High-elasticity antibacterial yarn production control method and system

Abstract

The invention discloses a high-elasticity antibacterial yarn production control method and system, and relates to the technical field of textile manufacturing. The method comprises the following steps: acquiring tension, temperature, antibacterial agent distribution and other multi-dimensional state information, and generating a comprehensive performance evaluation index based on a feature projection and gating residual fusion model; the index is compared with target performance, and key influence parameters such as fiber arrangement unevenness and antibacterial agent application distribution density are identified; calling a process response model to generate a parameter adjusting strategy, and driving an adjusting device to adjust process variables such as tension and temperature through a control instruction; and finally, based on a stochastic gradient descent algorithm and in combination with an expert knowledge base, adjusting rules are periodically updated, and performance self-adaptive control is achieved. According to the method, the intelligence and robustness of regulation and control of the elasticity and the antibacterial performance of the yarn are improved, and the method is suitable for industrial production of high-end textiles such as medical yarn and intelligent wearing fabric.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of textile manufacturing, and particularly relates to a high-elasticity antibacterial yarn production control method and system. BACKGROUND

[0002] With the rapid development of the textile industry towards functionalization and intelligentization, yarns with high elasticity and antibacterial properties are widely used in medical protection, high-performance sportswear, military materials and other fields, and higher requirements are put forward for the mechanical properties and functional properties of yarns. In order to realize large-scale production of high-performance yarns, more and more enterprises try to combine automatic production lines with performance monitoring systems to improve quality consistency and production efficiency.

[0003] However, in the existing yarn manufacturing process, the elasticity and antibacterial properties are coupled and affected by multiple process parameters such as raw materials, temperature, tension, and antibacterial agent application, and the control process has the following main technical problems: the yarn performance evaluation method is lagging behind: at present, off-line detection is generally used to carry out sampling inspection on finished yarns, it is difficult to obtain the elasticity and antibacterial properties of yarns in the production process in a timely manner, and real-time feedback cannot be provided for process adjustment. The control strategy is single and the feedback mechanism is missing: the existing system mainly depends on static control parameters, lacks the ability to adjust according to real-time production state, and cannot cope with the performance deviation caused by raw material fluctuation and environmental change. Lack of intelligent control mechanism for multiple performance targets: the performance of yarns is multi-dimensional, and the existing method is difficult to coordinate the mutual influence and control optimization between multiple performance indicators (such as elasticity and antibacterial property). The strategy updating mechanism relies on manual experience and lacks data-driven support: in terms of control strategy optimization, it mainly relies on manual experience accumulation at present, and has not formed a strategy updating mechanism based on the fusion of historical production data and knowledge rules, and lacks continuous adaptive ability.

[0004] Therefore, the existing technology urgently needs a high-performance yarn production control method and system which can obtain state information in real time, evaluate performance by fusing multi-dimensional features, identify key parameters and dynamically close-loop adjust process variables, and continuously optimize the control strategy by combining data-driven and knowledge-guided. SUMMARY

[0005] The present application aims to provide a high-elasticity antibacterial yarn production control method and system to solve the problems of lack of real-time performance evaluation, process adjustment lag, single control strategy and difficulty in adapting to batch differences and environmental fluctuations in the existing yarn production process, realize online perception of yarn elasticity and antibacterial properties, accurate identification of key influencing parameters, dynamic optimization and adaptive updating of process control strategy, and significantly improve the consistency of yarn products, intelligent control level and production stability.

[0006] The present application realizes the above-mentioned purpose through the following technical solutions:

[0007] In one aspect, the application provides a high-elasticity antibacterial yarn production control method, comprising the following steps:

[0008] Step 1: Collecting state information in the yarn production process, the state information including multi-dimensional parameter data representing the elasticity performance and antibacterial performance of the yarn;

[0009] Step 2: Performing feature fusion processing on the state information, and generating a comprehensive performance evaluation index using a multi-dimensional feature fusion model, the performance evaluation index reflecting the elasticity performance and antibacterial performance of the yarn at the same time;

[0010] Step 3: Comparing the performance evaluation index with a preset target performance value in terms of component level, and identifying key influence parameters causing performance deviation based on deviation determination results, the key influence parameters including fiber arrangement unevenness and antibacterial agent distribution density;

[0011] Step 4: Generating a control parameter adjustment strategy according to the identified key influence parameters, the control parameter adjustment strategy including process variables associated with the key influence parameters;

[0012] Step 5: Inputting the control parameter adjustment strategy into a control instruction generation module, generating and issuing an execution control instruction for driving a process adjustment action of a production process adjustment device, the production process adjustment action including set value adjustment of process variables associated with the key influence parameters;

[0013] Step 6: After completing the process adjustment, continuously collecting updated state information, and regenerating a performance evaluation index, dynamically modifying the control parameter adjustment strategy based on the performance feedback results, until the performance evaluation index meets the preset tolerance requirement;

[0014] Step 7: Taking performance data in multiple production batches as input samples, periodically updating adjustment rules for strategy generation using a stochastic gradient descent algorithm, and combining an expert knowledge base or external experience rules for auxiliary modification.

[0015] A further improvement of the application is that the multi-dimensional parameter data is obtained by a tension sensor, a displacement sensor, a temperature sensor, an antibacterial agent application rate monitoring device and an image acquisition device arranged on a yarn production line, and the multi-dimensional parameter data includes tension, deformation, temperature, antibacterial agent distribution density and yarn surface image information.

[0016] Further improvement of the present application is that the performance evaluation index is generated by a gating weighted fusion model that fuses yarn surface image features and production process state features, the gating weighted fusion model includes an image feature extraction module and a process state modeling module, which are respectively used for extracting a yarn surface image feature vector and a process state feature vector composed of tension, temperature and antibacterial agent distribution parameters, and includes the following steps:

[0017] A feature projection module is used for linearly mapping the process state feature vector to a feature space with the same dimension as the image feature vector to form a projected feature vector:

[0018] V' dyn = W surf ×V gate +B perf ;

[0019] A fusion gating module is used for generating a fusion adjustment coefficient Θ gate based on the spliced features [V' dyn , V gate ] and calculating a comprehensive performance evaluation vector through the following gating weighted formula:

[0020] V surf = Θ dyn ×V' surf +(1-Θ proj )×V proj ;

[0021] Wherein, V dyn is a dynamic process state feature vector in the yarn production process, containing real-time tension, temperature and antibacterial agent concentration changes; V gate is a yarn surface image feature vector, containing arrangement density, antibacterial particle distribution and surface texture image features; W perf and B perf are a projection weight matrix and a bias vector, used for projecting dynamic features to the same dimension as image features; V' target is a projected process state standard feature vector; Θ perf is a fusion gating coefficient vector, generated by the spliced image and dynamic features through an activation function (such as Sigmoid), ranging from 0 to 1; V perf is a weighted fused comprehensive performance feature vector, which serves as the basis for subsequent generation of yarn elasticity performance and antibacterial performance evaluation indexes.

[0022] Further improvement of the present application is that the step of identifying key influence parameters causing performance deviation based on the deviation determination result includes:

[0023] The comprehensive performance feature vector V target is compared with a target performance vector V impactA component level difference calculation is performed to obtain a performance deviation vector:

[0024] ΔV perf = V perf -V target ;

[0025] According to the absolute value size of each dimension in the deviation vector and the performance sensitivity of the input parameter, a key influence parameter causing performance deviation is identified, and the calculation formula is:

[0026]

[0027] Wherein: is the component of the performance deviation vector in the i-th dimension, indicating the difference between the actual and target performance in this item; S impact (i) is the sensitivity coefficient generated in the prior modeling stage, indicating the response strength of the i-th performance dimension to the input feature x i ; θ imp is the set minimum response threshold, and the input feature exceeding the value is determined as a key influence parameter; KeyParams is the finally identified key influence parameter set, used to guide the dynamic adjustment of the subsequent control strategy.

[0028] A further improvement of the application is that the step of generating a control parameter adjustment strategy by calling the process parameter influence response model comprises:

[0029] According to the identified key influence parameter set, the performance category associated with each parameter is determined, and the associated process variables are determined, the process variables including tension, temperature and antibacterial agent application rate;

[0030] The current deviation value of the key influence parameter is obtained, and a process parameter influence response model is called to calculate the target adjustment amount of each process variable, the process parameter influence response model being used to represent the corresponding relationship between the key influence parameter and the process variable;

[0031] According to the target adjustment amount, a control parameter adjustment strategy is constructed, the adjustment strategy defining the target setting value change direction and change amplitude of each process variable, used to adjust the stretch performance or antibacterial performance of the yarn.

[0032] A further improvement of the application is that the step of generating and issuing an execution control instruction for driving the process adjustment action of the production process adjustment device comprises:

[0033] The target adjustment amount of each process variable contained in the control parameter adjustment strategy is input to a control instruction generation module, and the control instruction generation module generates an execution control instruction corresponding to each process variable according to a preset control instruction mapping rule;

[0034] The execution control instructions are respectively sent to control units in the tension control system, the temperature control system and the antibacterial agent application system, for real-time adjustment of corresponding process execution devices.

[0035] The adjustment process includes adjusting the temperature control curve of the heating unit, the tension setting value of the drafting mechanism, and the spraying rate or application frequency of the antibacterial agent application device.

[0036] A further improvement of the present application is that the step of dynamically correcting the control parameter adjustment strategy based on the performance feedback result comprises:

[0037] After performing the production process adjustment action, real-time yarn state information after adjustment is obtained, and the performance evaluation index is recalculated;

[0038] The current performance evaluation index is compared with the preset target performance value to obtain a performance correction error vector;

[0039] It is judged whether the error vector continuously exceeds the preset tolerance threshold, and if so, the adjustment gain factor in the process parameter influence response model is dynamically adjusted according to the historical response data of control adjustment;

[0040] Based on the updated adjustment gain factor, a control parameter adjustment strategy is regenerated and used for next round of control instruction generation and process adjustment to realize convergence of performance error.

[0041] A further improvement of the present application is that the step of periodically updating the adjustment rule for strategy generation by using the machine learning algorithm comprises:

[0042] The mapping relationship between each control parameter adjustment strategy and the corresponding performance evaluation index is continuously recorded within multiple production cycles to construct a control strategy and performance response historical database;

[0043] The historical database is periodically analyzed to extract statistical correlation between parameter changes and performance improvement in the key adjustment path;

[0044] Based on the statistical correlation result, a random gradient descent method is used to update the adjustment rule in the control strategy generation module for deducing the process variable adjustment amount;

[0045] The updated adjustment rule includes a new adjustment gain factor, a parameter adjustment threshold and a feature selection priority, which is used to adapt to long-term changes in production conditions or performance requirements.

[0046] A further improvement of the present application is that the updating process of the adjustment rule comprises:

[0047] An expert knowledge base containing industry standard parameter ranges, experience adjustment suggestions and typical process bias response schemes is constructed for providing auxiliary adjustment suggestions; when updating the adjustment rules, the statistical analysis results from the production performance history data are compared with the experience rules in the expert knowledge base; if there is a significant deviation or rule conflict, the adjustment parameters are weighted and corrected according to the priority marking rules in the expert knowledge base; a composite adjustment rule that combines the data-driven results and the expert experience suggestions is formed to improve the stability and adaptability of the control strategy.

[0048] In another aspect, the present application provides a high-elasticity antibacterial yarn production control system applying a high-elasticity antibacterial yarn production control method as described above, the system comprising:

[0049] A state acquisition module for acquiring state information in the yarn production process, the state information including multi-dimensional parameter data representing the elasticity performance and the antibacterial performance of the yarn;

[0050] A fusion evaluation module for performing feature fusion processing on the state information, generating a performance evaluation index using a multi-dimensional feature fusion model, the performance evaluation index reflecting the elasticity performance and the antibacterial performance of the yarn at the same time;

[0051] A bias identification module for comparing the performance evaluation index with a preset target performance value in component level, identifying key influence parameters causing performance deviation based on the bias determination result, the key influence parameters including fiber arrangement unevenness and antibacterial agent distribution density;

[0052] A strategy generation module for generating a control parameter adjustment strategy according to the identified key influence parameters, calling a process parameter influence response model, the control parameter adjustment strategy including process variables associated with the key influence parameters;

[0053] An instruction generation module for converting the control parameter adjustment strategy into an execution control instruction and issuing it to a production process adjustment device, the execution control instruction being used to drive process adjustment actions to adjust the set values of the process variables associated with the key influence parameters;

[0054] A strategy correction module for continuously acquiring updated state information after completing the process adjustment, regenerating the performance evaluation index, and dynamically correcting the control parameter adjustment strategy based on the performance feedback result until the performance evaluation index meets the preset tolerance requirement;

[0055] A strategy update module for taking performance data in multiple production batches as input samples, periodically updating the adjustment rules used for control strategy generation using a machine learning algorithm, and combining expert knowledge base or external experience rules for auxiliary correction.

[0056] The beneficial effects of the present application are: the present application solves the key problems in the prior art that the yarn performance state cannot be obtained in real time, the process adjustment response is lagging, and the control strategy lacks adaptability, by constructing a full-process closed-loop control mechanism covering state perception, performance fusion evaluation, parameter identification, process regulation and strategy updating. The present application first collects multi-dimensional state information in the yarn production process, comprehensively covers parameters such as tension, temperature, antibacterial agent application characteristics, and other parameters highly related to elastic force and antibacterial performance, and performs feature extraction and performance evaluation on the above data based on a fusion model, which significantly improves the real-time identification accuracy of the comprehensive performance of the yarn. Through component level deviation analysis of the performance index and the preset target, the key factors affecting the performance fluctuation, such as fiber arrangement unevenness and antibacterial agent distribution density, can be quickly located, so as to realize accurate regulation. The process adjustment process automatically generates a regulation strategy based on a response model, and adjusts the process variables in real time through control instructions to drive the production device, effectively improving the response speed and pertinence of the control. At the same time, a self-learning mechanism based on stochastic gradient descent is introduced, and the adjustment rules are continuously optimized in combination with an expert knowledge base, so that the control strategy has the ability to dynamically evolve with batch changes, significantly enhancing the adaptability and robustness of the system to raw material differences and process fluctuations. Overall, the present application has the advantages of comprehensive performance perception, intelligent control path, and self-adaptive strategy updating, and can significantly improve the product consistency, production efficiency and quality control level of high-elasticity antibacterial yarn, and is suitable for high-performance yarn intelligent manufacturing scenes under complex process conditions. BRIEF DESCRIPTION OF DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0058] Among them:

[0059] Figure 1 The method flowchart of the present application is shown in the figure.

[0060] Figure 2 The system modularization diagram of the present application is shown in the figure. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the technical scheme of the embodiments of the present application will be described clearly and completely in combination with the drawings of the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all the embodiments. Based on the described embodiments of the present application, all other embodiments obtained by those skilled in the art belong to the scope of protection of the present application.

[0062] As Figure 1 shown, an embodiment of the present application provides a high-elasticity antibacterial yarn production control method, comprising the following steps:

[0063] Step 1: Collecting state information in the yarn production process, the state information including multi-dimensional parameter data representing the elasticity and antibacterial properties of the yarn;

[0064] The multi-dimensional parameter data is obtained by a tension sensor, a displacement sensor, a temperature sensor, an antibacterial agent application rate monitoring device, and an image acquisition device arranged on the yarn production line, and the multi-dimensional parameter data includes tension, deformation, temperature, antibacterial agent distribution density, and yarn surface image information.

[0065] In this embodiment, the state information collection link is arranged on the high-elasticity antibacterial yarn continuous production line, and the collection object includes multi-dimensional parameter data related to the elasticity and antibacterial properties of the yarn, mainly including tension, deformation, temperature, antibacterial agent application rate, antibacterial agent distribution density, and yarn surface image information.

[0066] Specifically:

[0067] The tension data is collected in real time by a micro tension sensor on the yarn guide rail, with the output unit being N (Newton), for reflecting the stress state of the yarn during the drafting process;

[0068] The deformation is obtained by a displacement sensor to obtain the cross-sectional or unit length change of the yarn, for calculating the draw ratio and elastic recovery ability;

[0069] The temperature information is collected by an infrared temperature sensor installed on the inner wall of the drafting channel, for monitoring the temperature control environment during the application of heat setting or antibacterial agent;

[0070] The antibacterial agent application rate is obtained by a dosing monitoring device of an antibacterial agent spraying module, and the application amount per unit time is obtained by integrating the time flow;

[0071] The antibacterial agent distribution density is obtained by an antibacterial agent component fluorescence response system and a yarn scanning acquisition device, and is converted into a density matrix after forming a distribution image;

[0072] The yarn surface image information is obtained by an industrial vision camera array, with the frame number per second being not less than 30 FPS, and the image is used for further extracting texture features, coating uniformity, and surface defect points.

[0073] The data of all sensors and image acquisition devices is converged through an edge processor, and is input to a subsequent fusion evaluation module after denoising, standardization, and time synchronization by a preprocessing module.

[0074] The embodiment can realize comprehensive perception of key state quantities of the production process, and guarantee accuracy and real-time performance of subsequent performance evaluation and control strategy generation.

[0075] In step 2, the state information is subjected to feature fusion processing, and a multi-dimensional feature fusion model is used to generate a comprehensive performance evaluation index, which reflects the elasticity performance and the antibacterial performance of the yarn.

[0076] In one embodiment, the performance evaluation index is generated by a gated weighted fusion model that fuses yarn surface image features and production process state features, and the gated weighted fusion model includes an image feature extraction module and a process state modeling module, which are respectively used to extract a yarn surface image feature vector and a process state feature vector composed of tension, temperature and antibacterial agent distribution parameters, and the process state feature vector includes the following steps:

[0077] A feature projection module is configured to linearly map the process state feature vector to a feature space with the same dimension as the image feature vector to form a projected feature vector:

[0078] V′ dyn =W proj ×V dyn +B proj ;

[0079] A fusion gating module is configured to generate a fusion adjustment coefficient Θ dyn based on the spliced features [V′ surf , V gate ] and calculate a comprehensive performance evaluation vector through the following gated weighted formula:

[0080] V perf =Θ gate ×V′ dyn +(1-Θ gate )×V surf ;

[0081] Wherein, V dyn is a dynamic process state feature vector in the yarn production process, which includes real-time tension, temperature and antibacterial agent concentration changes; V surf is a yarn surface image feature vector, which includes arrangement density, antibacterial particle distribution and surface texture image features; W proj and B proj are a projection weight matrix and a bias vector, which are used to project the dynamic features to the same dimension as the image features; V′ dyn is a projected process state standard feature vector; Θ gate is a fusion gating coefficient vector, which is generated by the spliced image and dynamic features through an activation function (such as Sigmoid), and the range is [0, 1]; V perfThe weighted fused comprehensive performance feature vector is taken as a basis for generating yarn elasticity performance and antibacterial performance evaluation indexes subsequently.

[0082] In this embodiment, to realize effective fusion of multi-source heterogeneous features in state information, the fusion module includes a nonlinear gate residual structure, which combines the complementarity of static image features and dynamic process state features, realizes flexible weighting by controlling the fusion strength coefficient, and ensures the dynamic adaptability of the fusion result to different source features.

[0083] The above fusion method avoids the feature conflict problem caused by direct splicing and improves the robustness and distinguishability of the performance index. In actual deployment, the gating parameter can be adaptively trained by a shallow neural network, and cross-entropy or least mean square error is used as the loss function to guide the network to learn the optimal fusion ratio under different production states.

[0084] Further, to ensure the reconfigurability and deployment efficiency of the fusion process, the fusion module can be implemented as a lightweight structure based on matrix multiplication and activation function, which is suitable for deployment on edge computing devices to meet real-time inference requirements in high-speed yarn production lines. The fused performance evaluation vector is sent to the subsequent deviation judgment module as the core reference index for measuring whether the current yarn elasticity and antibacterial performance meet the standards.

[0085] Through the gating weighted fusion mechanism in this embodiment, semantic mapping and collaborative expression of image information and process parameters in a unified space are realized, effectively improving the accuracy, stability and model perception ability to production condition changes of the performance evaluation index.

[0086] Step 3: Perform component-level comparison of the performance evaluation index and the preset target performance value, identify the key influence parameters causing performance deviation based on the deviation judgment result, and the key influence parameters include fiber arrangement unevenness and antibacterial agent distribution density.

[0087] In this embodiment, to identify the core influence factors causing the yarn comprehensive performance not to meet the standards, first, the comprehensive performance feature vector V perf fused in step 2 is compared with the preset target performance vector V target in component level difference calculation to obtain a performance deviation vector:

[0088] ΔV perf = V perf -V target ;

[0089] Wherein: represents the degree of deviation between the i-th performance component and the target value in the current production batch. To further evaluate the degree of influence of each input variable on the performance deviation, the system combines the sensitivity coefficient library constructed in advance during the experimental modeling stage. For each input parameter x i , the normalized sensitivity of the i-th performance is represented by the coefficient S impact (i) for quantifying the response strength of the input disturbance to the performance index fluctuation.

[0090] By combining the absolute value of the deviation component and the sensitivity of the input parameter, the system identifies the key influence parameter set through the following judgment formula:

[0091]

[0092] where: θ imp is the set minimum influence threshold. When the performance deviation amplification times the influence coefficient exceeds this threshold, the input variable x i plays a dominant role in the current working condition and needs to be targeted for subsequent process control. The system inputs the above KeyParams into the control strategy generation module as key influence parameters to dynamically construct the corresponding control variable adjustment scheme, ensuring that the control strategy has responsiveness and adaptability to the deviation.

[0093] Step 4: According to the identified key influence parameters, a process parameter influence response model is called to generate a control parameter adjustment strategy, which includes process variables associated with the key influence parameters;

[0094] In one embodiment, the step of calling the process parameter influence response model to generate the control parameter adjustment strategy includes:

[0095] According to the identified key influence parameter set, determine the performance category associated with each parameter, and determine the associated process variables, including tension, temperature, and antibacterial agent application rate;

[0096] Obtain the current deviation value of the key influence parameter, and call the process parameter influence response model to calculate the target adjustment amount of each process variable, which is used to represent the corresponding relationship between the key influence parameter and the process variable;

[0097] According to the target adjustment amount, construct a control parameter adjustment strategy, which defines the target set value change direction and change amplitude of each process variable, for adjusting the stretch performance or antibacterial performance of the yarn.

[0098] Step 5: input the control parameter adjustment strategy into a control instruction generation module to generate and issue an execution control instruction for driving a production process adjustment action of the production process adjustment device, the production process adjustment action including a set value adjustment on a process variable associated with the key influence parameter;

[0099] In one specific embodiment, the control instruction generation module generates an execution control instruction corresponding to each process variable according to a preset control instruction mapping rule;

[0100] The execution control instruction is respectively issued to a control unit in the tension control system, the temperature control system, and the antibacterial agent application system for real-time adjustment of the corresponding process execution device;

[0101] The adjustment process includes adjusting the temperature control curve of the heating unit, the tension set value of the drafting mechanism, and the spraying rate or application frequency of the antibacterial agent application device.

[0102] Step 6: after completing the process adjustment, continuously collect updated state information and regenerate performance evaluation indicators, and dynamically correct the control parameter adjustment strategy based on the performance feedback results until the performance evaluation indicators meet the preset tolerance requirement;

[0103] In one specific embodiment, the step of dynamically correcting the control parameter adjustment strategy based on the performance feedback results includes:

[0104] After performing the production process adjustment action, real-time acquisition of the adjusted yarn state information and recalculation of the performance evaluation indicators;

[0105] Compare the current performance evaluation indicators with the preset target performance value to obtain a performance correction error vector;

[0106] Determine whether the error vector continuously exceeds the preset tolerance threshold, and if so, dynamically adjust the adjustment gain factor in the process parameter influence response model according to the historical response data of the control adjustment;

[0107] Based on the updated adjustment gain factor, regenerate the control parameter adjustment strategy and use it for the next round of control instruction generation and process adjustment to realize the convergence of the performance error.

[0108] Step 7: use the performance data in multiple production batches as input samples to periodically update the adjustment rule for strategy generation using a stochastic gradient descent algorithm, and combine an expert knowledge base or external experience rules for auxiliary correction.

[0109] In one embodiment, the step of periodically updating the adjustment rule for strategy generation using a stochastic gradient descent algorithm includes:

[0110] Mapping relationship between each control parameter adjustment strategy and corresponding performance evaluation index is recorded continuously in multiple production cycles, and a control strategy and performance response history database is constructed;

[0111] The history database is analyzed periodically, and statistical correlation between parameter changes and performance improvement in a key adjustment path is extracted;

[0112] Based on the statistical correlation result, a random gradient descent method is used to update an adjustment rule for deriving a process variable adjustment amount in the control strategy generation module;

[0113] The updated adjustment rule includes a new adjustment gain factor, a parameter adjustment threshold value and a feature selection priority, which are used to adapt to long-term changes in production conditions or performance requirements.

[0114] In a specific embodiment, the updating process of the adjustment rule includes:

[0115] An expert knowledge base containing industry standard parameter ranges, experience adjustment suggestions and typical process deviation response schemes is constructed to provide auxiliary adjustment suggestions. When updating the adjustment rule, the statistical analysis result from the production performance history data is compared with the experience rules in the expert knowledge base. If there is a significant deviation or rule conflict, the adjustment parameters are weighted and corrected according to the priority marking rules in the expert knowledge base. A composite adjustment rule that integrates data-driven results and expert experience suggestions is formed to improve the stability and adaptability of the control strategy.

[0116] As Figure 2 shown, another embodiment of the present application provides a high-elasticity antibacterial yarn production control system, which applies a high-elasticity antibacterial yarn production control method as described above, and is characterized in that the system comprises:

[0117] A state acquisition module is configured to acquire state information in the yarn production process, and the state information includes multi-dimensional parameter data representing the elasticity performance and antibacterial performance of the yarn;

[0118] A fusion evaluation module is configured to perform feature fusion processing on the state information, and generate a performance evaluation index using a multi-dimensional feature fusion model, wherein the performance evaluation index reflects the elasticity performance and antibacterial performance of the yarn at the same time;

[0119] A deviation identification module is configured to compare the performance evaluation index with a preset target performance value in component level, and identify key influence parameters causing performance deviation based on the deviation determination result, wherein the key influence parameters include fiber arrangement unevenness and antibacterial agent distribution density;

[0120] A strategy generation module is configured to generate a control parameter adjustment strategy according to the identified key influence parameter, the control parameter adjustment strategy including a process variable associated with the key influence parameter;

[0121] An instruction generation module is configured to convert the control parameter adjustment strategy into an execution control instruction and send the execution control instruction to a production process adjustment device, the execution control instruction being used to drive a process adjustment action and adjust a set value of the process variable associated with the key influence parameter.

[0122] A strategy correction module is configured to continuously collect updated state information after the process adjustment is completed, regenerate the performance evaluation index, and dynamically correct the control parameter adjustment strategy based on the performance feedback result until the performance evaluation index meets a preset tolerance requirement.

[0123] A strategy update module is configured to use performance data in multiple production batches as input samples, periodically update an adjustment rule used for control strategy generation by using a machine learning algorithm, and assist in correction in combination with an expert knowledge base or external experience rules.

[0124] To sum up, the present application significantly improves the accuracy and response efficiency of the control of the yarn elasticity performance and the antibacterial performance by constructing a closed-loop regulation process integrating state information collection, heterogeneous feature fusion, performance deviation analysis, key parameter identification, and dynamic control strategy update. Compared with the conventional method relying on artificial experience or static rule control, the present application introduces a machine learning model and a process parameter response model to realize online adjustment of multiple process variables and adaptive optimization of the control strategy, and is particularly suitable for scenarios with high performance stability requirements such as high-performance medical yarns, functional fabrics, and intelligent fiber materials. The technology has the advantages of strong scalability, high regulation accuracy, and good industrial adaptability, and has a broad application prospect and significant industrial promotion value.

[0125] In the description of the present application, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, the skilled in the art can combine and combine the different embodiments or examples described in the present application and the features of the different embodiments or examples without contradiction.

[0126] Any process or method description in the flowchart or otherwise described herein can be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process. The scope of the preferred embodiments of the present application includes additional implementations in which the functions may be performed in a different order than shown or discussed, including in a substantially simultaneous manner or in a reverse order depending on the functions involved.

[0127] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included within the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A production control method for high-elasticity antibacterial yarn, characterized in that: The following steps are included: Step 1: Collecting status information during the yarn production process, wherein the status information includes multi-dimensional parameter data characterizing the elastic properties and antibacterial properties of the yarn; Step 2: Perform feature fusion processing on the state information and generate a comprehensive performance evaluation index using a multi-dimensional feature fusion model. The performance evaluation index simultaneously reflects the elastic properties and antibacterial properties of the yarn; Step 3: Compare the performance evaluation index with the preset target performance value at the component level, and identify the key influencing parameters that cause performance deviation based on the deviation determination result, wherein the key influencing parameters include fiber arrangement unevenness and antimicrobial agent distribution density; Step 4: Based on the identified key influencing parameters, call the process parameter impact response model to generate a control parameter adjustment strategy, wherein the control parameter adjustment strategy includes process variables associated with the key influencing parameters; Step 5: Input the control parameter adjustment strategy into the control instruction generation module to generate and issue execution control instructions for driving the process adjustment action of the production process adjustment device, wherein the production process adjustment action includes adjusting the set values ​​of process variables associated with key influencing parameters; Step 6: After the process adjustment is completed, the updated status information is continuously collected and the performance evaluation index is regenerated. The control parameter adjustment strategy is dynamically modified based on the performance feedback results until the performance evaluation index meets the preset tolerance requirements; Step 7: Using the performance data from multiple production batches as input samples, the stochastic gradient descent algorithm is used to periodically update the adjustment rules used for strategy generation, and auxiliary corrections are performed in combination with the expert knowledge base or external experience rules.

2. A high elasticity antibacterial yarn production control method according to claim 1, characterized in that: The multidimensional parameter data is collected by a tension sensor, displacement sensor, temperature sensor, antimicrobial agent application rate monitoring device and image acquisition device installed on the yarn production line. The multidimensional parameter data includes tension, deformation, temperature, antimicrobial agent distribution density and yarn surface image information.

3. The method for controlling the production of high-elasticity antibacterial yarn according to claim 1, characterized in that: The performance evaluation index is generated by a gated weighted fusion model that integrates yarn surface image features and production process state features. The gated weighted fusion model includes an image feature extraction module and a process state modeling module, which are respectively used to extract the yarn surface image feature vector and the process state feature vector composed of tension, temperature and antimicrobial agent distribution parameters, including the following steps: The feature projection module is used to linearly map the process state feature vector to a feature space with the same dimension as the image feature vector to form a projected feature vector: V′ dyn =W proj ×V dyn +B proj ; Fusion gating module for splicing features [V ′ dyn ,V surf ]Generate fusion adjustment coefficient Θ gate , and calculate the comprehensive performance evaluation vector using the following gating weighted formula: V perf =Θ gate ×V′ dyn +(1-Θ gate )×V surf ; Where: V dyn is the dynamic process state characteristic vector in the yarn production process; V surf is the yarn surface image feature vector; W proj 、B proj is the projection weight matrix and bias vector, which is used to project the dynamic features to the same dimension as the image features; V′ dyn is the standard feature vector of the process state after projection; Θ gate is the fusion gate coefficient vector, which is generated by the spliced ​​image and dynamic features through the activation function and has a range of [0,1]; V perf It is the comprehensive performance feature vector after weighted fusion, which serves as the basis for the subsequent generation of yarn elastic performance and antibacterial performance evaluation indicators.

4. The method for controlling the production of high-elasticity antibacterial yarn according to claim 1, wherein: The step of identifying the key influencing parameters causing the performance deviation based on the deviation determination result includes: The comprehensive performance characteristic vector V perf With the target performance vector V target Perform component-level difference calculations to obtain the performance deviation vector: ΔV perf =V perf -V target ; Based on the absolute value of each dimension in the deviation vector and the performance sensitivity of the input parameters, the key influencing parameters that cause performance deviation are identified. The calculation formula is: in: is the component of the performance deviation vector in the i-th dimension, indicating the difference between the actual and target performance in this item; S impact (i) is the sensitivity coefficient generated in the prior modeling stage, which represents the sensitivity of the i-th performance dimension to the input feature x i The response intensity of θ imp is the set minimum response threshold. Input features exceeding this value are determined to be key influencing parameters. KeyParams is the set of key influencing parameters finally identified.

5. The method for controlling the production of high-elasticity antibacterial yarn according to claim 1, characterized in that: The step of calling the process parameter impact response model to generate a control parameter adjustment strategy includes: Based on the identified set of key influencing parameters, determining the performance category associated with each parameter and determining the associated process variables, wherein the process variables include tension, temperature, and antimicrobial agent application rate; Obtaining the current deviation value of the key influencing parameter and calling the process parameter impact response model to calculate the target adjustment amount of each process variable. The process parameter impact response model is used to characterize the corresponding relationship between the key influencing parameter and the process variable; Based on the target adjustment amount, a control parameter adjustment strategy is constructed, and the adjustment strategy defines the change direction and change range of the target setting value of each process variable, which is used to adjust the elasticity or antibacterial performance of the yarn.

6. The method for controlling the production of high-elasticity antibacterial yarn according to claim 5, characterized in that: The step of generating and issuing an execution control instruction for driving a process adjustment action of a production process adjustment device includes: Inputting the target adjustment amount of each process variable included in the control parameter adjustment strategy into a control instruction generation module, the control instruction generation module generates an execution control instruction corresponding to each process variable according to a preset control instruction mapping rule; and issuing the execution control instruction to the control units in the tension control system, the temperature control system, and the antimicrobial agent application system, respectively, for real-time adjustment of the corresponding process execution devices; The adjustment process includes adjusting the temperature control curve of the heating unit, the tension setting value of the drafting mechanism, and the spraying rate or application frequency of the antimicrobial agent application device.

7. The method for controlling the production of high-elasticity antibacterial yarn according to claim 1, characterized in that: The step of dynamically correcting the control parameter adjustment strategy based on the performance feedback result includes: After executing the production process adjustment action, the adjusted yarn state information is obtained in real time, and the performance evaluation index is recalculated; the current performance evaluation index is compared with the preset target performance value to obtain the performance correction error vector; determining whether the error vector continuously exceeds a preset tolerance threshold, and if so, dynamically adjusting a regulation gain factor in a process parameter impact response model based on historical response data of control regulation; Based on the updated adjustment gain factor, the control parameter adjustment strategy is regenerated and used for the next round of control instruction generation and process adjustment to achieve convergence of performance error.

8. The method for controlling the production of high-elasticity antibacterial yarn according to claim 1, characterized in that: The step of periodically updating the adjustment rules for strategy generation using a machine learning algorithm includes: Continuously record the mapping relationship between each control parameter adjustment strategy and the corresponding performance evaluation indicators over multiple production cycles, and build a historical database of control strategies and performance responses; Performing periodic analysis on the historical database to extract statistical correlations between parameter changes and performance improvements in key regulation paths; and updating regulation rules for deriving process variable adjustments in a control strategy generation module using a stochastic gradient descent method based on the statistical correlation results. The updated regulation rules include new regulation gain factors, parameter adjustment thresholds, and feature selection priorities, and are used to adapt to long-term changes in production conditions or performance requirements.

9. The method for controlling the production of high-elasticity antibacterial yarn according to claim 1, characterized in that: The updating process of the adjustment rule includes: An expert knowledge base is constructed that includes industry standard parameter ranges, empirical adjustment suggestions, and typical process deviation response plans to provide auxiliary adjustment suggestions. When updating the adjustment rules, the statistical analysis results from the historical production performance data are compared with the empirical rules in the expert knowledge base. If there are significant deviations or rule conflicts, the adjustment parameters are weighted and corrected according to the priority marking rules in the expert knowledge base. A composite adjustment rule that integrates data-driven results and expert experience suggestions is formed to improve the stability and adaptability of the control strategy.

10. A high-elasticity antibacterial yarn production control system, using a high-elasticity antibacterial yarn production control method according to any one of claims 1 to 9, characterized in that: The system comprises: A status acquisition module is used to collect status information during the yarn production process, wherein the status information includes multi-dimensional parameter data representing the elastic properties and antibacterial properties of the yarn; a fusion evaluation module, configured to perform feature fusion processing on the state information and generate a performance evaluation index using a multi-dimensional feature fusion model, wherein the performance evaluation index simultaneously reflects the elastic properties and antibacterial properties of the yarn; a deviation identification module, configured to compare the performance evaluation index with a preset target performance value at a component level, and identify key influencing parameters causing performance deviation based on the deviation determination result, wherein the key influencing parameters include fiber arrangement unevenness and antimicrobial agent distribution density; a strategy generation module, configured to generate a control parameter adjustment strategy based on the identified key influencing parameters by calling a process parameter impact response model, wherein the control parameter adjustment strategy includes process variables associated with the key influencing parameters; An instruction generation module is used to convert the control parameter adjustment strategy into an execution control instruction and send it to the production process adjustment device. The execution control instruction is used to drive the process adjustment action and adjust the set value of the process variables associated with the key influencing parameters; The strategy correction module is used to continuously collect updated status information after completing process adjustment, regenerate performance evaluation indicators, and dynamically correct the control parameter adjustment strategy based on the performance feedback results until the performance evaluation indicators meet the preset tolerance requirements; the strategy update module is used to use the performance data within multiple production batches as input samples, use machine learning algorithms to periodically update the adjustment rules used for control strategy generation, and combine expert knowledge bases or external experience rules for auxiliary corrections.

Images