Method for dynamic regulation of a grid weaving process based on multivariate coupling

Abstract

The present application relates to the technical field of mesh fabric weaving intelligent regulation and control, in particular to a mesh fabric weaving process dynamic regulation and control method based on multivariate coupling, comprising the following steps: determining a multidimensional coupling regulation and control variable set of the mesh fabric weaving process, the coupling regulation and control variable set covering three categories of weaving process variables, yarn performance variables and environmental state variables; based on mesh fabric historical weaving full-process data, establishing a multivariate coupling correlation model, quantifying the coupling influence coefficients between variables, and constructing a coupling constraint equation containing cross-coupling terms; the present application first constructs a multidimensional coupling regulation and control variable set of three types of variables, establishes a multivariate coupling correlation model and a coupling constraint equation containing cross-coupling terms, eliminates the cross-coupling influence between variables from the root by a decoupling algorithm, and simultaneously solves the industry pain points of regulation and control lag and time sequence misplacement through a spindle synchronous acquisition and execution mechanism, thereby realizing full-process collaborative dynamic regulation and control of the weaving process.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology for mesh fabric weaving, specifically a dynamic control method for the mesh fabric weaving process based on multivariable coupling. Background Technology

[0002] As a core reinforcing substrate, mesh fabric is widely used in various fields such as building insulation, industrial filtration, electronic insulation, and aerospace composite materials. Its weaving precision, fabric surface flatness, and pore size consistency directly determine the performance and reliability of downstream products. Existing control technologies for the mesh fabric weaving process generally adopt a single-variable independent control mode: that is, fixed thresholds are set for individual process parameters such as weaving tension, warp feed speed, and spindle speed, and closed-loop adjustment is performed only when a deviation occurs in a certain parameter.

[0003] However, in actual industrial production, mesh fabric weaving is a dynamic process with strong coupling of multiple parameters, and existing technologies have four major technical defects: The inability to eliminate the cross-coupling effects between parameters easily leads to the problem of "adjusting one thing at the expense of another." There are strong coupling relationships between weaving process parameters, yarn performance parameters, and production environment parameters. For example, increasing the spindle speed will directly cause fluctuations in warp tension; changes in ambient temperature and humidity will alter the yarn breaking elongation, thus affecting the tension control effect; and yarn twist deviation will simultaneously affect the weft beating force and the interlacing stability of warp and weft yarns. Independent control of a single variable cannot solve the chain of deviations caused by coupling, resulting in quality defects such as uneven mesh fabric aperture, warp and weft yarn misalignment, and poor fabric flatness.

[0004] The regulation has a serious lag, and can only be passively corrected after quality deviations occur. It cannot achieve forward-looking and coordinated regulation based on multi-variable coupling and correlation, and it is difficult to meet the quality control requirements of high-precision mesh fabric.

[0005] Data acquisition and control execution are out of sync. The parameter acquisition and quality inspection of existing technologies are not synchronized. The adjustment actions of the execution mechanism are completed in cycles, which can easily cause continuous fluctuations in the weaving state, significantly increase the yarn breakage rate, and reduce production efficiency.

[0006] Poor quality consistency in production from multiple devices in the same batch makes it impossible to achieve production line-level collaborative optimization, which is insufficient to meet the consistency requirements of large-scale mass production. Summary of the Invention

[0007] The purpose of this invention is to provide a dynamic control method for the mesh fabric weaving process based on multivariable coupling, so as to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: A dynamic control method for the mesh fabric weaving process based on multivariable coupling includes the following steps: Step 1: Determine the multi-dimensional coupled control variable set for the mesh fabric weaving process. The coupled control variable set covers three major categories: weaving process variables, yarn performance variables, and environmental state variables. Step 2: Based on the historical weaving process data of the mesh fabric, establish a multivariate coupling correlation model, quantify the coupling influence coefficients between variables, and construct coupling constraint equations containing cross-coupling terms; Step 3: During the mesh fabric weaving process, real-time operating data of each variable in the coupled control variable set and online quality inspection data of the finished mesh fabric are collected synchronously in real time. Step 4: Input the collected real-time data into the multivariate coupled correlation model, calculate the deviation between the current weaving state and the preset target state, and eliminate the cross-coupling influence between variables through the decoupling control algorithm based on the coupling constraint equation to generate the coordinated control parameters of each variable. Step 5: The coordinated control parameters are synchronously sent to the corresponding actuators of the weaving equipment. The adjustment actions of all actuators are completed synchronously within the same rotation cycle of the weaving spindle to achieve coordinated dynamic adjustment of multiple variables. At the same time, based on the adjusted real-time operation data and online quality detection data, a closed-loop feedback is formed to iteratively optimize the multivariate coupled correlation model.

[0009] In a preferred embodiment of the present invention, the weaving process variables include the spinning spindle speed, warp feed amount, take-up tension, weft beat-up force, and heald frame opening amount; the yarn performance variables include the linear density, breaking strength, breaking elongation, and twist of the warp and weft yarns; and the environmental state variables include the ambient temperature, relative humidity, and dust concentration of the weaving workshop.

[0010] In a preferred embodiment of the present invention, the establishment of the multivariate coupling correlation model is specifically achieved by using response surface methodology combined with partial least squares regression to fit and analyze historical weaving data after removing abnormal operating conditions, quantifying the linear and nonlinear coupling influence coefficients between weaving process variables, yarn performance variables, and environmental state variables, constructing coupling constraint equations containing cross-coupling terms of the three types of variables, and setting corresponding safe operating threshold intervals for each variable.

[0011] In a preferred embodiment of the present invention, the decoupling control algorithm adopts a feedforward compensation decoupling algorithm combined with a fuzzy PID control algorithm. Based on the coupling constraint equation containing cross-coupling terms, it eliminates the cross-coupling influence between variables, generates independent and coordinated control increments for each variable, and ensures that the adjustment of a single variable does not cause the state deviation of other related variables.

[0012] In a preferred embodiment of the present invention, the acquisition of real-time operating data and online quality detection data adopts an equal time interval acquisition mode synchronized with the rotation speed of the weaving spindle. The acquisition frequency is positively correlated with the rotation frequency of the weaving spindle, and the acquisition frequency of operating data and quality data is completely consistent, ensuring that the acquired data are in the same weaving sequence.

[0013] In a preferred embodiment of the present invention, the execution of the coordinated dynamic adjustment adopts a synchronous triggering mechanism of the main shaft encoder, so that the adjustment actions of all actuators are completed within the same rotation cycle of the weaving main shaft, avoiding weaving state deviations such as warp tension fluctuations and warp and weft misalignment caused by asynchronous adjustments.

[0014] In a preferred embodiment of the present invention, when the real-time collected variable data exceeds the preset safe operation threshold range, or the online quality detection data exceeds the preset deviation threshold, the system immediately triggers a graded early warning, and at the same time generates abnormal source tracing results and emergency control parameters based on the multivariate coupling correlation model, and prioritizes the execution of yarn breakage shutdown protection or tension stabilization control.

[0015] In a preferred embodiment of the present invention, according to the core control objectives of the onboard stage, normal weaving stage, and finishing roll-changing stage of mesh fabric weaving, corresponding target state parameters and coupling constraint weights are preset respectively. The coupling constraint weights are assigned corresponding weight proportions to different types of variables based on the core control objectives, and directly act on the coupling influence coefficient of the coupling constraint equation. The corresponding model parameters and control strategies are automatically switched in different weaving stages to achieve adaptive control of the entire weaving process.

[0016] In a preferred embodiment of the present invention, the iterative optimization of the multivariate coupled correlation model specifically involves inputting the full-process operation data and final product quality inspection data of each completed weaving batch into the model, using incremental learning to iteratively optimize only the coupling influence coefficient deviation term corresponding to the batch data, updating the coupling influence coefficient and coupling constraint equation, and simultaneously eliminating abnormal working condition data to continuously improve the model's prediction accuracy.

[0017] In a preferred embodiment of the present invention, for multiple weaving machines of the same specification mesh fabric, a machine-specific coupling deviation correction coefficient is set for each machine to establish a machine-level multivariate coupling collaborative model; real-time operating data and control parameters of each machine are synchronously shared through the production line data bus, and the multivariate coupling correlation model is uniformly optimized based on the whole production line dataset to generate unified benchmark control parameters for the whole production line. Adaptive control parameters are generated by combining the correction coefficient of a single machine and distributed synchronously to achieve consistent quality control of mesh fabric in the same batch.

[0018] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.

[0019] This invention is the first to construct a multi-dimensional coupled control variable set of three types of variables, establish a multi-variable coupled correlation model and a coupling constraint equation containing cross-coupling terms, eliminate the cross-coupling effect between variables from the root through a decoupling algorithm, and solve the industry pain points of control lag and timing misalignment through the main axis synchronous acquisition and execution mechanism, and realize the full-process collaborative dynamic control of the weaving process; The weaving process variables, yarn performance variables, and environmental state variables of this invention specifically cover all core parameters affecting weaving quality, ensuring the comprehensiveness of the coupled model. The coupled model is constructed using response surface methodology combined with partial least squares regression, which can accurately quantify the linear and nonlinear coupling effects between variables. A control algorithm combining feedforward compensation decoupling and fuzzy PID effectively eliminates cross-coupling effects and achieves coordinated control. A data acquisition mode synchronized with the main shaft ensures the temporal consistency of operational and quality data. A synchronous execution mechanism with the main shaft during the same period avoids weaving fluctuations caused by asynchronous adjustments. A tiered early warning mechanism ensures production safety and stability. Phased adaptive control adapts to changes in coupling characteristics at different stages of the entire weaving process. An incremental learning iterative model continuously improves the model's prediction accuracy. Multi-device coordinated control ensures consistent product quality in mass production. Attached Figure Description

[0020] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 is a schematic diagram of the overall process of the dynamic control method for mesh fabric weaving based on multivariable coupling described in this invention; Figure 2 is a schematic diagram of the system architecture of the dynamic control method described in this invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the present invention clearer, a preferred embodiment of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Those skilled in the art, based on the disclosure of this invention, can repeatedly implement this method without creative effort, solve the corresponding technical problems, and achieve the expected technical effects.

[0023] This embodiment applies to the large-scale weaving production of alkali-resistant glass fiber mesh using rapier looms. The target product is a mesh fabric with a 5mm×5mm aperture and a weight of 80g / ㎡ for reinforcing exterior wall insulation. The weaving equipment consists of 6 high-speed rapier looms of the same model. The specific implementation process is as follows: Step 1: Determine the set of multidimensional coupling control variables.

[0024] In this embodiment, the set of coupled control variables is divided into three categories, covering all core variables affecting the weaving quality of the mesh fabric: The first category is weaving process variables, specifically including: weaving spindle speed, warp feed amount, take-up roller tension, weft beat-up force, and heald frame opening amount, which are the direct control objects of the weaving process; The second category is yarn performance variables, specifically including: linear density, breaking strength, breaking elongation, and twist of warp and weft yarns. In this embodiment, the linear density of the glass fiber untwisted roving is 33 tex, and the preset benchmark parameters are: breaking strength ≥ 0.25 N / tex, breaking elongation ≥ 3.0%, and twist of 80 twists / m; The third category is environmental state variables, specifically including: ambient temperature, relative humidity, and dust concentration in the weaving workshop. The preset benchmark production environment parameters are: temperature 25℃, relative humidity 65%, and dust concentration ≤ 0.5 mg / m³.

[0025] Step 2: Establish a multivariate coupled correlation model.

[0026] In this embodiment, historical data of 500 complete weaving batches of the same specification mesh fabric were collected over the past 12 months. Abnormal operating conditions such as yarn breakage, emergency shutdown, raw material abnormalities, and equipment failures were first removed, and 428 sets of valid operating condition data were retained. The response surface methodology combined with partial least squares regression was used to fit and analyze the valid historical data, quantify the linear and nonlinear coupling influence coefficients between weaving process variables, yarn performance variables, and environmental state variables, construct coupling constraint equations containing cross-coupling terms of the three types of variables, and set corresponding safe operation threshold ranges for each variable.

[0027] In this embodiment, the core coupling relationships obtained through model fitting include: for every 10 r / min increase in the weaving spindle speed, the coupling fluctuation of warp tension is 2.3 N; for every 5% decrease in relative humidity, the coupling decrease in yarn breaking elongation is 0.25%; for every 10 twists / m deviation in warp twist, the adjustment of weft insertion force is 8%; and for every 3°C increase in ambient temperature, the coupling adjustment of warp feed is 0.08%. All of these coupling relationships are quantified into the coupling constraint equations. Simultaneously, the safe operating range for spindle speed is set to 180-300 r / min, the safe operating range for warp tension is 15-30 N, and the safe operating range for heald frame opening is 80-120 mm to avoid exceeding the safe range and causing yarn breakage or equipment malfunction.

[0028] Step 3: Real-time data synchronization during the weaving process.

[0029] In this embodiment, sensors and detection devices are set at corresponding positions on each rapier loom to achieve real-time synchronous acquisition of all variables within the coupled variable set. At the same time, a machine vision online detection device is set up to collect online quality detection data of the mesh fabric.

[0030] The variable acquisition unit is synchronized with the loom spindle via a spindle encoder, employing an equal-time acquisition mode synchronized with the spindle rotation speed. The acquisition frequency is positively correlated with the spindle rotation frequency. In this embodiment, the spindle's rated speed is 250 r / min, corresponding to an acquisition frequency of 5 Hz. This means that one full variable synchronous acquisition is completed for every spindle rotation cycle. The machine vision online inspection device's acquisition frequency is completely synchronized with the variable acquisition frequency and undergoes nanosecond-level timing calibration with the variable acquisition unit via an industrial bus. This ensures that the acquired operational and quality data are within the same weaving timeline, completely avoiding model calculation deviations caused by timing misalignment. The acquired online quality data specifically includes mesh fabric aperture deviation, warp and weft yarn spacing deviation, fabric flatness, and yarn breakage count. The preset allowable range for aperture deviation is ±0.2 mm, and the preset allowable range for warp and weft yarn spacing deviation is ±0.15 mm.

[0031] Step 4: Deviation calculation and generation of coordinated control parameters.

[0032] The synchronously collected real-time data is input into a pre-established multivariate coupled correlation model to calculate the deviation between the current weaving state and the preset target state. A feedforward compensation decoupling algorithm combined with a fuzzy PID control algorithm is used. Based on the coupling constraint equation containing cross-coupling terms, the cross-coupling effects between variables are eliminated, and independent and coordinated control increments for each variable are generated to ensure that the adjustment of a single variable does not cause the state deviation of other related variables.

[0033] A typical control scenario in this embodiment is as follows: Real-time detection shows that the relative humidity of the environment drops to 58%. The model calculates that the yarn breaking elongation decreases by 0.35%, which will cause a warp tension fluctuation of 3.2N. At the same time, the mesh fabric aperture deviation expands to +0.25mm, exceeding the preset allowable range. At this time, the decoupling algorithm simultaneously generates two sets of collaborative control increments: one is to increase the warp beam feed by 0.12%, and the other is to decrease the take-up roller tension by 1.8N. The two sets of control increments are executed simultaneously, which not only offsets the tension fluctuation caused by the decrease in humidity, but also accurately corrects the aperture deviation. At the same time, it avoids the warp feed misalignment problem caused by adjusting the tension alone, thus eliminating the cross-coupling effect at the root.

[0034] Step 5: Synchronous control execution and closed-loop feedback optimization.

[0035] The generated coordinated control parameters are synchronously sent to the corresponding actuators of the rapier loom, such as the warp feed motor, take-up motor, beat-up mechanism, and heald frame drive mechanism, via the industrial bus. Through the synchronous trigger signal of the spindle encoder, it is ensured that the adjustment actions of all actuators are completed synchronously within the same rotation cycle of the weaving spindle, thus avoiding fluctuations in the weaving state caused by asynchronous adjustments.

[0036] Simultaneously, the adjusted real-time operational data and online quality inspection data are fed back to the model calculation module, forming a real-time closed-loop feedback to correct and adjust the control parameters in real time. After each complete weaving batch, the entire process operational data of that batch and the third-party laboratory testing data of the final product are input into the model. Using incremental learning, only the deviation term of the coupling influence coefficient corresponding to that batch's data is iteratively optimized, updating the coupling influence coefficient and coupling constraint equations. At the same time, abnormal operating condition data in that batch are removed, continuously improving the model's prediction accuracy. In this embodiment, after iterative optimization of three batches, the model's state prediction accuracy improved from the initial 92.3% to 98.7%.

[0037] Supplementary Implementation Method 1: Tiered Early Warning and Emergency Control.

[0038] In this embodiment, the tiered early warning system is divided into Level 1 and Level 2 warnings. The specific rules and execution logic are as follows: Level 1 warning is triggered when any variable data collected in real time exceeds the preset safe operating threshold range, or when the number of yarn breaks is ≥2 within 10 minutes. After triggering Level 1 warning, the system immediately reduces the weaving spindle speed to 80% of the rated speed and adjusts the warp tension to the middle value of the safe range, prioritizing weaving stability and avoiding continuous yarn breaks. Simultaneously, it generates anomaly tracing results based on a coupled correlation model, indicating abnormal variables and their coupling influence paths for on-site personnel to investigate. Level 2 warning is triggered when online quality inspection data exceeds twice the preset deviation threshold, or when the quality deviation continues to increase for three consecutive collection cycles. After triggering Level 2 warning, the system immediately stops the loom and outputs an anomaly investigation list, indicating the corresponding abnormal variables, coupling influence factors, and standardized investigation directions to prevent the generation of batches of non-conforming products.

[0039] Supplementary Implementation Method 2: Full-process phased adaptive control.

[0040] In this embodiment, based on the core control objectives of the weaving process in the onboarding stage, normal weaving stage, and finishing and changing-wrap stage, corresponding target state parameters and coupling constraint weights are preset. The allocation logic of the coupling constraint weights is as follows: based on the core control objectives of different weaving stages, corresponding weight ratios are set for different types of variables, and the sum of all weights is 1. The coupling constraint weights directly affect the coupling influence coefficient of the coupling constraint equation. The higher the weight, the higher the priority of the coupling constraint for that type of variable. Specifically, the weight allocation is as follows: the core objective of the onboarding stage is stable warp tension and smooth weft insertion, so a weight of 0.6 is set for tension-related process variables, a weight of 0.25 for yarn performance variables, and a weight of 0.15 for environmental state variables; the core objective of the normal weaving stage is aperture accuracy and fabric flatness, so a weight of 0.7 is set for weaving process variables, a weight of 0.2 for yarn performance variables, and a weight of 0.1 for environmental state variables; the core objective of the finishing and changing-wrap stage is take-up flatness and changing-wrap stability, so a weight of 0.55 is set for take-up-related process variables, a weight of 0.25 for yarn performance variables, and a weight of 0.2 for environmental state variables. The system automatically switches the corresponding model parameters and control strategies at different weaving stages to achieve adaptive control of the entire weaving process.

[0041] Supplementary Implementation Method 3: Multi-device Coordinated Consistency Control.

[0042] In this embodiment, a multi-variable coupled collaborative model is established for six identical rapier looms producing mesh fabric of the same specification. First, a unique coupling deviation correction coefficient is set for each loom to eliminate parameter deviations caused by differences in mechanical performance and wear levels between different looms. A production line-level data sharing bus is built via industrial Ethernet. Real-time operating data, control parameters, and quality inspection data from all looms are synchronously uploaded to the production line's central control unit at a frequency of 5Hz. Based on the full dataset from all looms, the central control unit uniformly optimizes the coupling influence coefficients and coupling constraint equations of the multi-variable coupled correlation model, generating unified benchmark control parameters for the entire production line. These parameters are then combined with the unique correction coefficients for each loom to generate adaptive control parameters for each loom, which are then synchronously distributed to all looms for execution. Through this multi-loom collaborative control method, the range of mesh fabric aperture deviation produced by the six looms in the same batch is reduced from 0.2mm to 0.08mm, and the batch product quality consistency pass rate is increased to 99.8%, fully meeting the consistency requirements for large-scale mass production.

[0043] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A dynamic control method for the mesh fabric weaving process based on multivariable coupling, characterized in that, Includes the following steps: Step 1: Determine the multi-dimensional coupled control variable set for the mesh fabric weaving process. The coupled control variable set covers three major categories: weaving process variables, yarn performance variables, and environmental state variables. Step 2: Based on the historical weaving process data of the mesh fabric, establish a multivariate coupling correlation model, quantify the coupling influence coefficients between variables, and construct coupling constraint equations containing cross-coupling terms; Step 3: During the mesh fabric weaving process, real-time operating data of each variable in the coupled control variable set and online quality inspection data of the finished mesh fabric are collected synchronously in real time. Step 4: Input the collected real-time data into the multivariate coupled correlation model, calculate the deviation between the current weaving state and the preset target state, and eliminate the cross-coupling influence between variables through the decoupling control algorithm based on the coupling constraint equation to generate the coordinated control parameters of each variable. Step 5: The coordinated control parameters are synchronously sent to the corresponding actuators of the weaving equipment. The adjustment actions of all actuators are completed synchronously within the same rotation cycle of the weaving spindle to achieve coordinated dynamic adjustment of multiple variables. At the same time, based on the adjusted real-time operation data and online quality detection data, a closed-loop feedback is formed to iteratively optimize the multivariate coupled correlation model.

2. The method for dynamic control of the mesh fabric weaving process based on multivariable coupling according to claim 1, characterized in that, The weaving process variables include the spindle speed, warp feed, take-up tension, weft beat-up force, and heald opening. The yarn performance variables include the linear density, breaking strength, breaking elongation, and twist of the warp and weft yarns. The environmental state variables include the ambient temperature, relative humidity, and dust concentration in the weaving workshop.

3. The dynamic control method for mesh fabric weaving process based on multivariable coupling according to claim 1, characterized in that, The establishment of the multivariate coupling correlation model specifically involves using response surface methodology combined with partial least squares regression to fit and analyze historical weaving data after removing abnormal operating conditions. This quantifies the linear and nonlinear coupling influence coefficients between weaving process variables, yarn performance variables, and environmental state variables, constructs coupling constraint equations containing cross-coupling terms of the three types of variables, and sets corresponding safe operating threshold ranges for each variable.

4. The dynamic control method for mesh fabric weaving process based on multivariable coupling according to claim 1, characterized in that, The decoupling control algorithm adopts a feedforward compensation decoupling algorithm combined with a fuzzy PID control algorithm. Based on the coupling constraint equation containing cross-coupling terms, it eliminates the cross-coupling effects between variables, generates independent and coordinated control increments for each variable, and ensures that the adjustment of a single variable does not cause the state deviation of other related variables.

5. The dynamic control method for mesh fabric weaving process based on multivariable coupling according to claim 1, characterized in that, The acquisition of real-time operation data and online quality inspection data adopts an equal time interval acquisition mode synchronized with the rotation speed of the weaving spindle. The acquisition frequency is positively correlated with the rotation frequency of the weaving spindle, and the acquisition frequency of operation data and quality data is completely consistent, ensuring that the acquired data are in the same weaving sequence.

6. The method for dynamic control of mesh fabric weaving process based on multivariable coupling according to claim 1, characterized in that, The execution of the coordinated dynamic adjustment adopts a synchronous triggering mechanism of the main shaft encoder. All adjustment actions of the actuators are completed within the same rotation cycle of the weaving main shaft, avoiding weaving state deviations such as warp tension fluctuations and warp and weft misalignment caused by asynchronous adjustments.

7. The dynamic control method for mesh fabric weaving process based on multivariable coupling according to claim 3, characterized in that, When the real-time collected variable data exceeds the preset safe operation threshold range, or the online quality detection data exceeds the preset deviation threshold, the system immediately triggers a graded warning. At the same time, based on the multivariate coupling correlation model, it generates abnormal source tracing results and emergency control parameters, and prioritizes the execution of yarn breakage shutdown protection or tension stabilization control.

8. The dynamic control method for mesh fabric weaving process based on multivariable coupling according to claim 1, characterized in that, Based on the core control objectives of the onboarding stage, normal weaving stage, and finishing and roll-changing stage of mesh fabric weaving, corresponding target state parameters and coupling constraint weights are preset respectively. The coupling constraint weights are assigned corresponding weight proportions to different types of variables based on the core control objectives, and directly affect the coupling influence coefficient of the coupling constraint equation. The corresponding model parameters and control strategies are automatically switched at different weaving stages to achieve adaptive control of the entire weaving process.

9. The method for dynamic control of mesh fabric weaving process based on multivariable coupling according to claim 1, characterized in that, The iterative optimization of the multivariate coupled correlation model involves inputting the full-process operation data and final product quality inspection data of each completed weaving batch into the model using an incremental learning approach. The model iteratively optimizes only the deviation term of the coupling influence coefficient corresponding to the batch data, updates the coupling influence coefficient and coupling constraint equation, and removes abnormal working condition data to continuously improve the model's prediction accuracy.

10. The method for dynamic control of the mesh fabric weaving process based on multivariable coupling according to claim 1, characterized in that, For multiple weaving machines of the same specification mesh fabric, a unique coupling deviation correction coefficient is set for each machine, and a multivariate coupling collaborative model at the machine level is established. Real-time operating data and control parameters of each machine are synchronously shared through the production line data bus. Based on the whole production line dataset, the multivariate coupling correlation model is uniformly optimized to generate unified benchmark control parameters for the whole production line. Adaptive control parameters are generated by combining the correction coefficient of each machine and distributed synchronously to achieve consistent quality control of mesh fabric in the same batch.

Images