Method for controlling the winding speed of a silk ribbon

Abstract

This invention discloses a method for synchronous control of the winding speed of a silk webbing, comprising: using a dynamically corrected eccentric disturbance filtering boundary as a constraint, employing a Kalman filter algorithm to fuse the effective signals below the filtering boundary in the webbing tension and radial runout trajectory to obtain a tension balance signal after filtering out eccentric disturbances; extracting the belt traction coordination relationship between the winding end and the unwinding end based on the tension balance signal after filtering out eccentric disturbances, processing the belt traction coordination relationship through a proportional-integral-derivative control algorithm to determine the synchronous control command; performing frequency domain analysis on the synchronous control command to identify residual eccentric correlation oscillation ripples, filtering out the oscillation ripples based on the rotation period characteristics to obtain a stable dual-end tension balance control command.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for synchronous control of the winding speed of silk ribbon. Background Technology

[0002] The webbing winding process is a crucial step in textile finishing equipment, directly determining the smoothness, winding tightness, and appearance quality of the finished webbing. In high-speed continuous production, it is essential to maintain a high degree of consistency between the unwinding and winding speeds while keeping the webbing tension constant. The stability of this synchronous control becomes a core factor affecting production efficiency and product quality. Currently, most control methods tend to calculate the rotational inertia of the unwinding end in real time based on changes in roll diameter and adjust the motor torque or speed setpoint accordingly to achieve dynamic matching of the speeds at both ends.

[0003] For example, a prior art webbing winding device and method, disclosed in CN118597864A, provides a webbing winding technology including a transmission mechanism, a tape feeding mechanism, and a winding mechanism. This technology achieves fully automatic webbing winding through automated cutting, tape application, and stacking mechanisms. However, it suffers from a problem where the periodic disturbance caused by the core tube's eccentricity amplifies as the unwinding diameter decreases, leading to severe tension fluctuations and speed synchronization instability. When the unwinding diameter gradually decreases from full roll to near the core tube, the moment of inertia significantly decreases, and the response speed of the unwinding end to speed commands increases markedly. This faster response should ideally help to quickly follow any small fluctuations in the winding end speed, but in reality, the core tube itself inevitably has a certain degree of eccentricity. When the roll diameter decreases, the same amount of eccentricity produces a larger radial runout on the roll surface. This runout occurs repeatedly with a cycle of one revolution of the core tube, forming a significant periodic disturbance. The reduced inertia makes the unwinding end exceptionally sensitive to this periodic radial runout, and the disturbance signal is transmitted to the webbing tension with almost no attenuation, and further coupled to the speed control loop at the winding end.

[0004] The result is a frequency mismatch and mutual chasing at both ends of the speed. The unwinding end attempts to quickly compensate for the sudden change in tension, while the winding end adjusts its output due to the fluctuations, creating a closed-loop oscillation. This causes severe tension fluctuations, resulting in uneven tension, wrinkles, and even breakage of the webbing during winding, seriously affecting the stability of continuous production. This disturbance amplification effect caused by eccentricity is most significant, with the largest oscillation amplitude and the most prone to instability, especially when the roll diameter is close to the core tube and the inertia is at its minimum.

[0005] Therefore, under the condition of continuously decreasing unwinding diameter and significantly reduced moment of inertia, how to effectively suppress the transmission of periodic radial runout caused by core tube eccentricity to the tension and speed synchronization control system, and avoid frequency mismatch coupling oscillations caused by excessively fast response at both ends, has become a key issue in ensuring the stability of tension balance and speed synchronization during high-speed winding of the woven tape. Summary of the Invention

[0006] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a synchronous control method for the winding speed of silk ribbon, which solves the problem that in the existing technology, when the winding diameter is close to the core tube and the inertia is the smallest, the disturbance amplification effect caused by eccentricity is the most significant, the oscillation amplitude is the largest, and the synchronization relationship is most prone to instability.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] This invention provides a method for synchronously controlling the winding speed of a silk ribbon, comprising:

[0009] The surface of the unwinding end core tube is continuously scanned by a laser rangefinder to obtain the current roll diameter and webbing tension. Combined with the rotation phase angle collected by the angle encoder, the periodic jumping profile of the roll diameter as the rotation phase angle changes is identified.

[0010] Based on the periodic jumping profile, the radial jumping trajectory and rotation period features caused by eccentricity are extracted. Combined with the current roll diameter, the evolution law of the radial jumping trajectory under the roll diameter decreasing state is analyzed to obtain the dynamically corrected eccentricity disturbance filtering boundary.

[0011] Using the dynamically corrected eccentric disturbance filtering boundary as a constraint, the Kalman filter algorithm is used to fuse the available signals below the eccentric disturbance filtering boundary in the webbing tension and the radial runout trajectory to obtain the tension balance signal after filtering out the eccentric disturbance.

[0012] Based on the tension balance signal after filtering out eccentric disturbances, the belt traction coordination relationship between the take-up end and the unwinding end is extracted, and the belt traction coordination relationship is processed by the proportional-integral-derivative control algorithm to determine the synchronous control command.

[0013] Frequency domain analysis is performed on the synchronization control command to identify residual eccentric correlation oscillation ripple in the synchronization control command. The residual eccentric correlation oscillation ripple is filtered out using the rotation period characteristic as a criterion to obtain a stable dual-end tension balance control command.

[0014] The stable dual-end tension balance control command is applied to the drive motor of the winding system, and the real-time roll diameter and real-time webbing tension are acquired. The fluctuation anomalies of the real-time roll diameter and real-time webbing tension are identified and used as iterative feedback, which is then sent back to the roll diameter scanning stage to form a closed-loop control to achieve synchronous and stable winding speed.

[0015] Furthermore, the step of continuously scanning the surface of the unwinding end core tube with a laser rangefinder to obtain the current roll diameter and webbing tension, and combining this with the rotational phase angle acquired by the angle encoder disk to identify the periodic jump profile of the roll diameter as the rotational phase angle changes, includes:

[0016] The real-time roll diameter value is obtained by converting the return time interval of the laser rangefinder to acquire distance sequence data.

[0017] The rotation phase angle determined by the pulse signal of the angle encoder disk is timestamped with the real-time roll diameter value to form a phase roll diameter correspondence table;

[0018] The extreme value positions in the phase roll diameter correspondence table are extracted and arranged to form the periodic jumping profile.

[0019] Furthermore, the step of extracting the radial jump trajectory and rotation period features caused by eccentricity based on the periodic jump contour, and combining this with the current roll diameter analysis to understand the evolution of the radial jump trajectory under the roll diameter decreasing state, yields a dynamically corrected eccentricity disturbance filtering boundary, including:

[0020] Extract the peak and valley positions of the periodic jumping profile, determine the radial jumping amplitude and fluctuation period caused by eccentricity, and use the fluctuation period as the rotation period feature;

[0021] The current roll diameter is compared with the full roll diameter value, and the roll diameter reduction stage is determined to be a small roll diameter stage or a large roll diameter stage based on the difference ratio.

[0022] The radial runout amplitudes of the small roll diameter stage and the large roll diameter stage are recorded separately. The difference in runout amplitude is compared and a mapping curve is established based on the decreasing trend of the difference with the roll diameter value, which serves as the evolution law of the runout trajectory.

[0023] Based on the mapping curve, the upper limit of the jump amplitude is found for the current roll diameter as the boundary threshold. The boundary threshold is synchronously updated within the phase angle range corresponding to the rotation period feature to obtain the dynamically corrected eccentric disturbance filtering boundary.

[0024] Furthermore, the step of using the dynamically corrected eccentricity disturbance filtering boundary as a constraint, and employing a Kalman filter algorithm to fuse the available signals below the eccentricity disturbance filtering boundary in the webbing tension and the radial runout trajectory to obtain the tension balance signal after filtering out the eccentricity disturbance, includes:

[0025] The tension fluctuation component is compared with the radial runout amplitude, and usable signals below the eccentric disturbance filtering boundary are retained;

[0026] By combining the amplitude and instantaneous velocity of the available signals to form a tension state vector, Kalman filtering is performed to estimate the state and obtain the optimal estimate.

[0027] The amplitude components of the optimal estimate are connected to form the tension balance signal after filtering out the eccentric disturbance.

[0028] Furthermore, based on the tension balance signal after filtering out eccentric disturbances, the tape traction coordination relationship between the take-up end and the unwinding end is extracted, and the tape traction coordination relationship is processed by a proportional-integral-derivative control algorithm to determine the synchronization control command, including:

[0029] The deviation characteristics of the conveyor belt traction coordination relationship are extracted, and the proportional-integral-derivative control algorithm is used to generate instantaneous correction, integral accumulation and advance compensation.

[0030] The combined control output value is then converted into speed regulation and torque following values ​​to generate the synchronous control command.

[0031] Furthermore, the combined control output value is converted into a speed adjustment amount and a torque following amount to generate the synchronous control command, including:

[0032] Extract the instantaneous deviation value at the moment when the belt speeds at both ends are the same, determine the synchronization and coordination relationship of the belt speeds at both ends based on the fluctuation amplitude and fluctuation direction of the instantaneous deviation value, and record the time difference between the issuance of the command and the change of the response as the drive response delay time.

[0033] The instantaneous deviation values ​​at different webbing travel distances are statistically analyzed. Positive values ​​are marked as webbing tightness and negative values ​​are marked as webbing looseness. The alternation of these states is taken as an indication of alternating tightness and looseness.

[0034] Based on the duration of the alternating tension and slack behavior and the drive response delay, deceleration adjustment is determined in the slack state and acceleration adjustment is determined in the slack state. The speed adjustment range of the speed adjustment amount is determined according to the product relationship. The state switching time interval is used as the adjustment execution cycle to generate the synchronous control command.

[0035] Furthermore, the step of performing frequency domain analysis on the synchronization control command to identify residual eccentric correlation oscillation ripple in the synchronization control command, and filtering out the residual eccentric correlation oscillation ripple using the rotation period characteristic as a criterion to obtain a stable dual-end tension balance control command includes:

[0036] The significant frequency components of the synchronization control command are extracted using Fourier transform;

[0037] The rotation periodicity feature is used to select the matching frequency as the residual eccentric correlated oscillation ripple;

[0038] The waveform obtained after suppression by a band-stop filter and inverse Fourier transform is used as the stable dual-end tension balance control command.

[0039] Furthermore, the step of applying the stable dual-end tension balance control command to the drive motor of the winding system, acquiring the real-time roll diameter and real-time webbing tension, identifying fluctuations in the real-time roll diameter and real-time webbing tension and using these as iterative feedback, and transmitting this feedback back to the roll diameter scanning stage, forms a closed-loop control to achieve synchronous and stable winding speed, including:

[0040] The stable dual-end tension balance control command is output to execute the speed change action, and the paired real-time roll diameter and the real-time webbing tension are recorded to form roll diameter tension timing data;

[0041] When the continuous change in the roll diameter tension time series data exceeds the fluctuation threshold, an abnormal feedback information is generated and transmitted back.

[0042] Adjust the sampling frequency of the laser rangefinder or the phase synchronization calibration value of the angle encoder disk and re-perform the roll diameter scan.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] This invention discloses a method for synchronous control of the winding speed of woven tape. Addressing the unique business scenario problem of periodic fluctuations in the roll diameter caused by core tube eccentricity during winding and unwinding, leading to a mismatch between tape tension and travel speed, the method acquires the fluctuation profile using laser scanning and an encoder disk, dynamically analyzes its evolution, generates filtering boundaries to constrain Kalman filtering, thereby separating an effective tension balance signal. Based on this signal, the travel coordination relationship is extracted and control commands are generated. Further filtering of residual eccentric oscillation ripple in the commands results in a stable dual-end tension balance control command driving the winding motor. Real-time data feedback forms a closed loop, achieving precise synchronization of winding speed and stable tension under high dynamic conditions. Attached Figure Description

[0045] Figure 1 This is a flowchart of a method for synchronously controlling the winding speed of a silk ribbon according to the present invention.

[0046] Figure 2 This is a schematic diagram of a synchronous control method for the winding speed of a silk ribbon according to the present invention.

[0047] Figure 3 This is another schematic diagram of a synchronous control method for the winding speed of a silk ribbon according to the present invention. Detailed Implementation

[0048] The present invention will now be described in detail through specific embodiments:

[0049] like Figures 1-3 The specific method for synchronous control of the winding speed of a silk ribbon in this embodiment may include:

[0050] Step S101: The surface of the unwinding end core tube is continuously scanned by the laser rangefinder to obtain the current roll diameter and webbing tension. Combined with the rotation phase angle collected by the angle encoder, the periodic jumping profile of the roll diameter as the rotation phase angle changes is identified.

[0051] A laser beam is emitted radially along the surface of the unwinding core tube using a laser rangefinder. The time interval from emission to return of the laser beam is collected. The instantaneous distance between the laser rangefinder and the core tube surface is calculated by dividing the product of the light speed and the time interval by two. This instantaneous distance value is continuously acquired at a fixed sampling frequency during the core tube's rotation, forming a distance sequence data covering the circumference of the core tube. The real-time roll diameter is obtained based on the difference between this distance sequence data and the laser rangefinder's mounting reference position. Simultaneously, a tension sensor collects the tension detection value of the webbing at the unwinding end. The pulse signal output from the angle encoder is acquired. Based on the ratio between the rising edge count of the pulse signal and the total number of graduations on the encoder, the current rotational phase angle of the core tube is determined. The rotational phase angle is timestamped and aligned with the corresponding real-time roll diameter value to establish a mapping relationship between the rotational phase angle and the roll diameter value, forming a phase-to-roll diameter correspondence table with the rotational phase angle as the horizontal axis and the roll diameter value as the vertical axis. Based on the distribution pattern of the roll diameter values ​​as a function of the rotation phase angle in the phase roll diameter correspondence table, the maximum and minimum values ​​of the roll diameter values ​​within a complete rotation cycle are extracted. For the relative deviation between the roll diameter value corresponding to each rotation phase angle and the maximum and minimum values, they are arranged sequentially along the rotation phase angle direction to form a periodic jumping profile describing the eccentricity characteristics of the core tube.

[0052] In one embodiment, a laser rangefinder is mounted on the unwinding end frame, with its emitting end fixed radially toward the core tube surface. The laser beam is emitted at an angle perpendicular to the core tube axis. After the laser beam reaches the core tube surface, it undergoes diffuse reflection, and part of the reflected light returns along the original path to the receiving end of the laser rangefinder. A photoelectric sensor inside the rangefinder records the complete time interval from the laser beam's emission to its return. Multiplying this time interval by the speed of light and dividing by two yields the instantaneous distance between the laser rangefinder and the core tube surface.

[0053] Specifically, during the continuous rotation of the core tube, the laser rangefinder continuously collects instantaneous distance values ​​at a fixed sampling frequency. The sampling frequency is selected to match the rotational speed of the core tube, ensuring that a sufficient number of distance sampling points are collected for each revolution of the core tube. All sampling points are arranged in chronological order to form a distance sequence data covering the circumference of the core tube. This distance sequence data reflects the radial distance distribution of each position on the surface of the core tube relative to the laser rangefinder.

[0054] For example, the angle encoder disk is coaxially mounted with the core tube, and the core tube rotates, causing the encoder disk to rotate synchronously. Several lines are evenly etched on the circumference of the encoder disk. The photoelectric reading head detects the changes in brightness as the lines pass through and outputs corresponding pulse signals. The rising edge of each pulse signal corresponds to the encoder disk rotating through one scale unit. By accumulating the rising edge count of the pulse signals and comparing this count with the total number of scale divisions on the encoder disk, the current rotational phase angle of the core tube can be obtained. The rotational phase angle ranges from 0 degrees to 360 degrees across the entire circumference.

[0055] In one embodiment, the rotation phase angle is timestamped with the real-time roll diameter value collected at the corresponding moment. Timestamping means that the moment when the laser rangefinder collects the instantaneous distance value is recorded synchronously with the moment when the angle encoder outputs the pulse signal, so that each roll diameter value can correspond to a uniquely determined rotation phase angle.

[0056] It should be noted that, based on the timestamp-aligned data, a phase-to-circuit diameter correspondence table is established with the rotation phase angle as the horizontal axis and the roll diameter value as the vertical axis. In this table, the roll diameter value exhibits a periodic fluctuation characteristic as the rotation phase angle changes. This fluctuation originates from the eccentricity of the core tube itself. By traversing all roll diameter values ​​within a complete rotation cycle, the maximum and minimum values ​​are extracted. The maximum value corresponds to the phase angle where the core tube's eccentricity direction is towards the laser rangefinder, and the minimum value corresponds to the phase angle where the core tube's eccentricity direction is away from the laser rangefinder. For each rotation phase angle, the relative deviation of the roll diameter value from the maximum and minimum values ​​is calculated. These deviations are then arranged sequentially along the rotation phase angle direction, forming a periodic jumping profile describing the core tube's eccentricity characteristics. This profile repeats with each full rotation of the core tube, providing fundamental data for subsequent analysis of eccentricity disturbances.

[0057] Step S102: Extract the radial jumping trajectory and rotation period features caused by eccentricity based on the periodic jumping contour, and analyze the evolution law of the jumping trajectory under the decreasing roll diameter state in combination with the current roll diameter to obtain the dynamically corrected eccentricity disturbance filtering boundary.

[0058] Based on the fluctuation distribution of the roll diameter value with the rotation phase angle in the periodic jumping profile, the peak and trough positions of the fluctuation curve are extracted. The radial jump amplitude caused by eccentricity is determined based on the amplitude difference between the peak and trough positions. The fluctuation period of the radial jump trajectory is determined based on the phase angle interval between the peak and trough positions, and the fluctuation period is used as the rotation period feature. The real-time roll diameter value at the current moment is obtained and compared with the full roll diameter value at the initial unwinding moment. The current roll diameter decreasing stage is determined based on the ratio of the difference between the two to the full roll diameter value. If the ratio exceeds a preset decreasing threshold, it is determined to be a small roll diameter stage; if the ratio is lower than the preset decreasing threshold, it is determined to be a large roll diameter stage. For the roll diameter decreasing stage, the radial jump amplitude is recorded in both the small and large roll diameter stages. The difference in radial jump amplitude between the two stages is compared. Based on the changing trend of the difference with the decreasing roll diameter value, a mapping curve between the jump amplitude and the roll diameter value is established as the evolution law of the jump trajectory. Based on the mapping curve in the evolution law of the jumping trajectory, the upper limit of the corresponding jumping amplitude is found for the current real-time roll diameter value. The upper limit of the jumping amplitude is used as the boundary threshold to distinguish between the eccentric disturbance signal and the effective tension signal. The boundary threshold is updated synchronously within the phase angle range corresponding to each rotation cycle feature to obtain the dynamically corrected eccentric disturbance filtering boundary.

[0059] In one implementation, the periodic jumping profile exhibits a fluctuating distribution pattern with the rotation phase angle as the horizontal axis and the roll diameter value as the vertical axis. This fluctuating distribution contains several local maxima and local minima. By traversing the roll diameter value sequence throughout the entire rotation period, all maxima and minima that meet the conditions are identified through point-by-point comparison. The maxima correspond to the position where the core tube's eccentricity direction is towards the ranging sensor, and the minima correspond to the position where the core tube's eccentricity direction is away from the ranging sensor.

[0060] Specifically, the extraction of peak and trough positions follows these principles: For any sampling point in the roll diameter value sequence, if the roll diameter value of that sampling point is simultaneously greater than the roll diameter values ​​of its two adjacent sampling points, then that sampling point is determined to be a peak position; if the roll diameter value of that sampling point is simultaneously less than the roll diameter values ​​of its two adjacent sampling points, then that sampling point is determined to be a trough position. Within a complete rotation cycle, due to the eccentricity of the core tube, there is usually a main peak and a main trough value. The difference in roll diameter values ​​between the two is the radial runout amplitude, which directly reflects the degree of influence of the core tube eccentricity on the webbing path under the current roll diameter conditions.

[0061] It should be noted that the extraction of rotational periodicity features is based on the phase angle interval between peak and valley positions. Since the radial runout caused by the eccentricity of the core tube is repeated in one revolution of the core tube, the phase angle interval between two adjacent peak positions should theoretically be close to 360 degrees. This interval value is used as the quantitative representation of the rotational periodicity features.

[0062] In one embodiment, the determination of the diameter reduction stage is based on the ratio between the real-time roll diameter value and the full roll diameter value. At the start of the unwinding process, the core tube is in a full roll state. As the webbing continues to be unwound, the roll diameter gradually decreases. The reduction in roll diameter is obtained by subtracting the full roll diameter value from the current real-time roll diameter value, and then divided by the full roll diameter value to obtain the roll diameter reduction ratio.

[0063] For example, the preset reduction threshold is usually set as a certain fixed proportion of the full roll diameter. When the roll diameter reduction proportion exceeds the reduction threshold, it indicates that the core tube has entered the small roll diameter stage. At this time, the roll diameter is close to the core tube body diameter, and the rotational inertia is significantly reduced. When the roll diameter reduction proportion is lower than the reduction threshold, it indicates that the core tube is still in the large roll diameter stage. The roll diameter is much larger than the core tube body diameter, and the rotational inertia is relatively high.

[0064] Understandably, establishing the evolution law of the runout trajectory is the core of this technical solution. In the large diameter stage, the webbing layer on the core tube surface is thicker, and the radial runout amplitude generated when the same core tube eccentricity is reflected on the outer surface is relatively small. This is because the webbing layer itself has a certain degree of uniformity in coverage, which can buffer the geometric deviation caused by eccentricity to a certain extent. However, in the small diameter stage, the webbing layer gradually thins until it approaches the core tube body, and the radial runout amplitude generated when the same core tube eccentricity is reflected on the outer surface increases significantly, amplifying the geometric influence of eccentricity. Therefore, recording the radial runout amplitude at different diameter reduction stages and comparing the amplitude differences between the two stages can reveal the trend of the runout amplitude increasing as the diameter decreases. Based on this trend, a mapping curve is established between the two, with the diameter as the independent variable and the radial runout amplitude as the dependent variable. This mapping curve is the quantitative expression of the runout trajectory evolution law. Furthermore, the mapping curve is constructed using a multi-point fitting method. Radial runout amplitudes were collected at various roll diameters during the unwinding process, forming several sets of data points corresponding to the roll diameter values ​​and runout amplitudes. By performing curve fitting on these data points, a smooth and continuous mapping curve was obtained. This mapping curve can reflect the expected runout amplitude under any roll diameter condition, providing a basis for subsequently determining the filtering boundary.

[0065] In one possible implementation, the determination of the eccentricity disturbance filtering boundary is based on a lookup operation of the mapping curve. For the real-time roll diameter value at the current moment, the corresponding fluctuation amplitude is found on the mapping curve, and this fluctuation amplitude is used as the upper limit of the fluctuation amplitude. During the webbing tension detection process, any tension fluctuation signal that is associated with the rotation period characteristics and whose amplitude does not exceed the upper limit of the fluctuation amplitude is judged as a disturbance signal caused by eccentricity rather than a real tension anomaly signal. Therefore, the upper limit of the fluctuation amplitude becomes the dividing threshold for distinguishing between eccentricity disturbance signals and effective tension signals.

[0066] Preferably, the boundary threshold is not fixed but dynamically updated as the roll diameter continuously decreases. Within the phase angle range corresponding to each rotation cycle feature, the mapping curve is re-found based on the latest acquired real-time roll diameter value to obtain the updated upper limit of the fluctuation amplitude, which is then used as the new boundary threshold. This synchronous update mechanism enables the eccentricity disturbance filtering boundary to adaptively track changes in roll diameter, forming a dynamically corrected eccentricity disturbance filtering boundary, thereby maintaining accurate identification and filtering capabilities for eccentricity disturbances throughout the unwinding process.

[0067] Step S103: Using the dynamically corrected eccentric disturbance filtering boundary as a constraint, the Kalman filter algorithm is used to fuse the effective signals below the filtering boundary in the webbing tension and radial jump trajectory to obtain the tension balance signal after filtering out the eccentric disturbance.

[0068] The current webbing tension detection value and radial runout trajectory data are acquired. The tension fluctuation component is obtained by subtracting the time series mean from the webbing tension detection value. The amplitude of the tension fluctuation component is compared with the runout amplitude at the corresponding phase angle position in the radial runout trajectory. If the amplitude of the tension fluctuation component is lower than the dynamically corrected eccentric disturbance filtering boundary, it is marked as a valid tension fluctuation signal. If the amplitude of the tension fluctuation component is higher than or equal to the filtering boundary, it is marked as an eccentric disturbance signal and discarded. For the valid tension fluctuation signal, the instantaneous rate of tension change is determined by dividing the difference between the valid tension fluctuation signals between adjacent sampling times by the sampling time interval. The amplitude and instantaneous rate of the valid tension fluctuation signal are arranged in order to form a tension state vector to be fused. A Kalman filter algorithm is used to estimate the tension state vector. The tension state estimate from the previous moment is used as the prior prediction for the current moment, and the tension state vector at the current moment is used as the observed value. The Kalman gain is determined based on the ratio of the prediction error covariance of the prior prediction to the observation noise covariance. The Kalman gain is used to weight and correct the deviation between the prior prediction and the observed value, resulting in the optimal tension state estimate for the current moment. The amplitude component is extracted from the optimal tension state estimate, and the amplitude components from multiple consecutive sampling moments are concatenated in chronological order to form a tension balance signal after filtering out eccentric disturbances.

[0069] In one embodiment, the webbing tension detection value is continuously acquired by a tension sensor installed at the unwinding end, and the detection value exhibits fluctuation characteristics over time. The tension fluctuation component is extracted from the webbing tension detection value by selecting continuous tension detection values ​​within a time window, calculating the arithmetic mean of all detection values ​​within the time window as the time series mean, and subtracting the time series mean from the tension detection value at each moment to obtain the tension fluctuation component at the corresponding moment.

[0070] Specifically, distinguishing between tension fluctuation components and eccentric disturbance signals is a crucial step in signal separation. For the tension fluctuation component at each sampling time, a dynamically corrected eccentric disturbance filtering boundary under the current roll diameter condition needs to be obtained. This boundary value B is dynamically calculated using the formula B=k×D+C, where D is the real-time roll diameter, k is the eccentric gain coefficient, and C is the basic amplitude constant. If the amplitude of the tension fluctuation component is lower than the boundary B, it indicates that the fluctuation originates from the actual tension change of the webbing, and it is marked as a valid signal and retained; if the amplitude is higher than or equal to the boundary B, it is determined to be a disturbance signal caused by core tube eccentricity and is removed from subsequent processing.

[0071] It should be noted that the instantaneous rate is calculated based on the change of the effective tension fluctuation signal between adjacent sampling times. The instantaneous rate of tension change is obtained by subtracting the effective tension fluctuation signal amplitude from the amplitude of the effective tension fluctuation signal at the previous sampling time from the amplitude of the effective tension fluctuation signal at the current sampling time, and then dividing by the time interval between the two sampling times.

[0072] In one embodiment, the tension state vector is constructed by arranging the amplitude and instantaneous rate of the effective tension fluctuation signal in a fixed order to form a two-dimensional vector. The first component of this two-dimensional vector is the amplitude of the effective tension fluctuation signal, and the second component is the instantaneous rate; together, they describe the state characteristics of the webbing tension at the current moment.

[0073] For example, the Kalman filter algorithm is a recursive state estimation method. Its core idea is to obtain the optimal estimate of the state by fusing predicted information and observed information. In the scenario of webbing tension processing, the Kalman filter algorithm uses the tension state estimated in the previous moment as prior knowledge for the current moment, and at the same time uses the tension state vector actually collected and constructed in the current moment as observed information. The influence of random noise is eliminated by weighted fusion of the two.

[0074] Understandably, the acquisition of prior predictions is based on the assumption of state transition. During the continuous unwinding of the webbing, the change in tension state between two adjacent sampling moments exhibits a certain continuity. Therefore, the optimal estimate of the tension state at the previous moment is directly used as the prior prediction for the current moment. The tension state vector collected and constructed at the current moment serves as the observation value, which contains true tension change information but may also be contaminated with measurement noise. Furthermore, determining the Kalman gain is a core computational step in the Kalman filtering algorithm. The Kalman gain reflects the weight that should be assigned to the observation value when fusing prior predictions and observations. Its determination is based on the ratio of the prediction error covariance to the observation noise covariance: the prediction error covariance characterizes the potential deviation between the prior prediction and the true state, while the observation noise covariance characterizes the degree to which the observation value is contaminated by measurement noise. When the prediction error covariance is relatively large, it indicates that the reliability of the prior prediction is low. In this case, the Kalman gain should be large, meaning that we should rely more on the observed values. When the observation noise covariance is relatively large, it indicates that the reliability of the observed values ​​is low. In this case, the Kalman gain should be small, meaning that we should rely more on the prior prediction values.

[0075] Preferably, the weighted correction process adjusts the prior prediction value based on the Kalman gain. The deviation between the observed value and the prior prediction value is calculated, multiplied by the Kalman gain to obtain the correction amount, and then the correction amount is added to the prior prediction value to obtain the optimal estimate of the tension state at the current moment. This optimal estimate comprehensively considers the evolution of historical states and the actual observation at the current moment, and can retain the true tension change information while filtering out random noise.

[0076] In one possible implementation, the amplitude component is extracted from the optimal estimate of the tension state as the tension value at the current moment after filtering out eccentric disturbances. The amplitude components from multiple consecutive sampling moments are connected sequentially in chronological order to form a smooth and continuous tension change curve. This curve is the tension balance signal after filtering out eccentric disturbances, which can accurately reflect the true tension state of the webbing during the unwinding process.

[0077] Step S104: Based on the tension balance signal after filtering out eccentric disturbances, extract the belt traction coordination relationship between the winding end and the unwinding end, process the belt traction coordination relationship through the proportional-integral-derivative control algorithm, and determine the synchronous control command.

[0078] The winding and unwinding speeds of the tension balance signal after filtering out eccentric disturbances are obtained. The real-time speed deviation between these two speeds is extracted. Based on the positive / negative direction and magnitude of the deviation, the current winding state is determined to be either winding-end leading or unwinding-end lagging, thus obtaining the deviation characteristics of the winding traction coordination relationship. Based on these deviation characteristics, a proportional-integral-derivative (PID) control algorithm is used. The proportional component generates an instantaneous correction based on the deviation magnitude; the integral component accumulates the deviation characteristics over time to eliminate steady-state residuals, obtaining an integral accumulation; and the derivative component generates a lead compensation based on the rate of change of the deviation characteristics. The instantaneous correction, integral accumulation, and lead compensation are added to obtain the comprehensive control output value. This comprehensive control output value is converted into a speed adjustment for the winding drive motor and a torque following for the unwinding drive motor. After amplitude limiting, the speed adjustment and torque following are combined and output to determine the synchronization control command.

[0079] During the winding process of the silk ribbon, the speed of the winding end and the unwinding end directly reflects whether the traction state at both ends is coordinated.

[0080] For example, the winding speed is calculated by multiplying the actual rotational speed of the winding drive motor by the current roll diameter, while the unwinding speed is calculated by multiplying the rotational angular velocity of the unwinding core tube by the corresponding roll diameter. In one embodiment, the real-time speed deviation is calculated by subtracting the unwinding speed from the winding speed. If the real-time speed deviation is positive, it indicates that the traction speed at the winding end is slower than the unwinding speed, and the webbing is in a slack state; if the real-time speed deviation is negative, it indicates that the unwinding speed is slower than the traction speed at the winding end, and the webbing is in a taut state. Based on the direction and magnitude of the deviation, the current traction coordination state can be determined, forming a deviation characteristic as the input basis for subsequent control.

[0081] Specifically, the proportional-integral-derivative (PID) control algorithm processes the deviation characteristics in three stages. The proportional stage directly multiplies the current deviation amplitude by a fixed proportional coefficient. This coefficient is pre-calibrated during the winding equipment debugging phase based on the webbing material and the range of the belt speed. The output instantaneous correction is linearly related to the deviation amplitude; the larger the deviation, the larger the correction. The integral stage accumulates the deviation characteristics at fixed time intervals. The accumulation result reflects the persistence of the deviation over a period of time. When there is a small but persistent speed difference, the integral accumulation gradually increases, thereby eliminating steady-state residuals that the proportional stage alone cannot eliminate. The derivative stage focuses on the rate of change of the deviation characteristics. By calculating the difference between the deviation values ​​at two adjacent sampling times, it determines whether the deviation is increasing or decreasing, and generates an advance compensation amount to apply correction before the deviation fully manifests.

[0082] It should be noted that the sum of the instantaneous correction, the integral accumulation, and the advance compensation constitutes the comprehensive control output value. This comprehensive control output value is a unified numerical value representing the coordinated adjustment range required at both the winding and unwinding ends at the current moment.

[0083] In one embodiment, the integrated control output value is converted into two parts: one part serves as the speed adjustment of the take-up drive motor, used to adjust the traction speed at the take-up end; the other part serves as the torque following of the unwinding drive motor, used to adjust the damping response at the unwinding end. Upper and lower limits are set for the speed adjustment and torque following, respectively. When the adjustment exceeds the limits, it is truncated to the limit value, completing the limiting process. The speed adjustment and torque following, after limiting, are combined and output to form a synchronous control command, which acts on the drive actuators at both the take-up and unwinding ends.

[0084] Based on the tension balance signal after filtering out eccentric disturbances, the winding end and unwinding end belt speeds are obtained. The synchronization coordination relationship and drive response delay time of the belt speeds at both ends are extracted. The alternating tension and looseness of the synchronization coordination relationship with the belt travel distance are analyzed. The speed adjustment amount in the aforementioned PID synchronization control command is optimized, and the speed adjustment strategy and adjustment execution cycle of the winding drive motor are formulated.

[0085] The take-up and unwinding belt speeds are extracted from the tension balance signal after filtering out eccentric disturbances. The instantaneous deviation value of the belt speeds at both ends is obtained by calculating the difference between their values ​​at the same moment. The synchronization and coordination relationship of the belt speeds at both ends is determined based on the fluctuation amplitude and direction of the instantaneous deviation value over time. Simultaneously, the time difference between the moment the take-up drive motor issues a command and the moment the unwinding belt speed responds, and the time difference between the moment the unwinding drive motor issues a command and the moment the take-up belt speed responds, are recorded as the drive response delay time. Based on the synchronization and coordination relationship, the length of the belt traveled along the belt travel direction is accumulated as the belt travel distance. The instantaneous deviation values ​​at different belt travel distances are statistically analyzed. If the instantaneous deviation value is positive, the position is marked as a tight belt state; if the instantaneous deviation value is negative, the position is marked as a loose belt state. The alternation between loose and tight states along the belt travel distance is used as the alternation of tightness and looseness. Based on the duration of the loose and tight states in the alternating loose and tight states and the drive response delay, the speed adjustment direction of the winding drive motor is determined to be deceleration in the loose state and acceleration in the tight state. The speed adjustment range is determined based on the product of the duration and the drive response delay, and the time interval between two adjacent loose and tight state switches is taken as the adjustment execution cycle.

[0086] In one implementation, the belt speeds at the take-up and unwind ends are extracted based on tension balance signals and the physical properties of the webbing.

[0087] Specifically, let the elastic modulus of the webbing be E, the cross-sectional area be A, and the real-time tension be F. By monitoring the rate of change of the tension signal dF / dt, and combining this with the relationship Δv=1 / (E×A)×dF / dt, where Δv is the instantaneous linear velocity difference between the take-up and unwinding ends, and given that the system reference speed v0 is the initial speed extracted based on the rotational speed and roll diameter mentioned earlier, the take-up end speed vc=v0+Δv / 2 and the unwinding end speed vp=v0-Δv / 2 can be corrected respectively, thereby improving the speed accuracy.

[0088] Specifically, the instantaneous deviation value reflects the degree of speed matching between the take-up and unwinding ends at the same moment. The instantaneous deviation value is obtained by subtracting the unwinding end's belt speed from the take-up end's belt speed. If this deviation value exhibits periodic fluctuations over time and the fluctuation amplitude remains within a preset range, the two ends are determined to be in a synchronized and coordinated state. If the deviation value shows a unidirectional, continuous increase or the fluctuation amplitude exceeds the preset range, the two ends are determined to be in a misaligned state. The synchronized and coordinated relationship is the quantitative representation of the above state.

[0089] It should be noted that the drive response delay duration is measured by recording a timestamp when the drive motor at the take-up end issues a speed adjustment command, and monitoring the moment when the belt speed at the unwinding end begins to change accordingly. The difference between the two timestamps is the drive response delay duration.

[0090] In one embodiment, the statistical analysis of alternating tension and slack is performed along the webbing travel distance. As the webbing travels from the unwinding end to the winding end, the accumulated webbing length is used as the coordinate of the webbing travel distance. For each segment of webbing travel distance, the instantaneous deviation value within that segment is read. A positive instantaneous deviation value indicates that the winding end speed is faster than the unwinding end, and the webbing is in a slack state; a negative instantaneous deviation value indicates that the winding end speed is slower than the unwinding end, and the webbing is in a slack state. The distribution positions of the slack and slack states along the webbing travel distance are marked to form a tension and slack alternation distribution map.

[0091] For example, the direction of speed adjustment is determined based on the current tension state of the webbing: in a slightly loose state, the winding drive motor should accelerate to tighten the webbing; in a slightly tight state, the winding drive motor should decelerate to loosen the webbing. The speed adjustment range is determined based on the product of the duration of the slightly loose or slightly tight state and the drive response delay; the longer the duration and the greater the response delay, the larger the adjustment range. Furthermore, the determination of the adjustment execution cycle is based on the switching frequency of the tension state. The time interval between two adjacent tension state switching is statistically analyzed and used as the adjustment execution cycle, ensuring that the speed adjustment rhythm of the winding drive motor is consistent with the rhythm of the webbing tension change.

[0092] Step S105: Perform frequency domain analysis on the synchronization control command to identify residual eccentric correlation oscillation ripples. Filter out the oscillation ripples using rotation period characteristics as a criterion to obtain a stable dual-end tension balance control command.

[0093] The time-domain waveform data of the synchronization control command is acquired, and the synchronization control command is transformed from the time domain to the frequency domain using Fourier transform to obtain the spectral distribution of the synchronization control command. The spectral distribution presents the energy proportion of each frequency component with frequency as the horizontal axis and amplitude as the vertical axis. Frequency components with amplitudes exceeding a preset amplitude threshold are extracted from the spectral distribution as significant frequency components. The fundamental frequency value corresponding to the core tube rotation is calculated based on the rotation period characteristic. The fundamental frequency value and its integer multiples are used as eccentric correlation frequencies. Frequency components matching the eccentric correlation frequencies are selected from the significant frequency components. If the deviation between the frequency value of the significant frequency component and the eccentric correlation frequency is within a preset frequency tolerance range, the frequency component is determined to be a residual eccentric correlation oscillation ripple. Based on the position of the eccentric correlation oscillation ripple in the spectral distribution, a band-stop filter is used to suppress the eccentric correlation frequency and its adjacent frequency bands, attenuating the amplitude of the eccentric correlation oscillation ripple to below a preset suppression threshold, resulting in a spectral distribution after filtering out the oscillation ripple. The frequency domain signal is converted back to the time domain by inverse Fourier transform of the spectrum distribution after filtering out the oscillation ripple, and the control command waveform after filtering out the residual components of the eccentric disturbance is obtained. The control command waveform is used as a stable two-end tension balance control command.

[0094] In one implementation, the Fourier transform is a mathematical transformation method that converts a time-domain signal into a frequency-domain signal. The synchronization control command, as a time-domain signal, exhibits a specific oscillation pattern as its waveform changes over time. Through the Fourier transform, the synchronization control command is decomposed into a superposition of several sinusoidal components of different frequencies. Each sinusoidal component corresponds to a specific frequency value and amplitude, and the set of all components constitutes a spectral distribution.

[0095] Specifically, the spectral distribution is presented with frequency on the horizontal axis and amplitude on the vertical axis. The frequency axis covers the range from zero to half the sampling frequency, while the amplitude axis reflects the energy proportion of each frequency component in the original signal. The peak positions in the spectral distribution indicate the main frequency components present in the original signal, and the height of the peak reflects the intensity of that frequency component.

[0096] It should be noted that the extraction of significant frequency components is based on the determination of a preset amplitude threshold. All frequency points in the spectral distribution are traversed; if the amplitude of a frequency point exceeds the preset amplitude threshold, then that frequency point and its corresponding amplitude are marked as significant frequency components. The preset amplitude threshold is typically set with reference to the average amplitude level in the spectral distribution, taking a multiple of the average amplitude as the determination criterion.

[0097] In one embodiment, the calculation of the fundamental frequency is closely related to the rotational characteristics of the die tube. The rotational periodicity characterizes the time it takes for the die tube to complete one full rotation, and the fundamental frequency is defined as the reciprocal of the rotational periodicity, physically representing the number of rotations the die tube makes per second. Since the radial runout caused by die tube eccentricity repeats with a period of one rotation of the die tube, the eccentricity disturbance in the frequency domain necessarily manifests as energy concentration at the fundamental frequency and its integer multiples. The frequency sequence formed by the fundamental frequency itself, twice the fundamental frequency, three times the fundamental frequency, and so on, is collectively referred to as the eccentricity-related frequencies.

[0098] For example, the existence of integer multiples of frequencies stems from the non-sinusoidal characteristics of the eccentric disturbance waveform. An ideal sine wave appears as a single-frequency spectral line in the frequency domain, while the actual eccentric disturbance waveform exhibits a complex shape due to factors such as irregularities on the core tube surface and nonlinear elastic response of the webbing. After Fourier decomposition, harmonic components will be generated at integer multiples of the fundamental frequency.

[0099] Understandably, the identification of eccentrically correlated oscillation ripple is achieved through frequency matching. For each extracted significant frequency component, the absolute value of the difference between its frequency value and each frequency in the eccentrically correlated frequency sequence is calculated. If the absolute value of the difference is less than a preset frequency tolerance, the significant frequency component is determined to be an eccentrically correlated oscillation ripple. The preset frequency tolerance is set considering the small fluctuations in the chip rotation speed and the frequency resolution limitations of the Fourier transform. Furthermore, a band-stop filter is a filtering device that suppresses signal energy within a specific frequency band while allowing the signal to pass through in other frequency bands. After locating the frequency position of the eccentrically correlated oscillation ripple in the spectral distribution, a stopband is constructed with that frequency as the center and a certain bandwidth as the range. The band-stop filter attenuates the frequency components within the stopband, reducing their amplitude to below a preset suppression threshold, while having no effect on frequency components outside the stopband. By applying band-stop filtering to all eccentrically correlated frequencies one by one, a spectral distribution after filtering out the oscillation ripple is formed, which no longer contains periodic disturbance components associated with chip eccentricity.

[0100] Preferably, the inverse Fourier transform is the reverse process of the Fourier transform, converting the frequency domain signal back into a time domain signal. Applying the inverse Fourier transform to the spectral distribution after filtering out oscillation ripples re-superimposes the remaining effective frequency components according to their respective frequencies, amplitudes, and phases, restoring the time domain waveform.

[0101] In one possible implementation, the time-domain waveform obtained by inverse Fourier transform is the control command waveform after filtering out residual components of eccentricity disturbance. Compared with the original synchronous control command, this control command waveform removes the periodic oscillation interference caused by core tube eccentricity, resulting in a more stable signal. It is then used as a stable dual-end tension balance control command output.

[0102] Step S106: Apply a stable dual-end tension balance control command to the drive motor of the winding system, acquire the real-time roll diameter and webbing tension, identify any abnormal fluctuations in the real-time roll diameter and webbing tension and use them as iterative feedback, and send them back to the roll diameter scanning stage to form a closed-loop control to achieve synchronous and stable winding speed.

[0103] The stable dual-end tension balance control command is output to the control port of the take-up drive motor. The take-up drive motor performs speed change actions according to the speed adjustment range and direction in the control command. During the speed change process, the real-time roll diameter value and webbing tension detection value at the take-up end are collected simultaneously. The real-time roll diameter value and webbing tension detection value are paired and recorded according to the sampling time to form roll diameter tension time series data. For the roll diameter tension time series data, the change range of the real-time roll diameter value and webbing tension detection value within the continuous sampling period is extracted. If the change range of the real-time roll diameter value exceeds the preset roll diameter fluctuation threshold, or the change range of the webbing tension detection value exceeds the preset tension fluctuation threshold, it is determined that there is a fluctuation anomaly. The occurrence time, anomaly type, and anomaly amplitude of the fluctuation anomaly are combined to form an anomaly feedback information. The abnormal feedback information is sent back to the roll diameter scanning stage. The adjustment object is determined according to the abnormality type. If the abnormality type is roll diameter fluctuation, the sampling frequency of the laser rangefinder is increased. If the abnormality type is tension fluctuation, the phase synchronization calibration value of the angle encoder is adjusted. The roll diameter scanning and periodic jump contour recognition are re-executed by adjusting the sampling frequency and phase synchronization calibration value to form a closed loop control to achieve synchronous and stable winding speed.

[0104] In one implementation, a stable dual-end tension balance control command is output to the control port of the take-up drive motor via a signal transmission line. Upon receiving the control command, the take-up drive motor executes a corresponding speed change action based on the speed adjustment range and direction information contained in the command, driving the take-up roller to rotate at the specified speed.

[0105] Specifically, during the process of the winding drive motor changing its speed, a laser rangefinder installed on the side of the winding roller continuously performs non-contact scanning on the surface of the core tube at the winding end to obtain the real-time roll diameter value. At the same time, a tension sensor installed in the webbing path synchronously collects the actual tension detection value of the webbing, and pairs the roll diameter value at each sampling moment with the corresponding tension detection value according to the timestamp to form roll diameter tension time sequence data.

[0106] It should be noted that the identification of fluctuation anomalies is based on a preset threshold judgment mechanism. For the real-time roll diameter values ​​within multiple consecutive sampling periods in the roll diameter tension time series data, the absolute value of the difference between the roll diameter values ​​at adjacent sampling times is calculated as the roll diameter change amplitude; for the webbing tension detection values ​​within multiple consecutive sampling periods, the absolute value of the difference between the tension detection values ​​at adjacent sampling times is calculated as the tension change amplitude. If the roll diameter change amplitude exceeds the preset roll diameter fluctuation threshold, or the tension change amplitude exceeds the preset tension fluctuation threshold, then a fluctuation anomaly is determined to exist.

[0107] In one embodiment, after an anomaly is identified, the time of occurrence, type, and magnitude of the anomaly are combined to form an anomaly feedback message. The anomaly types are categorized into two types: roll diameter fluctuation anomaly and tension fluctuation anomaly. The anomaly magnitude is a specific value exceeding the corresponding threshold. Further, after the anomaly feedback message is sent back to the roll diameter scanning stage, the adjustment target is determined based on the anomaly type. This stage refers to the process of real-time scanning of the roll diameter using a laser rangefinder and an angle encoder. If the anomaly type is a roll diameter fluctuation anomaly, it indicates that the current sampling frequency of the laser rangefinder is insufficient to capture rapid changes in the roll diameter. In this case, the sampling frequency of the laser rangefinder is increased to increase the number of sampling points per unit time. If the anomaly type is a tension fluctuation anomaly, it indicates a phase synchronization deviation between the angle encoder and the laser rangefinder. In this case, the phase synchronization calibration value of the angle encoder is corrected.

[0108] Preferably, the roll diameter scanning and periodic jump contour recognition are re-executed by adjusting the sampling frequency and phase synchronization calibration value. The periodic jump contour recognition refers to identifying the characteristics of the periodic jump contour based on the roll diameter scanning data. The input is the adjusted sampling data, and the output is the contour parameters. The process includes Fourier transform analysis of the periodic pattern. The recognition result is then input into the subsequent eccentricity disturbance filtering and control command generation stage, forming a complete closed loop from control command output, real-time data acquisition, fluctuation anomaly identification, parameter adjustment to re-scanning, so as to achieve synchronous and stable winding speed.

[0109] If the technical solution of this application involves the collection, processing, or application of personal information, the relevant products have, before implementing any personal information processing activities, fully and clearly informed individuals of the processing rules in accordance with the "Personal Information Protection Law of the People's Republic of China" and other current laws and regulations, and obtained their voluntary and explicit consent. If sensitive personal information is involved, the product has obtained the individual's separate consent before processing, and such consent is given in an explicit manner. For example, prominent signs are set up in the area where information collection devices such as cameras are located, clearly indicating "Entering is considered as consent to the collection of personal information"; or through pop-ups, checkboxes, user-initiated uploads, etc., under the premise of clearly listing the processor's identity, processing purpose, processing method, and information type, the user actively completes the authorization operation. The above mechanisms ensure that all personal information processing activities are based on legal authorization and fully comply with national compliance requirements regarding personal information protection.

[0110] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for synchronously controlling the winding speed of a silk webbing, characterized in that, include: The surface of the unwinding end core tube is continuously scanned by a laser rangefinder to obtain the current roll diameter and webbing tension. Combined with the rotation phase angle collected by the angle encoder, the periodic jumping profile of the roll diameter as the rotation phase angle changes is identified. Based on the periodic jumping profile, the radial jumping trajectory and rotation period features caused by eccentricity are extracted. Combined with the current roll diameter, the evolution law of the radial jumping trajectory under the roll diameter decreasing state is analyzed to obtain the dynamically corrected eccentricity disturbance filtering boundary. Using the dynamically corrected eccentric disturbance filtering boundary as a constraint, the Kalman filter algorithm is used to fuse the available signals below the eccentric disturbance filtering boundary in the webbing tension and the radial runout trajectory to obtain the tension balance signal after filtering out the eccentric disturbance. Based on the tension balance signal after filtering out eccentric disturbances, the belt traction coordination relationship between the take-up end and the unwinding end is extracted, and the belt traction coordination relationship is processed by the proportional-integral-derivative control algorithm to determine the synchronous control command. Frequency domain analysis is performed on the synchronization control command to identify residual eccentric correlation oscillation ripple in the synchronization control command. The residual eccentric correlation oscillation ripple is filtered out using the rotation period characteristic as a criterion to obtain a stable dual-end tension balance control command. The stable dual-end tension balance control command is applied to the drive motor of the winding system, and the real-time roll diameter and real-time webbing tension are acquired. The fluctuation anomalies of the real-time roll diameter and real-time webbing tension are identified and used as iterative feedback, which is then sent back to the roll diameter scanning stage to form a closed-loop control to achieve synchronous and stable winding speed.

2. The method for synchronous control of the winding speed of a silk webbing according to claim 1, characterized in that, The step of continuously scanning the surface of the unwinding end core tube with a laser rangefinder to obtain the current roll diameter and webbing tension, and combining this with the rotational phase angle acquired by the angle encoder disk to identify the periodic jump profile of the roll diameter as the rotational phase angle changes, includes: The real-time roll diameter value is obtained by converting the return time interval of the laser rangefinder to acquire distance sequence data. The rotation phase angle determined by the pulse signal of the angle encoder disk is timestamped with the real-time roll diameter value to form a phase roll diameter correspondence table; The extreme value positions in the phase roll diameter correspondence table are extracted and arranged to form the periodic jumping profile.

3. The method for synchronous control of the winding speed of a silk ribbon according to claim 1, characterized in that, The step involves extracting the radial jump trajectory and rotation period features caused by eccentricity based on the periodic jump contour, and combining this with the current roll diameter analysis to understand the evolution of the radial jump trajectory under the decreasing roll diameter state, thereby obtaining a dynamically corrected eccentricity disturbance filtering boundary, including: Extract the peak and valley positions of the periodic jumping profile, determine the radial jumping amplitude and fluctuation period caused by eccentricity, and use the fluctuation period as the rotation period feature; The current roll diameter is compared with the full roll diameter value, and the roll diameter reduction stage is determined to be a small roll diameter stage or a large roll diameter stage based on the difference ratio. The radial runout amplitudes of the small roll diameter stage and the large roll diameter stage are recorded separately. The difference in runout amplitude is compared and a mapping curve is established based on the decreasing trend of the difference with the roll diameter value, which serves as the evolution law of the runout trajectory. Based on the mapping curve, the upper limit of the jump amplitude is found for the current roll diameter as the boundary threshold. The boundary threshold is synchronously updated within the phase angle range corresponding to the rotation period feature to obtain the dynamically corrected eccentric disturbance filtering boundary.

4. The method for synchronous control of the winding speed of a silk webbing according to claim 1, characterized in that, The process involves using the dynamically corrected eccentricity filtering boundary as a constraint, and employing a Kalman filter algorithm to fuse the available signals below the eccentricity filtering boundary in the webbing tension and the radial runout trajectory, to obtain the tension balance signal after filtering out the eccentricity disturbance. This includes: The tension fluctuation component is compared with the radial runout amplitude, and usable signals below the eccentric disturbance filtering boundary are retained; By combining the amplitude and instantaneous velocity of the available signals to form a tension state vector, Kalman filtering is performed to estimate the state and obtain the optimal estimate. The amplitude components of the optimal estimate are connected to form the tension balance signal after filtering out the eccentric disturbance.

5. The method for synchronous control of the winding speed of a silk ribbon according to claim 1, characterized in that, The step involves extracting the belt traction coordination relationship between the take-up end and the unwinding end based on the tension balance signal after filtering out eccentric disturbances, processing the belt traction coordination relationship using a proportional-integral-derivative control algorithm, and determining the synchronization control command, including: The deviation characteristics of the conveyor belt traction coordination relationship are extracted, and the proportional-integral-derivative control algorithm is used to generate instantaneous correction, integral accumulation and advance compensation. The combined control output value is then converted into speed regulation and torque following values ​​to generate the synchronous control command.

6. The method for synchronous control of the winding speed of a silk ribbon according to claim 5, characterized in that, The combined control output value is converted into speed regulation and torque following quantities to generate the synchronous control command, including: Extract the instantaneous deviation value at the moment when the belt speeds at both ends are the same, determine the synchronization and coordination relationship of the belt speeds at both ends based on the fluctuation amplitude and fluctuation direction of the instantaneous deviation value, and record the time difference between the issuance of the command and the change of the response as the drive response delay time. The instantaneous deviation values ​​at different webbing travel distances are statistically analyzed. Positive values ​​are marked as webbing tightness and negative values ​​are marked as webbing looseness. The alternation of these states is taken as an indication of alternating tightness and looseness. Based on the duration of the alternating tension and slack behavior and the drive response delay, deceleration adjustment is determined in the slack state and acceleration adjustment is determined in the slack state. The speed adjustment range of the speed adjustment amount is determined according to the product relationship. The state switching time interval is used as the adjustment execution cycle to generate the synchronous control command.

7. The method for synchronous control of the winding speed of a silk ribbon according to claim 1, characterized in that, The step of performing frequency domain analysis on the synchronization control command to identify residual eccentric correlation oscillation ripple in the synchronization control command, and filtering out the residual eccentric correlation oscillation ripple using the rotation period characteristic as a criterion to obtain a stable dual-end tension balance control command includes: The significant frequency components of the synchronization control command are extracted using Fourier transform; The rotation periodicity feature is used to select the matching frequency as the residual eccentric correlated oscillation ripple; The waveform obtained after suppression by a band-stop filter and inverse Fourier transform is used as the stable dual-end tension balance control command.

8. The method for synchronous control of the winding speed of a silk webbing according to claim 1, characterized in that, The process involves applying the stable dual-end tension balance control command to the drive motor of the winding system, acquiring the real-time roll diameter and real-time webbing tension, identifying fluctuations in the real-time roll diameter and real-time webbing tension and using these as iterative feedback, and transmitting this feedback back to the roll diameter scanning stage to form a closed-loop control to achieve synchronous and stable winding speed. This includes: The stable dual-end tension balance control command is output to execute the speed change action, and the paired real-time roll diameter and the real-time webbing tension are recorded to form roll diameter tension timing data; When the continuous change in the roll diameter tension time series data exceeds the fluctuation threshold, an abnormal feedback information is generated and transmitted back. Adjust the sampling frequency of the laser rangefinder or the phase synchronization calibration value of the angle encoder disk and re-perform the roll diameter scan.

Images