically, this includes: extracting the axial and radial position deviation components from the cable hysteresis pose deviation matrix; calculating the arctangent trigonometric function values ​​of the axial and radial position deviation components to obtain the dynamic angle between the cable guide wheel axis and the cable centerline; using the flange geometric boundary parameters at the cable reel's reversal pole as boundary constraints to correct the dynamic angle value and determine the maximum geometric angle formed by the line connecting the cable guide wheel and the winding tangent point at the moment of flange edge reversal; obtaining the target tension issued by the upper-level process system of the Industrial Internet, and using a sine function to calculate the component of the target tension in the direction of the maximum geometric angle, which is defined as the lateral force along the cable reel axis.

5. The control method for flexible cable reel production equipment based on the Industrial Internet according to claim 1, characterized in that, The calculation of the elastic deformation elongation of the wire includes: continuously monitoring the calculated lateral force values ​​within a preset time window, extracting the difference between the maximum and minimum values ​​as the fluctuation of the lateral force; retrieving the elastic modulus and cross-sectional area parameters of the current wire, and obtaining the spatial straight-line distance between the center point of the current wire guide wheel and the winding tangent point as the free length of the suspended section; using the free length of the suspended section and the maximum geometric angle, vectorizing and converting the fluctuation of the lateral force into the additional tensile tension fluctuation along the wire axis; dividing the additional tensile tension fluctuation by the cross-sectional area to obtain the tensile stress variable, combining it with the elastic modulus to perform linear strain calculation, and integrating to obtain the elastic deformation elongation of the wire along the longitudinal direction.

6. The control method for flexible cable reel production equipment based on the Industrial Internet according to claim 1, characterized in that, The pulse quantity is converted into a position advance compensation pulse quantity, which controls the servo motor of the cable guide wheel to perform advance compensation. Specifically, this includes: obtaining the inherent pulse resolution per revolution of the servo motor of the cable guide wheel and the pitch parameter of the cable screw; multiplying the elastic deformation elongation by the sine value of the maximum geometric angle and projecting it into an equivalent displacement deviation along the axial direction of the cable reel; dividing the equivalent displacement deviation by the pitch parameter and multiplying it by the inherent pulse resolution per revolution to calculate the absolute value of the target pulse; performing phase advance prediction on the absolute value of the target pulse based on a preset advance time constant to generate a position advance compensation pulse quantity with time prediction characteristics, and injecting it into the underlying position loop controller of the servo motor of the cable guide wheel. This controls the servo motor to perform an equivalent advance displacement in advance before mechanical commutation, thus offsetting the winding position lag caused by the elastic deformation of the wire.

7. The control method for flexible cable reel production equipment based on the Industrial Internet according to claim 1, characterized in that, The extraction of harmonic characteristic residuals specifically includes: acquiring the three-phase stator current signal and back electromotive force signal of the take-up reel servo motor controller as the initial operating electrical signal; receiving production stage instructions issued by the upper-level process system of the Industrial Internet; inputting the macroscopic target torque and speed corresponding to the production stage instructions into a preset nonlinear electromagnetic state observer, and calculating the equivalent electrical signal components including macroscopic load step and inherent nonlinear harmonics inside the motor; performing high-frequency differential comparison between the initial operating electrical signal and the equivalent electrical signal components, stripping the background electrical signal, and extracting the harmonic characteristic residuals.

8. The control method for flexible cable reel production equipment based on the Industrial Internet according to claim 1, characterized in that, The stability calculation of the harmonic characteristic residuals specifically includes: constructing a Lyapunov numerical analysis function to quantitatively represent the energy convergence state of the accumulator disk rotation dynamics system; inputting the harmonic characteristic residuals as state variables into the Lyapunov numerical analysis function, and obtaining the energy change rate index by calculating the first-order time derivative of the state variables; if the energy change rate index is greater than zero for a consecutive preset number of servo sampling periods, an over-limit warning is triggered; comparing the cumulative number of control cycles with the energy change rate index being greater than zero with the root mean square value of the amplitude of the harmonic characteristic residuals, when either the cumulative number of control cycles or the root mean square value of the amplitude exceeds the corresponding preset tolerance upper limit, it is determined that the energy change rate index exceeds the stability threshold.

9. The control method for flexible cable reel production equipment based on the Industrial Internet according to claim 1, characterized in that, The process of generating a damping torque command in the opposite direction and injecting it into the underlying servo driver to control the take-up reel servo motor to output the reverse damping torque includes the following steps: extracting the dominant harmonic frequency, instantaneous sign, and amplitude of the harmonic characteristic residual when the energy change rate index exceeds the stability threshold; if the dominant harmonic frequency exceeds the safe phase margin bandwidth, mapping the amplitude to the compensation current amplitude based on the motor torque constant, and generating an initial command in combination with the opposite direction of the instantaneous sign; after the initial command is limited by a preset maximum allowable current threshold, generating an anti-disturbance quadrature-axis current control command; directly injecting the anti-disturbance quadrature-axis current control command into the current loop control register of the underlying servo driver to drive the take-up reel servo motor to generate an electromagnetic damping torque that tracks the harmonic characteristic residual fluctuations and is in the opposite direction, absorbing the excitation energy transmitted from the wire commutation impact to the axial direction of the take-up reel through the electromagnetic damping torque, thereby achieving physical suppression of nonlinear harmonic chattering.

Images