Self-adaptive control system and method for spinning heald frame

Abstract

The invention relates to the technical field of heald frame self-adaptive control, and discloses a spinning heald frame self-adaptive control system and method.The spinning heald frame self-adaptive control system comprises a sensor set used for collecting heald frame displacement, yarn tension and fabric surface state data in real time and generating heald frame state data, and the input end of a defect detection module is connected with the sensor set; receiving the checked yarn tension and fabric surface state data, performing defect detection on the fabric surface state data, generating a suppression signal for abrupt change of heald frame acceleration according to correlation analysis of a fabric state rating signal and real-time displacement, dynamically deciding and receiving a defect identification result, querying a defect type mapping table, and determining whether the defect type mapping table is abnormal or not. According to the method, a control parameter forced priority sequence is determined, weft errors and tension changes on the surface of the fabric are judged, self-adaptive adjustment of the spinning heald frame through weaving state detection is achieved, and weaving accuracy is improved.

Description

Technical Field

[0001] The invention relates to the technical field of heald frame adaptive control and discloses a textile heald frame adaptive control system and method. Background Art

[0002] Existing systems mostly use displacement sensors to independently monitor the displacement of the heald frames, and lack the synchronous collection of yarn tension and fabric surface conditions. When the yarn material changes, the machine wears out, or the warp breaks, it is impossible to identify tension anomalies or fabric defects (such as broken or loose wefts) based on displacement feedback alone, resulting in delayed quality control and easily causing batch defects. Fabric defect detection mostly relies on offline manual sampling or independently operated visual systems. The detection results cannot be fed back to the motion control end in real time, and the control parameter priority is fixed (such as always prioritizing tension stability). In high-speed weaving scenarios, when a sudden defect occurs on the fabric surface (such as weft misalignment), the system cannot dynamically adjust the priority, resulting in a delayed defect repair response or over-adjustment causing yarn breakage. Traditional PID control is difficult to simultaneously compensate for the displacement tracking error, tension mutation, and the associated disturbances of fabric defects. Summary of the Invention

[0003] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of this application to avoid obscuring the purpose of this section, the abstract and the title of the invention, and such simplifications or omissions should not be used to limit the scope of the present invention.

[0004] In order to solve the above technical problems, the main purpose of the present invention is to provide a textile heald frame adaptive control system, wherein a textile heald frame adaptive control system comprises:

[0005] The sensor group includes a displacement sensor, a pressure sensor, and an image sensor, which are used to collect the displacement of the heald frame, the yarn tension, and the surface state data of the fabric in real time, and generate the heald frame state data;

[0006] The input end of the defect detection module is connected to the multi-sensor group, receives the verified yarn tension and fabric surface state data, and performs defect detection on the fabric surface state data;

[0007] Generate suppression signals for sudden acceleration changes of heald frame based on correlation analysis between fabric status rating signals and real-time displacement;

[0008] The dynamic decision-making receives the defect identification results, queries the defect type mapping table, determines the mandatory priority sequence of the control parameters, and judges the weft yarn errors and tension changes on the fabric surface.

[0009] As a preferred solution of the textile heald frame adaptive control system of the present invention, wherein:

[0010] The sensor group is also provided with multi-sensor redundancy check;

[0011] Perform correlation check on the displacement sensor data and the warp and weft density of the fabric calculated by the image sensor. If the deviation exceeds the threshold, the displacement sensor calibration is triggered.

[0012] Based on the spatial distribution data of the pressure sensing array, the local tension abnormality area is identified and the area is directed and focused on by the image sensor;

[0013] The heald frame displacement, local tension and texture feature data are integrated to generate the heald frame status data.

[0014] As a preferred solution of the textile heald frame adaptive control system of the present invention, wherein:

[0015] The defect detection module includes a synchronous trigger unit, which is connected to the heald frame motion controller and is used to obtain the rising and falling edges of the heald frame motion cycle to trigger image capture, so that the image frame rate is synchronized with the heald frame motion frequency in integer multiples;

[0016] The defect detection module is also connected to the image processing unit, which outputs the fabric defect enhanced image, and the defect position mapper generates a fabric status signal including the defect type, defect size and corresponding heald frame position coordinates.

[0017] As a preferred solution of the textile heald frame adaptive control system of the present invention, wherein:

[0018] The image processing unit receives an original fabric surface image at its input end, performs motion correction on the image based on the heald frame displacement feedback signal, eliminates image smear caused by the high-speed movement of the heald frame through reverse displacement compensation, enhances fabric texture contrast through adaptive histogram equalization, and crops a region of interest in the weft yarn overlap area;

[0019] The image processing unit also generates a defect feature vector at the output end by receiving the preprocessed image of the region of interest, and outputs a weaving defect type label based on the defect feature vector, including yarn breakage, loose fibers, misaligned fibers and weaving defects.

[0020] As a preferred solution of the textile heald frame adaptive control system of the present invention, wherein:

[0021] The correlation analysis input terminal receives the defect type, defect size and corresponding heald frame position coordinates in the fabric status signal respectively, and binds the heald frame position coordinates at the time the defect occurs with the acceleration value obtained by the displacement sensor at the corresponding time through timestamp matching;

[0022] Based on the displacement of the heald frame, the local stress distribution of the defect position during the acceleration mutation is obtained. The defect-acceleration correlation mapping table generated by pre-stored textile process parameters and historical data training includes:

[0023] When the weft yarn misalignment defect is in the top area of the heald frame and the confidence level is not less than the maximum defect threshold, the associated acceleration mutation alarm is triggered;

[0024] The priority weight coefficient for triggering acceleration suppression when the yarn break defect is in the middle area of the heald frame.

[0025] As a preferred solution of the textile heald frame adaptive control system of the present invention, wherein:

[0026] The mutation suppression signal generator, whose input end is connected to the spatiotemporal alignment unit and the association rule base, executes:

[0027] When the real-time acceleration value is detected to exceed the mutation threshold associated with the current defect type and position, an amplitude-adjustable suppression signal is generated according to the priority weight coefficient;

[0028] Perform phase pre-compensation on the suppression signal so that it is within the safe motion cycle before the acceleration mutation occurs and acts on the servo motor drive end;

[0029] The output end superimposes the suppression signal with the linear output signal of the heald frame controller to control the movement of the heald frame.

[0030] As a preferred solution of the textile heald frame adaptive control system of the present invention, wherein:

[0031] The dynamic decision input terminal receives the defect type label, confidence score and real-time collected tension, heald frame displacement and texture data in the output result of the defect position mapper;

[0032] The risk factor is obtained by the difference between the real-time threshold of the preset priority weight and the critical threshold and the difference between the critical threshold, and a forced priority sequence of tension, heald frame displacement and texture is forcibly generated. A protection latch instruction is generated according to the forced priority sequence, and a parameter adjustment instruction is generated.

[0033] As a preferred solution of the textile heald frame adaptive control system of the present invention, wherein:

[0034] The parameter adjustment instructions include heald frame movement adjustment, heald frame displacement adjustment, opening height adjustment and local tension adjustment;

[0035] When a weft yarn misalignment defect occurs on the fabric surface, the speed reduction and height adjustment linkage strategy is triggered. While maintaining the opening height not lower than the minimum value of the process constraint, the speed of the heald frame movement is reduced until the defect is eliminated.

[0036] When the tension change rate exceeds the critical threshold, the emergency elastic buffer mode is activated, the yarn break risk period is predicted through the real-time data of the heald frame displacement, and the tension threshold is dynamically lowered.

[0037] As a preferred solution of the textile heald frame adaptive control system of the present invention, wherein:

[0038] When the coupled disturbance compensator detects a sudden acceleration change, the output signal of the linear control model is asymmetrically superimposed with the suppression signal to generate a pulse command to drive the servo motor;

[0039] Taking fabric condition rating as the main optimization objective, a fuzzy weight distribution algorithm is embedded in the PID control framework to adjust the weight coefficients of opening height, movement speed and tension threshold in real time.

[0040] As a preferred solution of the textile heald frame adaptive control method of the present invention, wherein:

[0041] The sensor group acquires the position of the heald frame, yarn tension and fabric surface status data in real time, generates heald frame status data, and pre-processes the acquired data;

[0042] Conduct defect detection on the fabric surface and analyze the type and size of weaving defects based on the defect detection results, including yarn breakage, loose fibers, misaligned fibers, and other weaving defects;

[0043] Receive the defect type, defect size and corresponding heald frame position coordinates in the fabric status signal, match the timestamp, perform correlation analysis, and output the defect-acceleration correlation mapping table;

[0044] The defect-acceleration association mapping table is used to prioritize the acceleration mutation of the defect position, weft yarn misalignment and yarn breakage defects, and the heald frame movement is controlled based on the ranking results.

[0045] Beneficial effects of the present invention:

[0046] Through the collaborative acquisition of displacement + tension + image multi-sensors, multi-dimensional data of heald frame displacement, yarn tension and fabric surface status can be obtained in real time, overcoming the blind spots of traditional single sensor monitoring and enabling the system to simultaneously identify mechanical motion deviations, tension anomalies and fabric defects.

[0047] The defect type-priority mapping table and dynamic calculation of multi-factor weights overcome the rigid decision-making issues of static rule bases. During high-speed weaving, the control parameter priority is automatically adjusted according to real-time working conditions. The fabric defect location is dynamically bound to the sudden acceleration change of the heald frame, and the influence of mechanical delay is offset by the phase advance suppression signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them:

[0049] Figure 1 This is a flow chart of a textile heald frame adaptive control system of the present invention;

[0050] Figure 2 The present invention provides a method for predicting and dynamically adjusting yarn break risk in a textile heald frame adaptive control system;

[0051] Figure 3 This is a process for implementing a textile heald frame adaptive control method of the present invention;

[0052] Figure 4 This is a schematic diagram of the heald frame structure of a textile heald frame adaptive control system of the present invention;

[0053] Table 1 shows the weight distribution logic of a textile heald frame adaptive control system. DETAILED DESCRIPTION

[0054] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0055] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0056] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0057] Example 1

[0058] like Figure 1 As shown, a textile heald frame adaptive control system includes:

[0059] The sensor group includes a displacement sensor, a pressure sensor, and an image sensor, which are used to collect the displacement of the heald frame, the yarn tension, and the surface state data of the fabric in real time, and generate the heald frame state data;

[0060] The sensor group is also provided with multi-sensor redundancy check;

[0061] Perform correlation check on the displacement sensor data and the warp and weft density of the fabric calculated by the image sensor. If the deviation exceeds the threshold, the displacement sensor calibration is triggered.

[0062] Based on the spatial distribution data of the pressure sensing array, the local tension abnormality area is identified and the area is directed and focused on by the image sensor;

[0063] The displacement, pressure and texture feature data are integrated to generate the heald frame status data.

[0064] The high frame rate image sensor comprises:

[0065] A synchronous trigger module, whose input end is connected to the periodic signal output end of the heald frame motion controller, is used to trigger image capture according to the rising / falling edge of the heald frame motion cycle, so that the image frame rate is synchronized with the integral multiple of the heald frame motion frequency;

[0066] The image preprocessing unit receives the original fabric surface image at the input and performs the following operations:

[0067] The image is motion-blurred based on the heald frame displacement feedback signal, and the image smear caused by the high-speed movement of the heald frame is eliminated through the reverse displacement compensation algorithm;

[0068] Adaptive histogram equalization is used to enhance the fabric texture contrast, and the region of interest in the weft overlap area is cropped. A lightweight CNN inference module is set to preprocess the region of interest map and output the defect feature vector.

[0069] Furthermore, the CNN model includes three convolutional layers and one global average pooling layer. The last layer is connected to a Softmax classifier to output defect type labels (including broken yarn, loose weft, and wrong weft) and confidence scores.

[0070] The model weights are initialized through transfer learning, and the training dataset contains augmented images labeled with multiple types of fabric defects under different lighting conditions;

[0071] The defect position mapper synchronously receives the defect type label output by CNN and the real-time position signal of the displacement sensor at the input end, and generates a fabric condition rating signal including the defect type, confidence score and the corresponding heald frame position coordinates at the output end.

[0072] The input end of the defect detection module is connected to the multi-sensor group, receives the verified yarn tension and fabric surface state data, and outputs nonlinear phase compensation for the yarn tension data;

[0073] Generate suppression signals for sudden acceleration changes of heald frame based on spatiotemporal correlation analysis between fabric status rating signals and real-time displacement

[0074] The dynamic decision-making receives the defect identification results, queries the defect type mapping table, determines the mandatory priority sequence of the control parameters, and judges the weft yarn errors and tension changes on the fabric surface.

[0075] A specific implementation method of a sensor group and a defect detection module includes:

[0076] A periodic signal acquisition line is connected in parallel to the output of the heald frame motion controller (such as a servo motor driver) and connected to the input interface of the synchronization trigger unit. This signal is a pulse signal reflecting the heald frame motion cycle (for example, a rising edge corresponds to the heald frame reaching top dead center, and a falling edge corresponds to bottom dead center). It can be generated by differentially analyzing the orthogonal signals of an encoder or displacement sensor.

[0077] The trigger input terminal of the image sensor (such as a linear array CCD or an area array CMOS) is connected to the output terminal of the synchronous trigger unit, supporting an external hardware trigger mode (such as a GPIO trigger).

[0078] The displacement sensor (such as grating scale or magnetic grating) collects the position of the heald frame in real time and transmits it synchronously to the image processing unit and the heald frame motion controller through the RS485 or Ethernet interface.

[0079] The image processing unit and defect mapping module adopt an industrial-grade embedded computing platform (such as NVIDIA Jetson or ARM industrial computer) with built-in high-speed DDR memory and FPGA acceleration module to support real-time image preprocessing and defect detection.

[0080] The defect position mapper is an independent logic circuit or software module. Its input includes the heald frame position coordinates at the time of image acquisition (from the displacement sensor) and the pixel coordinates of the defect in the image, and it outputs a physical coordinate mapping signal (such as the warp / weft position offset).

[0081] Synchronous triggering and image acquisition methods include:

[0082] Each time the heald frame completes a motion cycle (such as the opening-closing process), the heald frame motion controller outputs a pair of pulse signals (rising edge and falling edge) to the synchronization trigger unit, representing the phase of the heald frame movement (such as the rising edge corresponds to the start of the heald frame rising, and the falling edge corresponds to the start of falling).

[0083] The synchronous trigger unit has a built-in frequency phase-locked loop (PLL) that calculates the movement frequency of the heald frame in real time (such as 100-800 times / minute) and sets the image sensor frame rate to an integer multiple of the movement frequency (such as 2 times or 4 times), ensuring that a fixed number of images are captured in each movement cycle (such as 2 frames in the rising phase and 2 frames in the descending phase), avoiding image acquisition misalignment during movement.

[0084] Image acquisition is triggered during the uniform speed phase (not acceleration or deceleration) of the heald frame's motion, when the frame's speed is stable, minimizing motion blur. For example, after the synchronous trigger unit detects a rising edge signal, it delays for a fixed time (determined by displacement sensor feedback when the frame enters the uniform speed phase) before triggering the first acquisition. Subsequent acquisitions are then continued at a fixed interval (e.g., 1 / frame rate).

[0085] Image processing and defect detection implementation methods include:

[0086] Each time an image is captured, the real-time position value of the displacement sensor (such as the current height coordinate of the heald frame) is recorded synchronously, and the position difference (Δx, Δy) with the previous frame is calculated. Δx is the horizontal displacement difference, and Δy is the vertical displacement difference, which is used as the basis for compensating motion blur.

[0087] For each pixel in the original image, based on the direction of motion of the heald frame (determined by the sign of the displacement difference) and the speed (Δx / Δt), where Δt is the time difference, the pixel is shifted in the opposite direction of motion by an equivalent displacement (for example, if the heald frame moves 5 mm to the right, the image pixel is shifted to the left by the corresponding number of pixels). Bilinear interpolation is used to fill the blank areas and eliminate smear.

[0088] The region of interest cropping is based on prior knowledge of the fabric structure (e.g., the weft yarn overlapping area is located at a fixed position in the middle of the image). Through geometric transformation (rotation, scaling), the image is corrected to a standard perspective, and the weft yarn overlapping area is cropped (e.g., a horizontal strip with a height of 1 / 3 of the image and a width covering the entire width), reducing the amount of subsequent processing calculations.

[0089] The region of interest is divided into 8×8 pixel sub-blocks, and the histogram is calculated and equalized for each sub-block to enhance the local texture contrast (such as the grayscale difference of yarn hairiness and breakage) while avoiding noise amplification caused by global equalization.

[0090] Through the preprocessed region of interest image, features are extracted, including: reflecting the grayscale correlation between pixels, detecting texture uniformity, calculating edge continuity by identifying yarn edges (broken yarn is manifested as edge fracture), and counting the area and shape of connected regions by removing noise (distorted dimension is manifested as abnormal connected regions).

[0091] Based on the preprocessed feature vector, a trained classification model (such as a support vector machine or random forest, with labeled defect samples such as broken yarn, loose fibers, and misaligned fibers used for offline training) is used to output the defect type label and confidence score (for example, a confidence score > 80% is considered a valid defect).

[0092] The defect position mapper receives two parts of data: the pixel coordinates (u, v) of the defect in the area of interest and the real-time coordinates (X, Y) from the displacement sensor, where X is the horizontal position of the heald frame (corresponding to the warp direction of the fabric) and Y is the height (corresponding to the weft direction). Through a pre-calibrated mapping relationship (such as 1 pixel = 0.1mm, based on the lens focal length and object distance calibration), the pixel coordinates are converted into physical coordinates on the fabric surface (such as warp offset ±5mm, weft position Y ±2mm), generating a status rating signal containing the defect type, confidence level and physical location.

[0093] By setting up the coordination of synchronous triggering and motion compensation, the inherent contradiction between image acquisition and analysis in high-speed textile scenarios is resolved; targeting the local characteristics of textile defects (such as weft yarn misalignment manifested as short line anomalies), redundant network layers are reduced to ensure accuracy while meeting real-time requirements; the defect position is mapped to the coordinates of the heald frame, so that subsequent control strategies can locate and repair it (such as speed reduction at a specific location) rather than global parameter adjustments, reducing efficiency losses.

[0094] Example 2

[0095] A textile heald frame adaptive control system, the specific implementation method also includes:

[0096] Generate suppression signals for heald frames based on correlation analysis between fabric status rating signals and real-time displacement;

[0097] The correlation analysis input terminal receives the defect type, defect size and corresponding heald frame position coordinates in the fabric status signal respectively, and binds the heald frame position coordinates at the time the defect occurs with the acceleration value obtained by the displacement sensor at the corresponding time through timestamp matching;

[0098] Based on the displacement of the heald frame, the local stress distribution of the defect position during the acceleration mutation is obtained. The defect-acceleration correlation mapping table generated by pre-stored textile process parameters and historical data training includes:

[0099] When the weft yarn misalignment defect is in the top area of the heald frame and the confidence level is not less than the maximum defect threshold, the associated acceleration mutation alarm is triggered;

[0100] The priority weight coefficient for triggering acceleration suppression when the yarn break defect is in the middle area of the heald frame.

[0101] A specific implementation method for generating a suppression signal for a heald frame based on correlation analysis between a fabric state rating signal and real-time displacement includes:

[0102] A rotary encoder is installed on the heald frame drive shaft to generate a pulse signal synchronized with the motion cycle (e.g., one pulse per 10° of rotation). This pulse signal is used as an external trigger source and simultaneously connected to a high-frame-rate image sensor and displacement sensor to ensure that the image capture moment is strictly aligned with the clock reference of the displacement / acceleration data, eliminating timestamp errors.

[0103] After the image sensor captures the fabric surface image, the pixel coordinates of the defect in the image (such as the weft yarn misalignment point) are converted into the corresponding physical position coordinates of the heald frame (such as 15 cm ± 2 mm from the top of the heald frame) through the coordinate system conversion relationship pre-established by the mechanical calibration plate.

[0104] A simplified beam bending model is established based on the mechanical structural parameters of the heald frame (such as material and cross-sectional moment of inertia). When the displacement sensor detects a sudden acceleration change at a certain location on the heald frame, the theoretical stress increment (Δσ) at the defect location is quickly calculated based on this model. If Δσ exceeds 30% of the material's yield strength (an empirical safety threshold), it is considered a high-risk disturbance.

[0105] During the equipment debugging phase, typical defect conditions (such as yarn breakage and weft misalignment) are simulated and the corresponding acceleration waveform characteristics (such as peak value and rising slope) are recorded. The acceleration thresholds of each defect type at different positions (top / middle) are determined through statistical analysis, for example:

[0106] Weft misalignment (top): acceleration threshold is set to 1200 rad / s 2 , confidence level ≥ 85%.

[0107] Yarn breakage (middle section): acceleration threshold is set to 900 rad / s 2 , confidence level ≥ 90%.

[0108] The above thresholds and priority weights (such as yarn break weight coefficient 0.7, misalignment 0.5) are written into the association mapping table of the process knowledge base.

[0109] When a fabric defect is detected, the system automatically extracts the defect type, location, and confidence level, then queries a correlation mapping table to obtain the corresponding acceleration threshold and weight. Simultaneously, it reads the acceleration value measured by the displacement sensor. If it exceeds the threshold and the confidence level meets the requirements, a suppression signal is triggered.

[0110] Parameterized output of suppression signals including weft misalignment and yarn breakage:

[0111] The weft yarn misalignment (top) generates a negative pulse signal with an amplitude of 1.2 times the reference value and lasting for 2 motion cycles, forcing the acceleration of the top of the heald frame to be reduced;

[0112] The broken yarn (middle section) generates a bidirectional oscillation suppression signal with an amplitude of 1.5 times and a duration of 3 cycles, which gently releases the stress concentration in the middle section.

[0113] According to the response delay of the servo motor (such as the measured average lag of 50ms), advance time compensation (such as output 10ms in advance) is superimposed on the suppression signal to ensure that the suppression action is synchronized with the acceleration mutation.

[0114] Furthermore, priority weights are dynamically enforced, including conflict arbitration and single-line protection;

[0115] Furthermore, if multiple defects trigger suppression signals within the same time period (e.g., top misalignment and mid-segment yarn breakage), the suppression signal strength is allocated based on priority weights. For example, a yarn breakage weight of 0.7 results in a 70% suppression signal; a misalignment weight of 0.5 results in a 30% suppression signal. The final output signal is a weighted superposition of the two.

[0116] When the inhibition signal of the yarn break defect is continuously triggered for more than 5 cycles, the "speed reduction-pressure reduction" linkage strategy is automatically activated: the loom speed is reduced to 70% of the rated value, and the tension threshold is simultaneously relaxed to 80% of the critical value to prevent secondary yarn breakage.

[0117] For example, when the loom is running at 600 rpm, the image sensor detects the top weft yarn is misaligned (confidence level 88%), and the displacement sensor shows that the acceleration at that position suddenly increases to 1250 rad / s 2 After the system matches the mapping table, it outputs a suppression signal, and the acceleration drops back to 1100 rad / s within 2 cycles. 2 , the defects were repaired successfully and the fabric surface was flawless.

[0118] If the priority ranking of acceleration suppression is triggered, the dynamic decision input terminal receives the defect type label in the output result of the defect position mapper and the tension, heald frame displacement and texture data collected in real time;

[0119] The risk factor is obtained by the difference between the real-time threshold of the preset priority weight and the critical threshold and the critical threshold, and a forced priority sequence of tension, heald frame displacement and texture is forcibly generated. A protection latch instruction is generated according to the forced priority sequence, and a heald frame parameter adjustment instruction is generated.

[0120] When the real-time acceleration value is detected to exceed the mutation threshold associated with the current defect type position, an amplitude-adjustable suppression signal is generated according to the priority weight coefficient;

[0121] Perform phase pre-compensation on the suppression signal so that it is within the safe motion cycle before the acceleration mutation occurs and acts on the servo motor drive end;

[0122] The output end superimposes the suppression signal with the linear output signal of the heald frame controller to control the movement of the heald frame.

[0123] A specific implementation method includes:

[0124] Receive the currently detected fabric defect type (such as yarn breakage, texture abnormality, etc.) from the "Defect Location Mapper" in real time to clarify the type of problem that needs to be dealt with.

[0125] Three key operating parameters are obtained in real time through sensors:

[0126] Tension test: whether the tension of yarn is stable during weaving;

[0127] Heald frame displacement monitoring: whether the position of the heald frame (the device that controls the up and down movement of the yarn) is as expected;

[0128] Texture data collects the surface texture of the fabric through images or sensors to determine whether there are any defects.

[0129] Two sets of thresholds are set for the three parameters mentioned above (tension, heald frame displacement, texture):

[0130] “Safety limits” for critical threshold parameters, exceeding which may cause equipment failure or fabric defects;

[0131] The real-time threshold is a dynamically adjusted “alert value” based on the current defect type and equipment operating status.

[0132] Calculate the risk level by comparing the parameter values collected in real time with the above thresholds to determine the "risk level" of each parameter (for example, the more the tension exceeds the real-time threshold, the higher the risk).

[0133] Determine the processing order: Generate a mandatory priority sequence of the three parameters from high to low risk (for example: current tension has the highest risk and is processed first; followed by heald frame displacement and finally texture).

[0134] Generating control instructions includes protection and regulation instructions and inhibiting signal generation.

[0135] The protection and adjustment instructions are based on the priority sequence, first generating protection latch instructions for high-risk parameters (such as locking the current tension control parameters to prevent abnormal fluctuations);

[0136] At the same time, heald frame parameter adjustment instructions are generated (such as adjusting the heald frame movement speed and amplitude to adapt to the processing requirements of the current defect type).

[0137] Suppression signal generation: When it is detected that the real-time acceleration of the heald frame exceeds the "mutation threshold" associated with the current defect type (i.e., a sudden abnormal increase in acceleration that may cause equipment impact or fabric damage), an adjustable intensity suppression signal is generated based on the priority level (the higher the risk parameter, the greater the signal intensity).

[0138] The generated inhibition signal is adjusted in "time advance" to ensure that the signal takes effect within the safe motion cycle before the acceleration is about to suddenly change (for example: if it is predicted that the acceleration may exceed the limit at the next moment, the signal is sent in advance at the end of the current cycle to reserve reaction time).

[0139] The compensated signal is sent to the servo motor drive end to prepare for intervention in the movement of the heald frame.

[0140] The suppression signal is superimposed on the original linear output signal of the heald frame controller (the control signal during normal operation) to form the final control signal.

[0141] The superimposed signals are used to adjust the operating status of the servo motor in real time, such as reducing acceleration and correcting the displacement trajectory, to ensure that the heald frame moves smoothly near the defect location, avoiding equipment damage or fabric quality problems caused by sudden acceleration changes.

[0142] Dynamic decision-making receives defect identification results, queries the defect type mapping table, determines the mandatory priority sequence of control parameters, judges weft yarn errors and tension changes on the fabric surface, and generates control instructions based on the judgment results.

[0143] The parameter adjustment instructions include heald frame movement adjustment, heald frame displacement adjustment, opening height adjustment and local tension adjustment;

[0144] When a weft yarn misalignment defect occurs on the fabric surface, the speed reduction and height adjustment linkage strategy is triggered. While maintaining the opening height not lower than the minimum value of the process constraint, the speed of the heald frame movement is reduced until the defect is eliminated.

[0145] When the tension change rate exceeds the critical threshold, the emergency elastic buffer mode is activated, the yarn break risk period is predicted through the real-time data of the heald frame displacement, and the tension threshold is dynamically lowered.

[0146] When the coupled disturbance compensator detects a sudden acceleration change, the output signal of the linear control model is asymmetrically superimposed with the suppression signal to generate a pulse command to drive the servo motor;

[0147] Taking fabric condition rating as the main optimization objective, a fuzzy weight distribution algorithm is embedded in the PID control framework to adjust the weight coefficients of opening height, movement speed and tension threshold in real time.

[0148] A specific implementation method of parameter adjustment includes:

[0149] Deploy visual inspection systems or sensors to monitor the fabric surface in real time and identify weft yarn misalignment defects. When defects are detected continuously for more than a preset number of times (such as 3 times / minute), a linkage strategy is triggered.

[0150] The dynamic parameter adjustment process includes:

[0151] Read the current opening height value. If it is higher than the minimum value of the process constraint, gradually increase the opening height in a gradient (such as increasing by 0.2mm each time) until the maximum value allowed by the process is reached or the defect disappears.

[0152] The movement speed of the heald frame is reduced synchronously (e.g., by 5% each time), and a smooth transition is achieved through the PID control algorithm to avoid mechanical shock.

[0153] Continuously monitor the defect status. If the defect is eliminated, record the current parameter combination as a temporary process. If the opening height has reached the minimum value, only reduce the speed to the equipment's safety lower limit.

[0154] After the defects are eliminated, the original speed is gradually restored (increased by 3% each time) while keeping the opening height unchanged to ensure production stability

[0155] The emergency elastic buffer mode for tension overrun uses a high-precision tension sensor to collect data at a frequency of 100Hz and calculate the tension change rate (ΔT / Δt). When the change rate exceeds a critical threshold (such as 5N / s), the emergency mode is triggered.

[0156] like Figure 2 As shown: The yarn break risk prediction and dynamic adjustment method includes:

[0157] Acquire the heald frame displacement time series data in real time and predict the displacement fluctuation trend within the next 3 seconds.

[0158] If the predicted displacement deviation exceeds ±0.5mm, it is marked as a high-risk period and elastic buffering is activated.

[0159] The tension threshold is temporarily lowered to 80% of its original value, and the warp tension is dynamically adjusted through fuzzy logic control to ensure a smooth transition.

[0160] During the risk period, the displacement compensation of the heald frame is adjusted in a linked manner (such as fine-tuning ±0.1mm) to balance the tension, while reducing the acceleration of the heald frame and reducing sudden loads.

[0161] Hardware for parameter adjustment: servo motor (controls the movement and displacement of the heald frame), linear encoder (displacement feedback), tension sensor (±0.1N accuracy), industrial camera (defect detection). Model predictive control (MPC) is used to coordinate multi-parameter adjustment, and an adaptive learning module is embedded to optimize thresholds and response speeds.

[0162] Set a hard stop threshold (such as tension exceeding 120% of the equipment limit) to prevent mechanical damage.

[0163] A specific implementation method of disturbance compensation and signal control:

[0164] A three-axis acceleration sensor with high-frequency sampling (≥1000 times / second) is installed on the heald frame drive shaft to monitor mechanical vibration in real time. When the acceleration is detected to change by more than 0.5 times the acceleration of gravity within 1 millisecond, it is judged as mechanical impact or rebound.

[0165] Asymmetric compensation strategies include positive shocks and negative rebounds;

[0166] Among them, the positive impact (such as heald frame overshoot) generates a fast-decaying suppression signal, which preferentially weakens the impact of the sudden load on the motor.

[0167] Negative rebound (such as loose mechanism) adopts stepped suppression signal to reduce the response intensity for small jitter parts.

[0168] A specific implementation method of fabric state-driven fuzzy weighted PID control includes:

[0169] The uniformity of weft yarn arrangement is analyzed by industrial cameras (accounting for 40%).

[0170] Calculate the standard deviation of tension fluctuations in the last 10 seconds (accounting for 30%).

[0171] Statistical analysis of the deviation between the actual displacement of the heald frame and the target value (accounting for 30%).

[0172] The scoring rules include: ideal state (90-100 points): the weft yarn is uniform, the tension fluctuation is <1N, and the displacement deviation is <0.1mm; warning state (60-89 points): the single indicator exceeds the normal range but does not trigger a shutdown; abnormal state (<60 points): the speed reduction protection is immediately activated.

[0173] The weight distribution logic is shown in the following table:

[0174]

[0175] Table 1 Weight distribution logic

[0176] Furthermore, in each control cycle (10 ms), the control strength of the PID controller on the three parameters of opening height, movement speed, and tension threshold is dynamically allocated according to the latest weight, and the total weight is maintained at 100%.

[0177] The three-level response mechanism includes a millisecond execution layer, a ten-millisecond decision layer, and a hundred-millisecond monitoring layer;

[0178] The millisecond-level execution layer (1ms) is responsible for real-time operations such as motor pulse generation and emergency stop triggering, ensuring precise synchronization of physical actions.

[0179] The ten-millisecond decision layer (10ms) runs fuzzy weight calculations, PID parameter refreshes, and dynamically adjusts the control strategy.

[0180] The 100-millisecond monitoring layer (100ms) updates fabric scores, stores operational data, and warns of potential risks.

[0181] When a mechanical shock is detected, the weight adjustment is immediately frozen for 1 control cycle (10ms) to prioritize the disturbance.

[0182] Add a pre-compensation term for the acceleration change rate in the motor control command to offset the inertial impact in advance.

[0183] Example 3

[0184] like Figure 3 As shown, a textile heald frame adaptive control system also includes:

[0185] The sensor group acquires the position of the heald frame, yarn tension and fabric surface status data in real time, generates heald frame status data, and pre-processes the acquired data;

[0186] Conduct defect detection on the fabric surface and analyze the type and size of weaving defects based on the defect detection results, including yarn breakage, loose fibers, misaligned fibers, and other weaving defects;

[0187] Receive the defect type, defect size and corresponding heald frame position coordinates in the fabric status signal, match the timestamp, perform correlation analysis, and output the defect-acceleration correlation mapping table;

[0188] The defect-acceleration association mapping table is used to prioritize the acceleration mutation of the defect position, weft yarn misalignment and yarn breakage defects, and the heald frame movement is controlled based on the ranking results.

[0189] Example 3

[0190] like Figure 4 As shown, a textile heald frame adaptive control system also includes a textile heald frame structure schematic diagram.

[0191] The specific schematic diagram of the textile heald frame structure includes: a rotating screw 1 for controlling the movement of the rotating wheel 3, the rotating wheel 3 and the rotating screw 1 are combined to control the movement of the heald frame, the spare port 7 is used to observe the internal movement of the textile heald frame and provide a maintenance window when necessary, the vent 2 is used to dissipate the heat of the textile heald frame movement to prevent excessive heat from mechanical movement causing mechanical damage to the structure, and the power socket 4 is used to connect an external power cord to provide energy for the movement of the textile heald frame.

[0192] Furthermore, the flange panel 5 is a flange with flexible size, and the flange panel 5 is used to provide an interface panel for the manual operating rod 6. The manual operating rod 6 is used to provide a manual operating panel to stop the movement of the textile heald frame through manual operation when the adaptive control of the textile heald frame fails or malfunctions.

[0193] The control panel 3 is provided with a heald frame control communication interface, a heald frame data acquisition communication interface and a reset interface in sequence.

[0194] Furthermore, the heald frame control communication interface is used to receive heald frame motion control strategies or control signals.

[0195] The heald frame data acquisition communication interface is provided with sensors and other components for real-time acquisition of heald frame motion parameters and weaving parameters.

[0196] The reset interface is used to manually reset the heald frame movement when an adaptive control failure occurs.

[0197] Furthermore, components such as a drive motor are provided inside the textile heald frame structure for driving and controlling the movement of the heald frame.

[0198] It is important to note that the construction and arrangement of the present application shown in a number of different exemplary embodiments are merely illustrative. Although only two embodiments are described in detail in this disclosure, it should be readily understood by those who refer to this disclosure that, without departing substantially from the novel teachings and advantages of the subject matter described in this application, many modifications are possible, for example, the size, scale, structure, shape and proportion of various elements, as well as parameter values (e.g., temperature, pressure, etc.), mounting arrangements, use of materials, color, directional changes, etc. For example, an element shown as integrally formed can be composed of multiple parts or elements, the position of the element can be inverted or otherwise changed, and the nature or number or position of the discrete elements can be altered or changed. Therefore, all such modifications are intended to be included within the scope of the present invention. The order or sequence of any process or method steps can be changed or reordered according to alternative embodiments. Any "device plus function" clause is intended to cover the structure of the execution function described herein, and is not only structurally equivalent but also equivalent structures. Without departing from the scope of the present invention, other replacements, modifications, changes and omissions can be made in the design, operating conditions and arrangement of the exemplary embodiments. Therefore, the invention is not limited to the specific embodiments, but extends to various modifications that still fall within the scope of the appended claims.

[0199] Additionally, in order to provide a concise description of exemplary embodiments, all features of an actual embodiment (ie, those features that are not relevant to the best mode presently contemplated for carrying out the invention or those that are not relevant to implementing the invention) may not be described.

[0200] It should be understood that in the development of any actual embodiment, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort may be complex and time-consuming, but for those of ordinary skill having the benefit of this disclosure, the development effort will be a routine task of design, fabrication, and production without undue experimentation.

[0201] 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 the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A textile heald frame adaptive control system, characterized in that: include: The sensor group includes a displacement sensor, a pressure sensor, and an image sensor, which are used to collect the displacement of the heald frame, the yarn tension, and the surface state data of the fabric in real time, and generate the heald frame state data; The input end of the defect detection module is connected to the multi-sensor group, receives the verified yarn tension and fabric surface state data, and performs defect detection on the fabric surface state data; Generate suppression signals for heald frames based on correlation analysis between fabric status rating signals and real-time displacement; Dynamic decision-making receives defect identification results, queries the defect type mapping table, determines the mandatory priority sequence of control parameters, judges weft yarn errors and tension changes on the fabric surface, and generates control instructions based on the judgment results.

2. The textile heald frame adaptive control system according to claim 1, characterized in that: The sensor group is also provided with multi-sensor redundancy check; Perform correlation check on the displacement sensor data and the warp and weft density of the fabric calculated by the image sensor. If the deviation exceeds the threshold, the displacement sensor calibration is triggered. Based on the spatial distribution data of the pressure sensing array, the local tension abnormality area is identified and the area is directed and focused on by the image sensor; The heald frame displacement, local tension and texture feature data are integrated to generate the heald frame status data.

3. The textile heald frame adaptive control system according to claim 2, characterized in that: The defect detection module includes a synchronous trigger unit, which is connected to the heald frame motion controller and is used to obtain the rising and falling edges of the heald frame motion cycle to trigger image capture, so that the image frame rate is synchronized with the heald frame motion frequency in integer multiples; The defect detection module is also connected to the image processing unit, which outputs the fabric defect enhanced image, and the defect position mapper generates a fabric status signal including the defect type, defect size and corresponding heald frame position coordinates.

4. The textile heald frame adaptive control system according to claim 3, characterized in that: The image processing unit receives an original fabric surface image at its input end, performs motion correction on the image based on the heald frame displacement feedback signal, eliminates image smear caused by the high-speed movement of the heald frame through reverse displacement compensation, enhances fabric texture contrast through adaptive histogram equalization, and crops a region of interest in the overlapping area of the fabric; The image processing unit also generates a defect feature vector at the output end by receiving the pre-processed image of the region of interest, and outputs a weaving defect type label based on the defect feature vector, including broken yarn, loose fiber, wrong fiber and defect.

5. The textile heald frame adaptive control system according to claim 3, characterized in that: The correlation analysis input terminal receives the defect type, defect size and corresponding heald frame position coordinates in the fabric status signal respectively, and binds the heald frame position coordinates at the time the defect occurs with the acceleration value obtained by the displacement sensor at the corresponding time through timestamp matching; Based on the displacement of the heald frame, the local stress distribution of the defect position during the acceleration mutation is obtained, and the defect acceleration correlation mapping table generated by pre-stored textile process parameters and historical data training includes: When the weft yarn misalignment defect is in the top area of the heald frame and the confidence level is not less than the maximum defect threshold, the heald frame-associated acceleration mutation alarm is triggered; Prioritization of acceleration suppression is triggered when yarn breaks and defects are in the middle area of the heald frame.

6. The textile heald frame adaptive control system according to claim 3, characterized in that: If the priority ranking of acceleration suppression is triggered, the dynamic decision input terminal receives the defect type label in the output result of the defect position mapper and the tension, heald frame displacement and texture data collected in real time; The risk factor is obtained by the difference between the real-time threshold of the preset priority weight and the critical threshold and the critical threshold, and a forced priority sequence of tension, heald frame displacement and texture is forcibly generated. A protection latch instruction is generated according to the forced priority sequence, and a heald frame parameter adjustment instruction is generated.

7. The textile heald frame adaptive control system according to claim 1, characterized in that: When the real-time acceleration value is detected to exceed the mutation threshold associated with the current defect type position, an amplitude-adjustable suppression signal is generated according to the priority weight coefficient; Perform phase pre-compensation on the suppression signal so that it is within the safe motion cycle before the acceleration mutation occurs and acts on the servo motor drive end; The output end superimposes the suppression signal with the linear output signal of the heald frame controller to control the movement of the heald frame.

8. The textile heald frame adaptive control system according to claim 7, characterized in that: The parameter adjustment instructions include heald frame movement adjustment, heald frame displacement adjustment, opening height adjustment and local tension adjustment; When a weft yarn misalignment defect occurs on the fabric surface, the speed reduction and height adjustment linkage strategy is triggered. Under the premise of keeping the opening height not lower than the minimum value of the process constraint, the speed of the heald frame movement is reduced until the defect is eliminated. When the tension change rate exceeds the critical threshold, the emergency elastic buffer mode is activated, the yarn break risk period is predicted through the real-time data of the heald frame displacement, and the tension threshold is dynamically lowered.

9. The textile heald frame adaptive control system according to claim 1, characterized in that: When the coupled disturbance compensator detects a sudden acceleration change, the output signal of the linear control model is asymmetrically superimposed with the suppression signal to generate a pulse command to drive the servo motor; Taking fabric condition rating as the main optimization objective, a fuzzy weight distribution algorithm is embedded in the PID control framework to adjust the weight coefficients of opening height, movement speed and tension threshold in real time.

10. A textile heald frame adaptive control method, based on a textile heald frame adaptive control system according to any one of claims 1 to 9, characterized in that: include: The sensor group acquires the position of the heald frame, yarn tension and fabric surface status data in real time, generates heald frame status data, and pre-processes the acquired data; Conduct defect detection on the fabric surface and analyze the type and size of weaving defects based on the defect detection results, including yarn breakage, loose fibers, misaligned fibers and other weaving defects; Receive the defect type, defect size and corresponding heald frame position coordinates in the fabric status signal, match the timestamp, perform correlation analysis, and output the defect-acceleration correlation mapping table; The defect-acceleration association mapping table is used to prioritize the acceleration mutation of the defect position, weft yarn misalignment and yarn breakage defects, and the heald frame movement is controlled based on the ranking results.

Images