Composite brocade garment fabric textile process based on multi-source data fusion regulation and control

Abstract

The invention discloses a composite brocade garment fabric spinning process based on multi-source data fusion regulation and control, and belongs to the technical field of fabric spinning, the spinning process comprises the following specific steps: I, collecting a multi-modal culture pattern, generating an original pattern, and rehearsing the yarn interweaving effect of the generated pattern; iI, acquiring physiological data and different environmental data of a wearer in real time, and adjusting a brocade fabric organization structure regulation scheme according to the acquired data; iII, performing linkage design on the pattern and the organization structure according to the pattern generation result and the structure regulation and control scheme, and constructing a dynamic organization database to store design data; according to the method, self-adaptive regulation and control of the fabric on the somatosensory comfort of a wearer can be achieved, the wearing experience is remarkably improved, the design reusability and precision are enhanced, the design accuracy is improved, the trial weaving period is shortened, and the material and labor cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of fabric weaving, and in particular to a composite brocade clothing fabric weaving process based on multi-source data fusion and regulation. Background Art

[0002] Brocade, a culmination of traditional textile art, fuses sophisticated weaving techniques with rich cultural patterns, carrying a rich legacy of historical value and craftsmanship. However, with the increasing demand for fabric functionality, sustainability, and personalized aesthetics in the modern apparel industry, traditional brocade craftsmanship faces multiple technical and process bottlenecks in adapting to the modern market. On the one hand, traditional brocade weaving relies heavily on the experience and judgment of skilled workers, resulting in low production efficiency and delayed adjustments, making it difficult to adapt to personalized and small-batch customization needs. On the other hand, its pattern generation and process transfer lack systematic digital support, hindering the innovative expression and rapid iteration of cultural elements. Furthermore, traditional fabrics often lack functional responsiveness (such as temperature control and color change), limiting their expansion into emerging fields such as functional apparel and smart wearables. Therefore, it is crucial to develop a weaving process for composite brocade apparel fabrics based on multi-source data fusion and control.

[0003] The existing weaving process for composite brocade clothing fabrics cannot achieve adaptive regulation of the fabric's physical comfort to the wearer, which reduces the wearing experience, has low design reusability and precision, reduces design accuracy, lengthens the trial weaving cycle, and increases material and labor costs. To this end, we propose a weaving process for composite brocade clothing fabrics based on multi-source data fusion regulation. Summary of the Invention

[0004] The purpose of the present invention is to solve the defects in the prior art and to propose a composite brocade clothing fabric weaving process based on multi-source data fusion and regulation.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions: A composite brocade clothing fabric weaving process based on multi-source data fusion control, the specific steps of the weaving process are as follows: I. Collect multimodal cultural patterns and generate original patterns, while previewing the yarn interweaving effects of the generated patterns; II. Real-time acquisition of the wearer's physiological data and different environmental data, and adjustment of the brocade fabric structure control plan based on the collected data; III. Based on the pattern generation results and the structural control plan, the pattern and tissue structure are designed in a linked manner, and a dynamic tissue database is constructed to store the design data; IV. Based on the design data, the jacquard loom is driven to automatically weave according to the optimized organization and pattern information, while the weaving data is collected and verified in real time; Ⅴ. After weaving is completed, all weaving data will be uploaded to the chain for storage, and a non-homogeneous certificate corresponding to the brocade clothing fabric used will be generated.

[0006] As a further solution of the present invention, the specific steps of generating an original pattern in step I are as follows: S1.1: Collect data from various data sources, including images of historical artifacts, textual records of traditional brocade weaving techniques, and graphic data on fashion trends. Categorize the collected data by data type and then map each type of data into a unified public semantic space to generate different sets of modal vectors. S1.2: The obtained modal vectors are input into the style encoding network. The style encoding network performs multiple rounds of convolution processing on each modal vector through the built-in convolution block to extract local style features. The attention module then performs weighted summation processing on the local style features extracted by the convolution block to generate the corresponding global style features. S1.3: Analyze the potential structural rules of the patterns in each modal vector through a graph neural network and generate the structural semantics of the corresponding pattern. Use a bidirectional attention mechanism to fuse the collected global style features and the structural semantics of the pattern, and output the corresponding composite semantic features. S1.4: Each set of composite semantic features is input into the CM-GAN network. The CM-GAN network conducts a generator-discriminator game based on the received composite semantic features. The loss values ​​of the generator and discriminator are recorded in real time after each round of the game. Based on the obtained loss values, the parameters of the generator and discriminator are adjusted respectively. The game is repeated alternately until the generator loss value converges to a preset range. S1.5: Input real-time composite semantic features into the CM-GAN network, and use the trained generator to generate original pattern images with cultural semantics. Then, use the brocade physical simulator to perform yarn interweaving calculations on the original pattern images to verify the clarity, symmetry, and integrity of the pattern under the specific weaving structure.

[0007] As a further solution of the present invention, the specific steps of regulating the weave structure of the brocade fabric in step II are as follows: S2.1: A flexible sensor array integrated into the fabric collects the user's physiological and environmental data in real time, synchronizing the timestamps of each sensor transmission channel. Each set of collected physiological and environmental data is then normalized and weighted based on the user's current state. S2.2: Divide the fabric into multiple structural units. Initialize the topological parameters of each structural unit based on the standard brocade parameter library. Then, establish a corresponding adaptive function based on the contribution of each structural unit to the tester's comfort index and its impact on the mechanical performance constraints of the fabric. S2.3: Based on the initial topological parameters of each organizational structure unit, the fitness of each organizational structure unit is calculated using an adaptive function. The adaptive function is then solved using the gradient descent method, and the topological parameters of the corresponding organizational structure unit are updated based on the solution. S2.4: Iteratively update the topological parameters of each tissue structure unit until the fitness of each tissue structure unit converges to a preset range, and then convert the final topological parameters of each tissue structure unit into a corresponding brocade fabric tissue structure control scheme.

[0008] As a further solution of the present invention, the specific steps of linking the pattern and the tissue structure according to the pattern generation result and the structure control scheme in step III are as follows: S3.1: Collect the generated original pattern images and the brocade fabric structure control scheme, use image segmentation and texture clustering technology to divide the original pattern images into multiple groups of sub-regions with consistent styles, and assign semantic attributes of the corresponding pattern image to each sub-region manually or through a sample library; S3.2: Calculate the complexity score of the required weave structure for each sub-region based on the intensity of texture detail and the rate of color level change. Then, based on the complexity score of each sub-region, retrieve the appropriate basic weave type. Then, establish the corresponding weave parameter set based on the semantic attributes and weaving complexity of each sub-region. S3.3: Adjust the brocade fabric structure control scheme based on the established weave parameter set, including the warp and weft yarn density, interweaving pattern, and float length limit parameter values. Then, input the adjusted brocade fabric structure control scheme into the weaving process simulator, and calculate the error value between the ideal pattern image and the simulated pattern image through the difference feedback function; S3.4: Based on the error value, a joint optimization strategy is adopted to adjust the pattern and weave parameters simultaneously. Through multiple rounds of iterative optimization, until the error value converges to the preset range, the optimized weave parameters of each sub-region are integrated to generate the weaving structure matrix of the overall fabric. Based on the established weaving structure matrix of the overall fabric, corresponding operation instructions are issued to the CNC loom.

[0009] As a further solution of the present invention, the specific steps of constructing a dynamic organization database to store design data in step III are as follows: S4.1: Collect the weave structure information of various brocade fabrics through image scanning, microscopic section analysis and numerical control machine files, and mark the use scenarios and performance, and then convert the collected fabric structure sample information into corresponding weave feature vectors through a feature coding model; S4.2: Collect the performance indicators of each fabric structure sample under different use scenarios, establish the weave structure features and performance, and then calculate the similarity between each fabric structure sample according to the coded weave feature vectors, and construct a dynamic weave database based on a graph or tree structure; S4.3: The dynamic weave database receives the newly generated brocade fabric weave structure and performance data in real time, and performs incremental updating and error correction on the existing database, and then standardizes all weave sample data into a unified data structure format and stores it.

[0010] As a further scheme of the present application, the real-time collection and verification of weaving data in step IV S5.1: Through various sensing nodes arranged on the loom and the production environment, the parameters in the brocade fabric production process are periodically collected, and the parameters are preprocessed, and the unique identity code is marked for each parameter, and the marked parameters are integrated into a parameter set, and then the parameter set is coded into a data format recognized by the blockchain; S5.2: Generate a hash fingerprint through the SHA256 encryption algorithm, and package the hash fingerprint, timestamp and device node number into a new block, and then upload the corresponding parameter set to the new block, and when the parameter set is uploaded to the new block, trigger the smart contract containing the legal range of the preset process parameters to automatically verify the parameter set; S5.3: If any parameter in the parameter set exceeds the set tolerance interval, the abnormal state will be recorded and a notification will be sent, and an audit request will be automatically sent to the quality control personnel, and the smart contract will also count the verification results of all uploaded parameters in the current time period to generate the compliance rate index of the current production process, and the compliance rate index will be attached to the corresponding block in the chain as a state label.

[0011] Compared with the prior art, the present application has the following advantages: The composite brocade garment fabric textile process based on multi-source data fusion regulation integrates flexible sensors to collect wearer physiological and environmental data, initializes the organizational structure unit in combination with the standard brocade parameter library, establishes an adaptive function according to comfort and mechanical constraints, optimizes the topological parameters through the gradient descent method, forms a dynamic organizational structure regulation scheme, then analyzes the original pattern using image segmentation and texture clustering technology, adjusts the organizational parameters in combination with semantic attributes and complexity scores, iteratively optimizes the pattern and structure through a weaving simulator, generates the overall weaving matrix and sends it to a numerical control loom, establishes an organization-performance database through image analysis and microstructure sampling, dynamically updates the feature vector and integrates it into the standardized structure, and then writes the encrypted parameters into the blockchain during the production process, verifies the traceability mechanism through the smart contract, realizes the full-process intelligent textile process from data collection, structure optimization to chain verification of the brocade fabric, can realize adaptive regulation of the fabric to the wearer's sensory comfort, significantly improve the wearing experience, enhance the design reusability and precision, improve the design accuracy, shorten the trial weaving period, and reduce the material and labor costs. BRIEF DESCRIPTION OF DRAWINGS

[0012] The accompanying drawings are included to provide a further understanding of the application, and constitute a part of the specification, illustrate the application together with the embodiments thereof, and are used to explain the application, and do not constitute a limitation on the application.

[0013] Figure 1 A flowchart of a composite brocade garment fabric textile process based on multi-source data fusion regulation is proposed for the application. DETAILED DESCRIPTION Example 1

[0014] Reference Figure 1 A composite brocade garment fabric textile process based on multi-source data fusion regulation, the specific steps of the textile process are as follows: Collect multi-modal cultural patterns and generate original pattern images, and pre-play the yarn interweaving effect of the generated pattern.

[0015] Specifically, data from different data sources are collected, including images of historical relics, written records of traditional brocade weaving techniques, and graphic data of popular trends. The collected data are classified according to the data type, and then the various types of data are mapped into a unified public semantic space to generate different groups of modal vectors. The obtained modal vectors are input into the style encoding network. The style encoding network performs multiple rounds of convolution processing on each modal vector through the built-in convolution block to extract local style features. The local style features extracted by the convolution block are then weighted and summed through the attention module to generate the corresponding global style features. The potential structural rules of the patterns in each modal vector are analyzed through the graph neural network, and the structural semantics of the corresponding pattern are generated. The collected The global style features and the structural semantics of the pattern are integrated, and the corresponding composite semantic features are output. Each group of composite semantic features is input into the CM-GAN network. The CM-GAN network performs a generator-discriminator game based on the received composite semantic features, and records the loss values ​​of the generator and the discriminator after each round of game in real time. The parameters of the generator and the discriminator are adjusted based on the obtained loss values, and the game is alternated until the generator loss value converges to a preset range. The real-time composite semantic features are input into the CM-GAN network, and the trained generator is used to generate original pattern images with cultural semantics. The original pattern images are then subjected to yarn interweaving calculations through a brocade physical simulator to verify the clarity, symmetry and integrity of the pattern under the specific weaving organization.

[0016] Acquire the wearer's physiological data and different environmental data in real time, and adjust the brocade fabric structure control plan based on the collected data.

[0017] Specifically, through the flexible sensor array integrated in the fabric, various physiological data and surrounding environmental data of the test user are collected in real time, and the timestamps of the transmission channels of each sensor are synchronized. Then, each set of collected physiological data and environmental data are normalized, and corresponding weight values ​​are assigned to each set of data according to the current state of the test user. The fabric is divided into multiple tissue structure units, and the topological parameters of each tissue structure unit are initialized according to the standard brocade parameter library. Then, according to the contribution of each tissue structure unit to the comfort index of the test personnel and the influence on the mechanical performance constraints of the fabric, a corresponding adaptive function is established. Based on the initial topological parameters of each tissue structure unit, the fitness of each tissue structure unit is calculated by the adaptive function. Then, the adaptive function is solved by the gradient descent method, and the topological parameters of the corresponding tissue structure unit are updated according to the solution results. The topological parameters of each tissue structure unit are iteratively updated until the fitness of each tissue structure unit converges to the preset range. Then, the final topological parameters of each tissue structure unit are converted into the corresponding brocade fabric structure control scheme. Example 2

[0018] Reference Figure 1 A composite brocade clothing fabric weaving process based on multi-source data fusion and control, the specific steps of the weaving process are as follows: According to the pattern generation results and the structural control plan, the pattern and tissue structure are designed in a linked manner, and a dynamic tissue database is built to store the design data.

[0019] Specifically, the generated original pattern images and brocade fabric structure control scheme are collected, and the image segmentation and texture clustering technology is used to divide the original pattern images into multiple groups of sub-regions with consistent styles. The semantic attributes of the corresponding pattern image are assigned to each sub-region through manual or sample library. According to the texture detail intensity and color level change rate of each sub-region, the complexity score of the required structure of each sub-region is calculated. Based on the complexity score of each sub-region, the adapted basic structure type is retrieved. Then, according to the semantic attributes and weaving complexity of each sub-region, the corresponding structure parameter set is established. The brocade fabric is adjusted according to the established structure parameter set. The weaving structure control scheme is adjusted, including the parameter values ​​of warp and weft yarn density, interweaving method and floating length limit. Then the adjusted brocade fabric weaving structure control scheme is input into the weaving process simulator, and the error value between the ideal pattern image and the simulated pattern image is calculated through the difference feedback function. Based on the error value, the pattern and weaving parameters are adjusted at the same time by adopting a joint optimization strategy. Through multiple rounds of iterative optimization, until the error value converges to the preset range, the optimized weaving parameters of each sub-region are fused to generate the weaving structure matrix of the overall fabric. Based on the established weaving structure matrix of the overall fabric, the corresponding operation instructions are issued to the CNC loom.

[0020] Specifically, through image scanning, microscopic cross-section analysis and CNC machine files, the organizational structure information of various brocade fabrics is collected, and the usage scenarios and performance are marked. Then, the collected fabric structure sample information is converted into corresponding organizational feature vectors through a feature coding model, and the performance indicators of each fabric structure sample in different usage scenarios are collected to establish organizational structure characteristics and performance. Then, the similarity between each fabric structure sample is calculated based on the encoded organizational feature vector, and a dynamic organizational database based on a graph or tree structure is constructed. The dynamic organizational database receives the newly generated brocade fabric organizational structure and performance data in real time, and performs incremental updates and error corrections on the existing database. Then, all organizational sample data are standardized into a unified data structure format and stored.

[0021] According to the design data, the jacquard loom is driven to automatically weave according to the optimized organization and pattern information, and the weaving data is collected and verified in real time. Specifically, various sensor nodes are deployed on the loom and production environment to periodically collect various parameters in the production process of brocade fabrics, pre-process each parameter, mark each parameter with a unique identity code, and integrate the marked parameters into a parameter set. The parameter set is then encoded into a data format recognized by the blockchain, and a hash fingerprint is generated through the SHA256 encryption algorithm. The hash fingerprint, timestamp and device node number are packaged and processed into a new block, and the corresponding parameter set is uploaded to the new block. When the parameter set is uploaded to the new block, the smart contract containing the legal range of the preset process parameters is triggered to automatically check the parameter set. If any parameter in the parameter set exceeds the set tolerance range, the abnormal status will be immediately recorded and a notification will be issued, and an audit request will be automatically sent to the quality control personnel. At the same time, the smart contract will count the verification results of all uploaded parameters in the current time period, generate a compliance rate indicator for the current production process, and attach it to the corresponding block in the chain as a status label.

[0022] After weaving is completed, all weaving data will be uploaded to the chain for storage, and a non-homogeneous certificate corresponding to the brocade clothing fabric used will be generated.

Claims

1. A composite brocade garment fabric weaving process based on multi-source data fusion control, characterized in that: The specific steps of the textile process are as follows: I. Collect multimodal cultural patterns and generate original patterns, while previewing the yarn interweaving effect of the generated patterns; II. Real-time acquisition of the wearer's physiological data and different environmental data, and adjustment of the brocade fabric structure control plan based on the collected data; III. Based on the pattern generation results and the structural control plan, the pattern and tissue structure are designed in a linked manner, and a dynamic tissue database is constructed to store the design data; IV. Based on the design data, the jacquard loom is driven to automatically weave according to the optimized organization and pattern information, while the weaving data is collected and verified in real time; Ⅴ. After weaving is completed, all weaving data will be uploaded to the chain for storage, and a non-homogeneous certificate corresponding to the brocade clothing fabric used will be generated.

2. The composite brocade garment fabric weaving process based on multi-source data fusion control according to claim 1 is characterized in that: The specific steps for generating an original pattern described in step I are as follows: S1.1: Collect data from various data sources, including images of historical artifacts, textual records of traditional brocade weaving techniques, and graphic data on fashion trends. Categorize the collected data by data type and then map each type of data into a unified public semantic space to generate different sets of modal vectors. S1.2: The obtained modal vectors are input into the style encoding network. The style encoding network performs multiple rounds of convolution processing on each modal vector through the built-in convolution block to extract local style features. The attention module then performs weighted summation processing on the local style features extracted by the convolution block to generate the corresponding global style features. S1.3: Analyze the potential structural rules of the patterns in each modal vector through a graph neural network and generate the structural semantics of the corresponding pattern. Use a bidirectional attention mechanism to fuse the collected global style features and the structural semantics of the pattern, and output the corresponding composite semantic features. S1.4: Each set of composite semantic features is input into the CM-GAN network. The CM-GAN network conducts a generator-discriminator game based on the received composite semantic features. The loss values ​​of the generator and discriminator are recorded in real time after each round of the game. Based on the obtained loss values, the parameters of the generator and discriminator are adjusted respectively. The game is repeated alternately until the generator loss value converges to a preset range. S1.5: Input real-time composite semantic features into the CM-GAN network, and use the trained generator to generate original pattern images with cultural semantics. Then, use the brocade physical simulator to perform yarn interweaving calculations on the original pattern images to verify the clarity, symmetry, and integrity of the pattern under the specific weaving structure.

3. The composite brocade garment fabric weaving process based on multi-source data fusion control according to claim 2 is characterized in that: The specific steps of regulating the weave structure of the brocade fabric in step II are as follows: S2.1: A flexible sensor array integrated into the fabric collects the user's physiological and environmental data in real time, synchronizing the timestamps of each sensor transmission channel. Each set of collected physiological and environmental data is then normalized and weighted based on the user's current state. S2.2: Divide the fabric into multiple structural units. Initialize the topological parameters of each structural unit based on the standard brocade parameter library. Then, establish a corresponding adaptive function based on the contribution of each structural unit to the tester's comfort index and its impact on the mechanical performance constraints of the fabric. S2.3: Based on the initial topological parameters of each organizational structure unit, the fitness of each organizational structure unit is calculated using an adaptive function. The adaptive function is then solved using the gradient descent method, and the topological parameters of the corresponding organizational structure unit are updated based on the solution. S2.4: Iteratively update the topological parameters of each tissue structure unit until the fitness of each tissue structure unit converges to a preset range, and then convert the final topological parameters of each tissue structure unit into a corresponding brocade fabric tissue structure control scheme.

4. The composite brocade garment fabric weaving process based on multi-source data fusion control according to claim 3 is characterized in that: The specific steps for linking the design of pattern and tissue structure based on the pattern generation results and the structure control scheme described in step III are as follows: S3.1: Collect the generated original pattern images and the brocade fabric structure control scheme, use image segmentation and texture clustering technology to divide the original pattern images into multiple groups of sub-regions with consistent styles, and assign semantic attributes of the corresponding pattern image to each sub-region manually or through a sample library; S3.2: Calculate the complexity score of the required weave structure for each sub-region based on the intensity of texture detail and the rate of color level change. Then, based on the complexity score of each sub-region, retrieve the appropriate basic weave type. Then, establish the corresponding weave parameter set based on the semantic attributes and weaving complexity of each sub-region. S3.3: Adjust the brocade fabric structure control scheme based on the established weave parameter set, including the warp and weft yarn density, interweaving pattern, and float length limit parameter values. Then, input the adjusted brocade fabric structure control scheme into the weaving process simulator, and calculate the error value between the ideal pattern image and the simulated pattern image through the difference feedback function; S3.4: Based on the error value, a joint optimization strategy is adopted to adjust the pattern and weave parameters simultaneously. Through multiple rounds of iterative optimization, until the error value converges to the preset range, the optimized weave parameters of each sub-region are integrated to generate the weaving structure matrix of the overall fabric. Based on the established weaving structure matrix of the overall fabric, corresponding operation instructions are issued to the CNC loom.

5. The composite brocade garment fabric weaving process based on multi-source data fusion control according to claim 4 is characterized in that: The specific steps for constructing a dynamic organization database to store design data in step III are as follows: S4.1: Collect the structural information of various brocade fabrics through image scanning, microscopic cross-section analysis, and CNC machine files, and annotate the usage scenarios and performance. Then, convert the collected structural information of each fabric sample into the corresponding structural feature vector through the feature encoding model; S4.2: Collect the performance indicators of each fabric structure sample under different usage scenarios, establish the weave structure characteristics and performance, and then calculate the similarity between each fabric structure sample based on the encoded weave feature vector to construct a dynamic weave database based on a graph or tree structure; S4.3: The dynamic tissue database receives the newly generated brocade fabric tissue structure and performance data in real time, and performs incremental updates and error corrections on the existing database. After that, all tissue sample data are standardized into a unified data structure format and stored.

6. The composite brocade garment fabric weaving process based on multi-source data fusion control according to claim 4 is characterized in that: Real-time collection and verification of weaving data as described in step IV S5.1: Various sensor nodes deployed on looms and in the production environment periodically collect various parameters during the brocade fabric production process, pre-process each parameter, label each parameter with a unique identity code, and integrate the labeled parameters into a parameter set. The parameter set is then encoded into a data format recognized by the blockchain. S5.2: Generate a hash fingerprint using the SHA256 encryption algorithm, and package the hash fingerprint, timestamp, and device node number into a new block. The corresponding parameter set is then uploaded to the new block. When the parameter set is uploaded to the new block, a smart contract containing the legal range of the preset process parameters is triggered to automatically verify the parameter set. S5.3: If any parameter in the parameter set exceeds the set tolerance range, the abnormal status will be immediately recorded and a notification will be issued, and an audit request will be automatically sent to the quality control personnel. At the same time, the smart contract will collect statistics on the verification results of all uploaded parameters in the current time period, generate a compliance rate indicator for the current production process, and attach it to the corresponding block in the chain as a status tag.

Images