Tension feed-forward based warp let-off servo control method and system

Abstract

This invention relates to the field of automatic control technology for textile machinery, and particularly to a warp feed servo control method and system for looms based on tension feedforward. The method includes: acquiring the spindle speed signal, warp tension signal, and warp tension setpoint; detecting load fluctuations to obtain the required adjustment amount of the warp feed servo speed; fitting the spindle speed signal to obtain acceleration / deceleration stage identifiers, generating a torque distribution scheme, and outputting a warp feed servo torque adjustment command; decomposing the warp tension signal to obtain the dominant fluctuation frequency, and determining the response lag based on the control loop response characteristics; acquiring servo status data, updating the tension feedforward parameters, predicting disturbances, generating torque compensation and speed compensation adjustment sequences, and driving the warp feed servo to keep the warp tension within the allowable deviation range.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology for textile machinery, and in particular to a servo control method and system for warp feeding of looms based on tension feedforward. Background Technology

[0002] During weaving, warp tension directly affects shedding stability, weft density consistency, and breakage rate, thus impacting fabric quality and production efficiency. From a process control perspective, when the loom spindle starts, stops, accelerates, decelerates, or experiences load fluctuations, the warp feed mechanism is prone to response lag due to inertia, friction, and transmission clearance, leading to transient impacts and periodic oscillations in warp tension. Simultaneously, the mechanical resistance and dynamic characteristics vary significantly under different fabric types and operating conditions. Fixed-parameter process control strategies struggle to maintain tension consistency during both start / stop and stable phases, resulting in overshoot, oscillations, or insufficient follow-through, affecting continuous production and quality stability. Summary of the Invention

[0003] This invention provides a warp feed servo control method and system for looms based on tension feedforward, which is used to at least solve the problem of how to achieve fast and stable following of warp tension by the warp feed servo under the conditions of loom acceleration / deceleration and load fluctuation.

[0004] In a first aspect, the present invention provides a servo control method for warp feeding on a loom based on tension feedforward, comprising the following steps: The spindle speed signal, warp tension signal, and warp tension setpoint are acquired. Load fluctuation detection is performed on the spindle speed signal and warp tension signal to obtain the required adjustment of the warp feed servo speed. The spindle speed signal is fitted with a speed curve to obtain the acceleration and deceleration stage identifier. Based on the acceleration and deceleration stage identifier and the warp feed servo speed adjustment requirement, a torque distribution scheme is generated. Based on the torque distribution scheme and the warp tension signal, a warp feed servo torque adjustment command is generated. The tension fluctuation of the warp tension signal is decomposed to obtain the dominant fluctuation frequency. The dominant fluctuation frequency is matched with the response characteristics of the warp feed servo control loop to obtain the response hysteresis determination result. The system acquires warp feed servo status data, updates tension feedforward parameters based on response lag determination results, generates predicted disturbance data, generates torque compensation and speed compensation based on predicted disturbance data, updates dynamic thresholds to generate adjustment sequences, and drives the warp feed servo to keep the warp tension within the allowable deviation range of the warp tension set value.

[0005] In one possible implementation, load fluctuation detection of the spindle speed signal and warp tension signal includes: statistically analyzing the change and fluctuation amplitude of the spindle speed signal within a preset time window, and statistically analyzing the tension deviation and fluctuation amplitude of the warp tension signal within a preset time window, and determining the required adjustment amount of the warp feed servo speed based on the correlation coefficient between the change of the spindle speed signal and the change of the warp tension signal.

[0006] In one possible implementation, the warp tension signal is acquired by a warp tension sensor, the fluctuation amplitude of the spindle speed signal is the difference between the maximum and minimum values ​​of the spindle speed signal within a preset time window, the fluctuation amplitude of the warp tension signal is the difference between the maximum and minimum values ​​of the warp tension signal within a preset time window, and the tension deviation of the warp tension signal is the difference between the warp tension signal and the set warp tension value.

[0007] In one possible implementation, obtaining the acceleration / deceleration stage identifier by fitting the spindle speed signal to the speed curve includes: performing polynomial fitting on the spindle speed signal to obtain the spindle speed trend curve, and determining the acceleration / deceleration stage identifier based on the difference between adjacent sampling points of the spindle speed trend curve.

[0008] In one possible implementation, the torque distribution scheme includes an inertial compensation torque component, a friction compensation torque component, and a tension feedforward torque component, and sets torque amplitude constraints and torque change rate constraints on the servo torque adjustment command.

[0009] In one possible implementation, generating the feed servo torque adjustment command based on the torque distribution scheme includes: generating a torque compensation signal based on the torque distribution scheme; filtering the torque compensation signal and performing amplitude limiting and slope limiting processing to obtain a smooth torque compensation signal; and performing feedback calibration on the smooth torque compensation signal based on the deviation between the feed servo torque observation value and the smooth torque compensation signal to output the feed servo torque adjustment command.

[0010] In one possible implementation, decomposing the warp tension signal into tension fluctuations includes: performing spectral analysis on the warp tension signal to obtain the tension spectrum, and determining the dominant fluctuation frequency based on the main peak of the tension spectrum.

[0011] In one possible implementation, matching the dominant fluctuation frequency with the response characteristics of the warp feed servo control loop to obtain the response lag determination result includes: comparing the dominant fluctuation frequency with the response bandwidth of the warp feed servo control loop, determining the equivalent time delay based on the cross-correlation analysis of the warp tension signal and the warp feed servo torque adjustment command, and obtaining the response lag determination result based on the comparison of the equivalent time delay and the lag threshold.

[0012] In one possible implementation, the tension feedforward parameters include tension feedforward gain and phase compensation parameters; generating predicted disturbance data includes trend extrapolation of the warp tension signal and the spindle speed signal to obtain predicted disturbance data; the warp feed servo status data includes warp feed servo motor current and warp feed servo motor speed; updating dynamic thresholds includes updating the adjustment threshold used to generate the torque distribution scheme and the hysteresis threshold used to obtain the response hysteresis determination result based on the warp feed servo status data; and the adjustment sequence includes a speed compensation sequence and a torque compensation sequence.

[0013] Secondly, the present invention provides a tension-feedforward-based warp feed servo control system for looms, used to implement the tension-feedforward-based warp feed servo control method for looms, the system comprising: The fluctuation detection module is used to acquire the spindle speed signal, warp tension signal and warp tension set value, and to perform load fluctuation detection on the spindle speed signal and warp tension signal to obtain the required adjustment amount of the warp feed servo speed. The fitting and allocation module is used to fit the spindle speed signal to obtain the acceleration and deceleration stage identifier, generate a torque allocation scheme based on the acceleration and deceleration stage identifier and the warp feed servo speed adjustment requirement, and generate a warp feed servo torque adjustment command based on the torque allocation scheme and the warp tension signal. The hysteresis determination module is used to decompose the tension fluctuation of the warp tension signal to obtain the dominant fluctuation frequency, and then perform frequency matching between the dominant fluctuation frequency and the response characteristics of the warp feed servo control loop to obtain the response hysteresis determination result. The collaborative adjustment module is used to acquire warp feed servo status data, update tension feedforward parameters based on response lag judgment results and generate predicted disturbance data, generate torque compensation and speed compensation based on predicted disturbance data, update dynamic threshold to generate adjustment sequence, and drive warp feed servo to keep warp tension within the allowable deviation range of warp tension set value.

[0014] Compared with the prior art, the advantages and beneficial effects of the present invention are as follows: By employing load fluctuation detection and correlation determination techniques, the required warp feed servo speed adjustment is effectively extracted when the spindle operating conditions change, reducing noise triggering and erroneous adjustments. Through speed curve fitting and acceleration / deceleration stage identification techniques, the torque distribution scheme is configured specifically for different motion stages, reducing tension impacts caused by inertia and friction changes. Through tension fluctuation decomposition and dominant fluctuation frequency extraction techniques, the main cause of tension oscillations is located. Through frequency matching and equivalent time delay determination techniques, the response lag of the control loop is quantitatively identified. Through adaptive updating of tension feedforward parameters and predictive disturbance compensation techniques, coordinated output of torque compensation and speed compensation is achieved, keeping warp tension within the allowable deviation range of the set value, improving weaving stability and fabric consistency. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the execution flow of the method of the present invention; Figure 2 This is a structural block diagram of the system of the present invention. Detailed Implementation

[0016] The embodiments of the present disclosure will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the disclosure. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the present disclosure for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts of the present disclosure.

[0017] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0018] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0019] Tension feedforward is a feedforward compensation strategy in process control. In warp feed servo control, it not only utilizes the feedback deviation of warp tension for closed-loop correction but also converts observable measurements or trends that predict tension changes into control compensation quantities in advance, applying them to the torque or speed commands of the warp feed servo before the tension deviation significantly increases. The basic idea is to establish a correspondence between the warp tension signal and external disturbance sources such as the spindle speed signal. By identifying the intensity and timing of the disturbance, a compensation quantity that is in phase with or ahead of the disturbance is generated to offset the response lag caused by the inertia, friction, and transmission lag of the warp feed mechanism. Thus, tension control changes from "correcting after deviation occurs" to "pre-compensating when disturbance occurs," enabling the warp feed servo to maintain timely and smooth tension adjustment under acceleration / deceleration and load fluctuation conditions. Based on the above process control concept, this invention further combines load fluctuation detection, acceleration / deceleration stage identification, and response lag determination to adaptively update the tension feedforward parameters and introduce predictive disturbance compensation, thereby forming an implementable collaborative adjustment scheme.

[0020] like Figure 1As shown, a servo control method for warp feed in a loom based on tension feedforward includes the following steps: The spindle speed signal, warp tension signal, and warp tension setpoint are acquired. Load fluctuation detection is performed on the spindle speed signal and warp tension signal to obtain the required adjustment of the warp feed servo speed. In one implementation, the spindle speed signal is acquired by a rotary encoder mounted on the loom spindle, and the warp tension signal is acquired by a tension sensor mounted on the warp path. The warp tension setpoint is written to the controller via the human-machine interface and used for subsequent deviation calculation. The controller synchronously samples the spindle speed signal and the warp tension signal, and performs low-pass filtering on the warp tension signal to suppress high-frequency noise. Then, it calculates the change and fluctuation amplitude of the spindle speed signal in units of a sliding time window, and calculates the tension deviation and fluctuation amplitude of the warp tension signal relative to the warp tension setpoint. Based on the correlation coefficient between the spindle speed signal change and the tension deviation, it identifies whether the spindle speed disturbance causes abnormal warp tension, thereby completing load fluctuation detection. When load fluctuation is detected, the controller fuses the spindle speed change and the tension deviation to obtain the warp feed servo speed adjustment demand, and outputs the warp feed servo speed adjustment demand to the subsequent torque distribution step to generate the corresponding warp feed servo control command. The expression for calculating the warp feed servo speed adjustment demand is: To meet the requirements of servo speed regulation, This represents the change in the spindle speed signal within the sliding time window. The difference between the warp tension signal and the warp tension set value. The weighting coefficient for spindle speed variation. This is the tension deviation weighting coefficient.

[0021] Load fluctuation detection of spindle speed signal and warp tension signal includes: statistically analyzing the change and fluctuation amplitude of spindle speed signal within a preset time window, and statistically analyzing the tension deviation and fluctuation amplitude of warp tension signal within a preset time window, and determining the required adjustment of warp feed servo speed based on the correlation coefficient between the change of spindle speed signal and the change of warp tension signal.

[0022] In one embodiment, the controller synchronously samples the spindle speed signal and the warp tension signal, and establishes a buffer queue with a preset time window within the controller. The preset time window is determined by the sampling period and the number of sampling points within the window. The buffer queue is used to store the spindle speed signal sequence and the warp tension signal sequence within the preset time window. To reduce the impact of sensor noise on the statistical results, the warp tension signal undergoes low-pass filtering before entering the buffer queue, and the spindle speed signal undergoes deburring processing before entering the buffer queue, ensuring data alignment at the same sampling time.

[0023] Load fluctuation detection is performed after each cache queue update. Within a preset time window, the change in the spindle speed signal is taken as the difference between the sampled value at the end of the preset time window and the sampled value at the beginning of the preset time window, and the fluctuation amplitude of the spindle speed signal is taken as the difference between the maximum and minimum values ​​of the spindle speed signal within the preset time window; the tension deviation of the warp tension signal is taken as the statistical value of the deviation of the warp tension signal from the set warp tension value within the preset time window, and the fluctuation amplitude of the warp tension signal is taken as the difference between the maximum and minimum values ​​of the warp tension signal within the preset time window. These statistics are used to characterize the degree of synchronization between changes in spindle operating conditions and abnormal warp tension, thereby suppressing false triggers caused solely by random noise.

[0024] To quantify the correlation between changes in spindle speed signal and changes in warp tension signal, the controller calculates a correlation coefficient based on the spindle speed signal sequence and the warp tension signal sequence within a preset time window. The correlation coefficient is determined using the following expression: The correlation coefficient, For the first time window within the preset time window Each spindle speed signal sample value For the first time window within the preset time window Each warp tension signal sample value, The average value of the spindle speed signal within a preset time window. The average value of the warp tension signal within a preset time window. This represents the number of sampling points within a preset time window.

[0025] The controller determines load fluctuations based on correlation coefficients and preset correlation thresholds. When the correlation coefficients meet the correlation determination conditions, it is determined that the spindle speed disturbance has a significant impact on warp tension. Based on the change in the spindle speed signal, the fluctuation amplitude of the spindle speed signal, the tension deviation of the warp tension signal, and the fluctuation amplitude of the warp tension signal, the controller determines the adjustment requirement of the warp feed servo speed. The calculation of the warp feed servo speed adjustment requirement uses the aforementioned calculation expression combined with the correlation coefficient for weighting, prioritizing adjustments for conditions with high correlation. Therefore, in scenarios involving spindle acceleration / deceleration and sudden load changes, the controller can identify the source of tension anomalies earlier and output the adjustment requirement, providing stable input for subsequent torque distribution and feedforward compensation, and reducing warp tension fluctuations.

[0026] The warp tension signal is acquired by the warp tension sensor. The fluctuation amplitude of the spindle speed signal is the difference between the maximum and minimum values ​​of the spindle speed signal within a preset time window. The fluctuation amplitude of the warp tension signal is the difference between the maximum and minimum values ​​of the warp tension signal within a preset time window. The tension deviation of the warp tension signal is the difference between the warp tension signal and the set warp tension value.

[0027] In one embodiment, the warp tension signal is acquired by a warp tension sensor, which is positioned at the tension measuring roller or guide roller along the warp feed path. This causes the warp yarn to wrap around the tension measuring component corresponding to the warp tension sensor, forming a wrap angle. The warp tension sensor converts the warp tension into a voltage or current signal and outputs it to the controller's analog-to-digital converter interface. To ensure repeatability across different loom batches and installation conditions, the controller performs zero-point calibration on the warp tension sensor during the power-on self-test phase and writes a sensitivity coefficient during the calibration phase, thereby converting the warp tension sensor's output signal into a warp tension signal. The spindle speed signal is acquired by a spindle rotary encoder. The controller synchronously samples the spindle speed signal and the warp tension signal using a unified sampling period and establishes a buffer queue with a preset time window. The preset time window is used to limit the range of continuous sampling points participating in the statistics.

[0028] During load fluctuation detection, the controller calculates the fluctuation amplitudes of the spindle speed signal and the warp tension signal within a preset time window. The fluctuation amplitude of the spindle speed signal is defined as the difference between the maximum and minimum values ​​of the spindle speed signal within the preset time window, and the fluctuation amplitude of the warp tension signal is defined as the difference between the maximum and minimum values ​​of the warp tension signal within the preset time window. To reduce computational load, the controller synchronously maintains the maximum and minimum values ​​within the preset time window when updating the buffer queue. When a new sampling point is added to the buffer queue and an old sampling point is removed, if the removed sampling point is the current maximum or minimum value, a new maximum or minimum value is searched again within the preset time window. The tension deviation of the warp tension signal is defined as the difference between the warp tension signal and the warp tension setpoint. The warp tension setpoint is written to the human-machine interface or issued by the process formula and stored in the controller, serving as a reference for tension control. Based on the above definitions, the fluctuation amplitude of the spindle speed signal is used to characterize the intensity of spindle operating condition fluctuation, the fluctuation amplitude of the warp tension signal is used to characterize the intensity of tension fluctuation, and the tension deviation of the warp tension signal is used to characterize the direction and degree of tension deviation from the set operating condition. This provides a consistent, feasible, and easy-to-deploy basic statistical quantity for subsequent correlation determination and calculation of the required adjustment of the warp feed servo speed, thereby improving the stability of tension anomaly identification under acceleration / deceleration and load change scenarios.

[0029] The spindle speed signal is fitted with a speed curve to obtain the acceleration and deceleration stage identifier. Based on the acceleration and deceleration stage identifier and the warp feed servo speed adjustment requirement, a torque distribution scheme is generated. Based on the torque distribution scheme and the warp tension signal, a warp feed servo torque adjustment command is generated. In one embodiment, the controller performs speed curve fitting on the spindle speed signal, using polynomial least squares fitting to obtain the spindle speed trend curve, and determines the acceleration / deceleration stage identifier based on the difference between adjacent sampling points of the spindle speed trend curve. A torque allocation scheme is generated based on the acceleration / deceleration stage identifier and the warp feed servo speed adjustment requirement. The torque allocation scheme includes an inertial compensation torque component, a friction compensation torque component, and a tension feedforward torque component. The direction of the inertial compensation torque component is determined according to the acceleration / deceleration stage identifier. The warp feed servo speed adjustment requirement is converted proportionally to obtain a basic torque command, which is then calibrated using a preset mapping table. A tension feedforward torque component is generated based on the tension deviation of the warp tension signal, and these components are superimposed to obtain the output torque command. Amplitude and rate of change constraints are applied to the output torque command to obtain the warp feed servo torque adjustment command, which is then sent to the servo driver.

[0030] The process of fitting the spindle speed signal to obtain acceleration / deceleration stage identifiers includes: performing polynomial fitting on the spindle speed signal to obtain a spindle speed trend curve, and determining the acceleration / deceleration stage identifiers based on the difference between adjacent sampling points of the spindle speed trend curve.

[0031] In one embodiment, after determining the required adjustment amount of the servo speed, the controller performs speed curve fitting on the spindle speed signal to obtain the acceleration / deceleration stage identifier. Specifically, the controller continuously samples the spindle speed signal based on a preset sampling period and forms a spindle speed signal sequence in units of preset time windows. To avoid interference from instantaneous spikes on the fitting results, the controller first performs de-spiking and moving average processing on the spindle speed signal, so that the spindle speed signal sequence maintains trend information while suppressing high-frequency jitter.

[0032] During the spindle speed curve fitting process, the controller employs a least-squares fitting method to fit the spindle speed signal sequence within a preset time window to the spindle speed trend curve. The spindle speed trend curve is represented by a polynomial of a preset order as follows: For the first The spindle speed trend value corresponding to each sampling point The sampling point number, Let be the order of the polynomial. These are the polynomial coefficients.

[0033] The controller determines the acceleration / deceleration stage identifier based on the difference between adjacent sampling points of the spindle speed trend curve. The difference value is taken as the difference between the spindle speed trend value at the current sampling point and the spindle speed trend value at the previous sampling point. When the difference value meets the acceleration judgment condition, the acceleration / deceleration stage identifier is marked as an acceleration stage; when the difference value meets the deceleration judgment condition, the acceleration / deceleration stage identifier is marked as a deceleration stage; and when the difference value meets the constant speed judgment condition, the acceleration / deceleration stage identifier is marked as a constant speed stage. To improve the stability of the identifier, the controller sets a maintenance rule for the acceleration / deceleration stage identifier, ensuring that the identifier switches only after multiple consecutive sampling points meet the same judgment condition, thereby avoiding frequent jumps in the acceleration / deceleration stage identifier caused by small fluctuations in the spindle speed signal.

[0034] In the above way, the acceleration / deceleration stage indicator can reflect the trend change of the spindle speed signal. The controller outputs the acceleration / deceleration stage indicator to the subsequent torque distribution scheme generation stage to constrain the inertia compensation direction and tension feedforward injection timing. This allows the warp feed servo to obtain an adjustment basis consistent with the spindle motion state under the spindle acceleration / deceleration conditions, reducing sudden changes in warp feed servo torque caused by misjudged conditions, thereby reducing warp tension fluctuations.

[0035] The torque distribution scheme includes inertial compensation torque components, friction compensation torque components, and tension feedforward torque components, and sets torque amplitude constraints and torque change rate constraints for the servo torque adjustment commands sent to it.

[0036] In one embodiment, after receiving the acceleration / deceleration phase identifier and the required warp feed servo speed adjustment, the controller generates a torque distribution scheme and accordingly forms a warp feed servo torque adjustment command. The torque distribution scheme consists of an inertia compensation torque component, a friction compensation torque component, and a tension feedforward torque component, used to respectively address the changes in warp feed inertia load caused by spindle acceleration / deceleration, changes in mechanical friction and transmission resistance of the warp feed mechanism, and the feedforward compensation requirements caused by warp tension deviation.

[0037] The inertial compensation torque component is oriented based on the acceleration / deceleration phase indicators and its amplitude is determined by the required adjustment of the warp feed servo speed. This allows the warp feed servo to provide torque compensation in the acceleration direction during acceleration and in the deceleration direction during deceleration, thereby reducing tension spikes caused by warp feed response lag. The friction compensation torque component is obtained through a preset friction compensation table. This table uses the warp feed servo speed and direction as indices, recording friction compensation values ​​corresponding to different speed ranges and motion directions. The controller retrieves the friction compensation torque component from the preset friction compensation table based on the current warp feed servo operating status. The tension feedforward torque component is determined based on the difference between the warp tension signal and the warp tension setpoint. The sign of the difference determines the compensation direction, and the magnitude of the difference determines the compensation amplitude, ensuring that the warp feed servo injects compensation torque early in the occurrence of tension deviations, reducing closed-loop feedback lag.

[0038] In one embodiment, the servo torque adjustment command is obtained by superimposing three types of torque components, and its core calculation expression is: To send servo torque adjustment commands, For inertial compensation torque component, For friction compensation torque component, This represents the tension feedforward torque component.

[0039] To prevent the warp feed servo torque adjustment command from exceeding the capacity of the servo driver and the warp feed mechanism, the controller sets torque amplitude constraints and torque change rate constraints on the warp feed servo torque adjustment command. The torque amplitude constraint limits the warp feed servo torque adjustment command to between a preset maximum torque and a preset minimum torque; the torque change rate constraint limits the change in warp feed servo torque adjustment command between adjacent control cycles, thereby suppressing mechanism shocks and secondary tension fluctuations caused by sudden torque steps. Through the above torque distribution scheme and dual constraint processing, the warp feed servo can achieve smoother torque output under spindle acceleration / deceleration and load fluctuation scenarios, reducing the fluctuation amplitude of the warp tension signal and improving the stability of the weaving process.

[0040] The generation of feed servo torque adjustment commands based on the torque distribution scheme includes: generating a torque compensation signal based on the torque distribution scheme; filtering the torque compensation signal and performing amplitude and slope limiting processing to obtain a smooth torque compensation signal; and performing feedback calibration on the smooth torque compensation signal based on the deviation between the feed servo torque observation value and the smooth torque compensation signal to output the feed servo torque adjustment command.

[0041] In one embodiment, after generating a torque distribution scheme, the controller first converts the scheme into a torque compensation signal. This torque compensation signal characterizes the compensation torque to be applied under the current acceleration / deceleration phase and current tension deviation conditions. The torque compensation signal is obtained by superimposing the inertial compensation torque component, the friction compensation torque component, and the tension feedforward torque component. The superposition relationship is implemented using the method shown in the aforementioned torque adjustment command calculation expression. To prevent signal glitches and instantaneous fluctuations from being directly transmitted to the servo driver, the controller filters the torque compensation signal. The filtering method uses first-order low-pass filtering or moving average filtering. The filtering parameters are pre-configured and stored in the controller parameter area according to the sampling period, ensuring that the filtered signal maintains the adjustment trend while suppressing high-frequency jitter.

[0042] After filtering, the controller performs amplitude limiting and slope limiting on the torque compensation signal. Amplitude limiting restricts the torque compensation signal within the allowable output range of the warp feed servo, preventing protective shutdown caused by exceeding the rated capacity of the servo driver. Slope limiting limits the variation in the torque compensation signal between adjacent control cycles, preventing sudden torque changes from causing impact on the warp feed mechanism and secondary fluctuations in warp tension. The resulting smooth torque compensation signal, obtained through filtering, amplitude limiting, and slope limiting, serves as the basis for generating the warp feed servo torque adjustment command.

[0043] To improve the controllability and consistency of torque compensation, the controller further incorporates warp feed servo torque observations for feedback calibration of the smoothed torque compensation signal. The warp feed servo torque observations are estimated by the servo driver based on motor current and parameters and then transmitted back to the controller, or calculated by the controller based on the warp feed servo motor current. The controller calculates the deviation between the warp feed servo torque observations and the smoothed torque compensation signal, and calibrates the smoothed torque compensation signal based on this deviation. The calibration method involves adding a correction amount proportional to the deviation to the smoothed torque compensation signal, thereby compensating for torque output errors caused by friction variations, load changes, or parameter drift. After feedback calibration, a warp feed servo torque adjustment command is output and sent to the servo driver for execution. This ensures that the warp feed servo maintains smooth torque output consistent with tension requirements under acceleration / deceleration and load fluctuation conditions, reducing warp tension fluctuations and improving weaving stability.

[0044] The tension fluctuation of the warp tension signal is decomposed to obtain the dominant fluctuation frequency. The dominant fluctuation frequency is matched with the response characteristics of the warp feed servo control loop to obtain the response hysteresis determination result. In one embodiment, the controller extracts a warp tension signal sequence within a preset time window, performs spectral analysis on the sequence to obtain the tension spectrum, and selects the frequency component with the largest amplitude after removing the DC component as the dominant fluctuation frequency. The response characteristics of the warp feed servo control loop are pre-stored by the controller or obtained online. These characteristics include at least the response bandwidth and equivalent delay of the warp feed servo control loop. During frequency matching, the controller compares the dominant fluctuation frequency with the response bandwidth of the warp feed servo control loop. If the dominant fluctuation frequency is higher than the response bandwidth or the equivalent delay exceeds the hysteresis threshold, a response hysteresis determination result is generated indicating the presence of hysteresis. If the dominant fluctuation frequency is lower than the response bandwidth and the equivalent delay does not exceed the hysteresis threshold, a response hysteresis determination result is generated indicating no hysteresis. This determination result provides a basis for subsequent tension feedforward parameter updates and disturbance prediction, thereby reducing tension oscillations under acceleration and deceleration conditions.

[0045] Tension fluctuation decomposition of warp tension signals includes: performing spectral analysis on the warp tension signals to obtain the tension spectrum, and determining the dominant fluctuation frequency based on the main peak of the tension spectrum.

[0046] In one embodiment, after receiving the warp tension signal, the controller performs tension fluctuation decomposition on the warp tension signal to identify the main periodic components of the warp tension fluctuation. The tension fluctuation decomposition takes the warp tension signal sequence within a preset time window as input. The preset time window is determined by both the sampling period and the window length, and the window length is used to cover the main tension fluctuation process during the acceleration and deceleration phases of the loom. To improve the stability of the spectrum analysis, the controller first performs mean-reduction processing on the warp tension signal sequence within the preset time window to weaken the influence of the warp tension setpoint and long-term drift on the spectrum, and then applies a window function weighting to the warp tension signal sequence to reduce spectral leakage caused by truncation.

[0047] Spectrum analysis is implemented using Fast Fourier Transform (FFT). The controller converts the warp tension signal sequence into a tension spectrum, which includes a frequency axis and an amplitude axis. The frequency axis represents each frequency component, and the amplitude axis represents the amplitude intensity of the corresponding frequency component. The controller searches for the dominant peak within the tension spectrum; the dominant peak is the peak point corresponding to the frequency component with the largest amplitude. To avoid noise spikes being misidentified as dominant peaks, the dominant peak search is performed within a preset frequency range. This preset frequency range is pre-configured based on the loom spindle speed range, the transmission ratio of the warp feed mechanism, and the bandwidth of the warp tension sensor. Simultaneously, the controller sets amplitude judgment conditions for the dominant peak. When the amplitude of the dominant peak is lower than a preset noise threshold, the controller does not update the dominant fluctuation frequency and retains the dominant fluctuation frequency from the previous cycle, thereby avoiding unstable frequency output under conditions of weak tension fluctuations.

[0048] When the main peak meets the amplitude judgment condition, the controller determines the frequency corresponding to the main peak as the dominant fluctuation frequency and outputs the dominant fluctuation frequency to the frequency matching stage for comparison with the response characteristics of the warp feed servo control loop. By determining the dominant fluctuation frequency based on the main peak, the controller can capture the periodic component that has the greatest impact on tension stability in the warp tension fluctuation, so that subsequent response lag judgment and tension feedforward parameter update have a clear target frequency basis, thereby more effectively suppressing tension oscillations in scenarios of spindle acceleration / deceleration and load fluctuation.

[0049] The process of matching the dominant fluctuation frequency with the response characteristics of the warp feed servo control loop to obtain the response lag determination result includes: comparing the dominant fluctuation frequency with the response bandwidth of the warp feed servo control loop, determining the equivalent time delay based on the cross-correlation analysis of the warp tension signal and the warp feed servo torque adjustment command, and obtaining the response lag determination result based on the comparison of the equivalent time delay and the lag threshold.

[0050] In one embodiment, after obtaining the dominant fluctuation frequency, the controller matches the dominant fluctuation frequency with the response characteristics of the servo control loop to output a response hysteresis determination result. The response characteristics of the servo control loop are provided by the bandwidth parameter table of the servo driver or written into the controller parameter area during the debugging phase, and at least include the servo control loop response bandwidth and hysteresis threshold. The controller first compares the dominant fluctuation frequency with the servo control loop response bandwidth. When the dominant fluctuation frequency is higher than the servo control loop response bandwidth, the response hysteresis determination result is directly determined to indicate the presence of hysteresis, which is used to trigger subsequent tension feedforward parameter updates and disturbance prediction compensation.

[0051] When the dominant fluctuation frequency is lower than the response bandwidth of the warp feed servo control loop, the controller further performs cross-correlation analysis based on the warp tension signal and the warp feed servo torque adjustment command to determine the equivalent time delay. The cross-correlation analysis uses the same preset time window as the dominant fluctuation frequency extraction. The controller performs mean-reduction processing on the warp tension signal to obtain a tension deviation sequence, and simultaneously performs mean-reduction processing on the warp feed servo torque adjustment command to obtain a torque command sequence. Within a preset hysteresis search range, the cross-correlation value is calculated, and the time shift corresponding to the largest cross-correlation value is taken as the equivalent time delay. The core calculation expression is: The time shift is The cross-correlation value, For the first time window within the preset time window Each tension deviation sample value, For the first time window within the preset time window One torque command sample value, The number of sampling points within a preset time window. The time shift corresponding to the maximum cross-correlation value. For equivalent time delay, The sampling period.

[0052] The controller determines the response lag based on a comparison between the equivalent delay and a lag threshold. When the equivalent delay is greater than the lag threshold, the response lag determination result is that lag exists; when the equivalent delay is not greater than the lag threshold, the response lag determination result is that there is no lag. Through a combination of frequency band comparison and cross-correlation analysis, the controller can distinguish between two types of operating conditions: high-frequency non-following and low-frequency conditions with phase lag. This provides a stable basis for subsequent tension feedforward parameter updates and reduces warp tension oscillations during acceleration and deceleration.

[0053] The system acquires warp feed servo status data, updates tension feedforward parameters based on response lag determination results, generates predicted disturbance data, generates torque compensation and speed compensation based on predicted disturbance data, updates dynamic thresholds to generate adjustment sequences, and drives the warp feed servo to keep the warp tension within the allowable deviation range of the warp tension set value.

[0054] In one embodiment, the controller acquires warp feed servo status data, including the warp feed servo motor current and speed. The controller updates the tension feedforward parameters based on the response hysteresis determination result. These parameters include the tension feedforward gain and phase compensation parameters. If the response hysteresis determination result indicates hysteresis, the tension feedforward gain is increased and the phase compensation parameters are adjusted according to the update rules configured in the parameter area. If the response hysteresis determination result indicates no hysteresis, the tension feedforward parameters are maintained or the controller reverts to a historical stable value according to the rollback rules. The controller performs trend extrapolation on the warp tension signal and the spindle speed signal to obtain predicted disturbance data, and converts this data into warp feed servo torque compensation and warp feed servo speed compensation. The controller updates the dynamic threshold based on the sliding average and fluctuation amplitude of the warp feed servo motor current, generates an adjustment sequence containing a warp feed servo speed compensation sequence and a warp feed servo torque compensation sequence, and sends it to the servo driver. This ensures that the deviation of the warp tension relative to the set warp tension value remains within the allowable deviation range, which is pre-configured in the controller's parameter area.

[0055] The tension feedforward parameters include tension feedforward gain and phase compensation parameters. The generation of predicted disturbance data includes trend extrapolation of the warp tension signal and the spindle speed signal to obtain the predicted disturbance data. The warp feed servo status data includes the warp feed servo motor current and the warp feed servo motor speed. The dynamic threshold update includes updating the adjustment threshold used to generate the torque distribution scheme and the hysteresis threshold used to obtain the response hysteresis determination result based on the warp feed servo status data. The adjustment sequence includes the speed compensation sequence and the torque compensation sequence.

[0056] In one embodiment, tension feedforward parameters are used to convert the deviation change of the warp tension signal into a compensation amount for the warp feed servo control loop in advance. The tension feedforward parameters include tension feedforward gain and phase compensation parameters. The tension feedforward gain is used to determine the conversion ratio from warp tension signal deviation to torque compensation, and the phase compensation parameter is used to adjust the timing of the compensation amount to compensate for the response lag of the warp feed servo control loop. After obtaining the response lag determination result, the controller updates the tension feedforward gain and phase compensation parameters based on the result. The update rules are pre-configured in the controller parameter area. The update process includes weighted fusion of the current parameters and historical stable parameters, and amplitude constraints are applied to the updated parameters to avoid oscillations in the warp feed servo torque adjustment command caused by parameter abrupt changes.

[0057] When generating predicted disturbance data, the controller caches the warp tension signal sequence and the spindle speed signal sequence within a preset prediction time window, and performs trend extrapolation on both types of signals. Trend extrapolation is implemented using a first-order differential extrapolation method. The controller calculates the predicted warp tension disturbance sequence and the predicted spindle speed disturbance sequence forward by a preset number of prediction steps, and uses both sequences together as the predicted disturbance data. Through this processing, the predicted disturbance data can reflect the trend of spindle speed changes and warp tension fluctuations in the short time domain, allowing for the early generation of compensation amounts rather than relying solely on feedback errors.

[0058] The warp feed servo status data includes the warp feed servo motor current and speed. The warp feed servo motor current is sampled by the servo driver and transmitted back to the controller, while the warp feed servo motor speed is transmitted back to the controller by the servo driver encoder or motor speed estimation module. When updating the dynamic threshold, the controller calculates the mean and fluctuation amplitude of the warp feed servo motor current and the mean and fluctuation amplitude of the warp feed servo motor speed within a preset time window. Based on these statistics, the controller updates the adjustment threshold used to generate the torque distribution scheme and the hysteresis threshold used to obtain the response hysteresis judgment result. This allows the threshold to adaptively adjust with changes in load level and execution capability, thereby reducing the probability of misjudgment under different weaving conditions using a fixed threshold.

[0059] The adjustment sequence includes a speed compensation sequence and a torque compensation sequence. The controller generates a speed compensation sequence based on the predicted spindle speed disturbance sequence, ensuring the warp feed servo speed is synchronized with the spindle trend during acceleration and deceleration. The controller also generates a torque compensation sequence based on the predicted warp tension disturbance sequence and updated tension feedforward parameters, injecting compensating torque into the warp feed servo before the warp tension deviation increases. The controller aligns the speed compensation sequence and torque compensation sequence according to the control cycle and synthesizes them into an adjustment sequence, which is then sent to the servo driver for execution, keeping the warp tension deviation relative to the setpoint within the allowable range.

[0060] like Figure 2 As shown, a tension-feedforward-based warp feed servo control system for a loom is used to implement the tension-feedforward-based warp feed servo control method for a loom. The system includes: The fluctuation detection module is used to acquire the spindle speed signal, warp tension signal, and warp tension setpoint. It performs load fluctuation detection on the spindle speed and warp tension signals to determine the required adjustment of the warp feed servo speed. The fluctuation detection module consists of a signal acquisition unit, a synchronous sampling unit, and an edge computing unit. The signal acquisition unit is electrically connected to the spindle rotary encoder to acquire the spindle speed signal and to the warp tension sensor to acquire the warp tension signal. The synchronous sampling unit, composed of an analog-to-digital converter and a clock synchronization circuit, samples the spindle speed and warp tension signals at the same time period and writes them to a circular buffer. The edge computing unit, composed of a microcontroller or digital signal processor, performs statistical calculations on fluctuation amplitude, tension deviation, and correlation within a preset time window to output the required adjustment of the warp feed servo speed. The warp tension setpoint is written to non-volatile memory by the human-machine interface board or host computer via an industrial communication interface and read by the edge computing unit.

[0061] The fitting and allocation module is used to fit the spindle speed signal to a speed curve to obtain acceleration / deceleration stage identifiers. Based on the acceleration / deceleration stage identifiers and the warp feed servo speed adjustment requirements, it generates a torque allocation scheme and, based on the torque allocation scheme and the warp tension signal, generates a warp feed servo torque adjustment command. The fitting and allocation module includes a computation acceleration unit, a parameter storage unit, and a servo command output unit. The computation acceleration unit, composed of a digital signal processor or a processor with a floating-point unit, performs speed curve fitting on the spindle speed signal and generates acceleration / deceleration stage identifiers. The parameter storage unit stores the fitting order, mapping table, and parameters related to the torque allocation scheme. The servo command output unit, composed of a pulse interface or an industrial Ethernet interface, sends the warp feed servo torque adjustment command to the servo driver in the form of a pulse sequence or bus message, while simultaneously limiting the amplitude and rate of change of the command to meet the rated capacity of the driver and the safety requirements of the mechanism.

[0062] The hysteresis determination module decomposes the warp tension signal into tension fluctuations to obtain the dominant fluctuation frequency. It then performs frequency matching between the dominant fluctuation frequency and the response characteristics of the warp feed servo control loop to obtain the response hysteresis determination result. The hysteresis determination module consists of a frequency domain analysis unit and a loop characteristic storage unit. The frequency domain analysis unit, composed of a digital signal processor or a processor with a Fast Fourier Transform instruction set, performs spectral analysis on the warp tension signal and extracts the dominant fluctuation frequency. The loop characteristic storage unit stores the response characteristic parameters of the warp feed servo control loop, which can be reported by the servo driver or written during the debugging phase. The hysteresis determination logic executes on the processor, performing frequency matching between the dominant fluctuation frequency and the response characteristics of the warp feed servo control loop and outputting the response hysteresis determination result.

[0063] The coordinated adjustment module acquires warp feed servo status data, updates tension feedforward parameters based on response lag determination results, generates predicted disturbance data, generates torque compensation and speed compensation based on the predicted disturbance data, updates the dynamic threshold to generate an adjustment sequence, and drives the warp feed servo to keep the warp tension within the allowable deviation range of the warp tension set value. The coordinated adjustment module includes a status acquisition unit, a prediction calculation unit, and a closed-loop output unit in hardware. The status acquisition unit acquires warp feed servo status data via a servo bus or analog interface, including at least the warp feed servo motor current and speed. The prediction calculation unit uses the processor to perform tension feedforward parameter updates and trend extrapolation calculations to generate predicted disturbance data, and converts the predicted disturbance data into torque compensation and speed compensation. The closed-loop output unit updates the dynamic threshold and generates an adjustment sequence, synthesizes the adjustment sequence with the torque adjustment command, and outputs it to the servo driver to drive the warp feed servo to keep the warp tension within the allowable deviation range of the warp tension set value.

[0064] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects.

[0065] The above are merely embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A servo control method for warp feed in a loom based on tension feedforward, characterized in that, Includes the following steps: The spindle speed signal, warp tension signal, and warp tension setpoint are acquired. Load fluctuation detection is performed on the spindle speed signal and warp tension signal to obtain the required adjustment of the warp feed servo speed. The spindle speed signal is fitted with a speed curve to obtain the acceleration and deceleration stage identifier. Based on the acceleration and deceleration stage identifier and the warp feed servo speed adjustment requirement, a torque distribution scheme is generated. Based on the torque distribution scheme and the warp tension signal, a warp feed servo torque adjustment command is generated. The tension fluctuation of the warp tension signal is decomposed to obtain the dominant fluctuation frequency. The dominant fluctuation frequency is matched with the response characteristics of the warp feed servo control loop to obtain the response hysteresis determination result. The system acquires warp feed servo status data, updates tension feedforward parameters based on response lag determination results, generates predicted disturbance data, generates torque compensation and speed compensation based on predicted disturbance data, updates dynamic thresholds to generate adjustment sequences, and drives the warp feed servo to keep the warp tension within the allowable deviation range of the warp tension set value.

2. The method according to claim 1, characterized in that, Load fluctuation detection of spindle speed signal and warp tension signal includes: statistically analyzing the change and fluctuation amplitude of spindle speed signal within a preset time window, and statistically analyzing the tension deviation and fluctuation amplitude of warp tension signal within a preset time window, and determining the required adjustment of warp feed servo speed based on the correlation coefficient between the change of spindle speed signal and the change of warp tension signal.

3. The method according to claim 2, characterized in that, The warp tension signal is acquired by the warp tension sensor. The fluctuation amplitude of the spindle speed signal is the difference between the maximum and minimum values ​​of the spindle speed signal within a preset time window. The fluctuation amplitude of the warp tension signal is the difference between the maximum and minimum values ​​of the warp tension signal within a preset time window. The tension deviation of the warp tension signal is the difference between the warp tension signal and the set warp tension value.

4. The method according to claim 1, characterized in that, The process of fitting the spindle speed signal to obtain acceleration / deceleration stage identifiers includes: performing polynomial fitting on the spindle speed signal to obtain a spindle speed trend curve, and determining the acceleration / deceleration stage identifiers based on the difference between adjacent sampling points of the spindle speed trend curve.

5. The method according to claim 1, characterized in that, The torque distribution scheme includes inertial compensation torque components, friction compensation torque components, and tension feedforward torque components, and sets torque amplitude constraints and torque change rate constraints for the servo torque adjustment commands sent to it.

6. The method according to claim 5, characterized in that, The generation of feed servo torque adjustment commands based on the torque distribution scheme includes: generating a torque compensation signal based on the torque distribution scheme; filtering the torque compensation signal and performing amplitude and slope limiting processing to obtain a smooth torque compensation signal; and performing feedback calibration on the smooth torque compensation signal based on the deviation between the feed servo torque observation value and the smooth torque compensation signal to output the feed servo torque adjustment command.

7. The method according to claim 1, characterized in that, Tension fluctuation decomposition of warp tension signals includes: performing spectral analysis on the warp tension signals to obtain the tension spectrum, and determining the dominant fluctuation frequency based on the main peak of the tension spectrum.

8. The method according to claim 1, characterized in that, The process of matching the dominant fluctuation frequency with the response characteristics of the warp feed servo control loop to obtain the response lag determination result includes: comparing the dominant fluctuation frequency with the response bandwidth of the warp feed servo control loop, determining the equivalent time delay based on the cross-correlation analysis of the warp tension signal and the warp feed servo torque adjustment command, and obtaining the response lag determination result based on the comparison of the equivalent time delay and the lag threshold.

9. The method according to claim 1, characterized in that, The tension feedforward parameters include tension feedforward gain and phase compensation parameters. The generation of predicted disturbance data includes trend extrapolation of the warp tension signal and the spindle speed signal to obtain the predicted disturbance data. The warp feed servo status data includes the warp feed servo motor current and the warp feed servo motor speed. The dynamic threshold update includes updating the adjustment threshold used to generate the torque distribution scheme and the hysteresis threshold used to obtain the response hysteresis determination result based on the warp feed servo status data. The adjustment sequence includes the speed compensation sequence and the torque compensation sequence.

10. A tension-feedforward-based warp feed servo control system for a loom, used to implement the tension-feedforward-based warp feed servo control method for a loom as described in any one of claims 1-9, characterized in that, The system includes: The fluctuation detection module is used to acquire the spindle speed signal, warp tension signal and warp tension set value, and to perform load fluctuation detection on the spindle speed signal and warp tension signal to obtain the required adjustment amount of the warp feed servo speed. The fitting and allocation module is used to fit the spindle speed signal to obtain the acceleration and deceleration stage identifier, generate a torque allocation scheme based on the acceleration and deceleration stage identifier and the warp feed servo speed adjustment requirement, and generate a warp feed servo torque adjustment command based on the torque allocation scheme and the warp tension signal. The hysteresis determination module is used to decompose the tension fluctuation of the warp tension signal to obtain the dominant fluctuation frequency, and then perform frequency matching between the dominant fluctuation frequency and the response characteristics of the warp feed servo control loop to obtain the response hysteresis determination result. The collaborative adjustment module is used to acquire warp feed servo status data, update tension feedforward parameters based on response lag judgment results and generate predicted disturbance data, generate torque compensation and speed compensation based on predicted disturbance data, update dynamic threshold to generate adjustment sequence, and drive warp feed servo to keep warp tension within the allowable deviation range of warp tension set value.

Images