Intelligent management system and method for textile handicrafts

Abstract

The invention discloses an intelligent management system and method for textile handicrafts, and belongs to the technical field of textile handicraft management, and the method specifically comprises the steps: collecting the production related data of the textile handicrafts in real time, carrying out the preprocessing of the collected production related data of the textile handicrafts, and carrying out the intelligent management of the textile handicrafts according to the preprocessed production related data of the textile handicrafts. According to the method, the quality of the textile is predicted and analyzed, the textile process is optimized, the quality and defects of the textile handicrafts are detected by using an image processing technology, and process parameters in textile production are automatically adjusted according to the quality and defect detection results of the textile handicrafts. A supply chain is dynamically adjusted according to market demands, production progress and inventory conditions, and a distribution path, a time table and production are optimized; related data in the production process are monitored in real time, real-time quality control and process parameter adjustment are carried out according to set standards, and the optimal production schedule is automatically generated through intelligent scheduling, so that the textile handicraft management efficiency is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of textile handicraft management, and specifically relates to an intelligent management system and method for textile handicrafts. Background Art

[0002] With the continuous development of the global textile industry and the changes in market demand, the production and management of textile handicrafts are facing more and more challenges. The traditional production management mode of textile handicrafts usually relies on manual operation and a single manual recording method, which is not only prone to errors, but also lacks flexibility and real-time performance.

[0003] This model has exposed a series of problems in the modern production environment, mainly including: 1) Low production efficiency: In the production process of textile handicrafts, the process flow is cumbersome, relies on manual operation, and has low efficiency; 2) Information transmission lag: It is difficult to achieve real-time update and monitoring of information such as production progress, inventory, and quality, resulting in delayed and inaccurate production plans; 3) Lack of intelligent scheduling: When faced with changes in production tasks, traditional production scheduling methods are often difficult to adapt quickly and difficult to make rapid adjustments according to changes in market demand; 4) Quality control is difficult: Due to the lack of real-time quality monitoring methods, textiles are prone to quality defects in the production process, and these defects can often only be discovered after the product is completed, with a high rework rate, which increases production costs; 5) Improper inventory management: Traditional inventory management methods cannot work closely with production progress and market demand, resulting in excessive or insufficient backlogs of raw materials or finished products, affecting production continuity and on-time delivery of orders.

[0004] In order to address these problems, the textile industry urgently needs a more intelligent and automated management method that can improve production efficiency, reduce costs, improve product quality and achieve optimal allocation of resources. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes an intelligent management system and method for textile handicrafts, which can monitor production progress, quality and inventory in real time, intelligently schedule production tasks, optimize resource allocation, and continuously optimize production processes through data analysis, thereby improving production efficiency, reducing costs, and improving quality to meet the needs of modern textile production.

[0006] To achieve the above object, the present invention provides the following technical solutions: An intelligent management method for textile handicrafts comprises the following specific steps: Step S1: collecting data related to the production of textile handicrafts in real time, and preprocessing the collected data related to the production of textile handicrafts; Step S2: predicting and analyzing the quality of textiles and optimizing the textile process according to the pre-processed textile handicraft production related data; Step S3: using image processing technology to detect the quality and defects of textile handicrafts; Step S4: automatically adjusting the process parameters in textile production according to the quality and defect detection results of the textile handicrafts; Step S5: Dynamically adjust the supply chain according to market demand, production progress and inventory status to optimize delivery routes, schedules and production.

[0007] Specifically, the specific steps of step S2 are: Step S201: setting the pre-processed textile handicraft production related data at time t as D(t); Step S202: Establish a quality prediction function, the specific formula is: , Where Q(t) represents the quality prediction function, that is, the quality prediction value of textile handicrafts at time t, D i (t) represents the i-th production-related data of textile handicrafts after preprocessing at time t, n represents the number of types of textile handicraft production-related data, m represents the number of historical data features, D j (t) represents the jth production-related data of textile handicrafts after preprocessing at time t, represents the weight coefficient of the i-th production-related data, represents the interaction effect coefficient between the i-th production-related data and the j-th production-related data, represents the cross terms between process parameters, represents the characteristic value of the k-th historical data feature at time t, Representing historical data characteristics The timing weight coefficient of It represents the temporal dependency between the quality prediction value at time t and the historical process data.

[0008] Specifically, the specific steps of step S2 also include: Step S203: Based on the prediction result of the quality prediction function, the quality deviation of the textile handicraft is minimized, and the quality loss function is defined by the error between the quality prediction function and the actual quality. The specific formula of the quality loss function is: , Where L(t) represents the quality loss function, and represents the weight coefficient, represents the quality prediction value at time t, represents the quality assessment value at time t, represents the absolute value function, represents the ideal process parameter value; Step S204: Utilize the dynamic weight learning method of the feedback mechanism to solve the optimal process parameter configuration, and automatically and adaptively adjust the parameters in the quality prediction function after each production of the textile handicraft is completed.

[0009] Specifically, the specific steps of step S3 are: The images of the finished textile crafts are acquired through industrial cameras and preprocessed, including image denoising and image enhancement; The multi-scale convolutional neural network is used to extract the feature map of the preprocessed textile craft image. The specific formula is: , in, represents the weighted feature map of the textile craft image after preprocessing, represents the activation function, WFC represents the weight of the fully connected layer, b represents the bias of the fully connected layer, Fglobal represents the global feature vector, p and q represent the size of the convolution kernel, respectively. represents the weight of the sth convolution kernel, (x, y) represents the coordinate position of the preprocessed textile handicraft image in the convolution operation, I(x, y) represents the preprocessed textile handicraft image, Represents the textile craft image after preprocessing in the convolution operation; Step S303: Defect classification is performed on the weighted feature map of the pre-processed textile handicraft image. The specific formula is: , in, represents the predicted probability of the cth category, W c represents the weight vector of the cth category, T represents the transpose of the vector, and b c represents the bias term; Step S304: Set the loss function to sum the errors of each category, optimize the multi-scale convolutional neural network, and select the category with the highest prediction probability as the defect category of the textile handicraft.

[0010] Specifically, the specific steps of step S5 are: Step S501: Based on the dynamic inventory management model of forecast error, the inventory level is adjusted in real time. The basic goal of inventory management is set to minimize inventory costs while ensuring timely delivery of orders. The target formula of the inventory management model is: ; Among them, min represents the minimization function, K represents the inventory of textile handicrafts at each moment, , represents the inventory of textile handicrafts at time t, Sc represents the total time, C hold represents the inventory holding cost, C short represents the out-of-stock cost, represents the shortage of textile handicrafts at time t; Step S502: Optimizing the distribution path of textile handicrafts while satisfying the time window constraint; Step S503: Dynamically schedule the supply chain of textile handicrafts. The objective function of dynamic scheduling is: , Among them, A represents the number of production tasks, G represents the number of production lines, and C delay represents the cost of production delay, Dy a represents the delay time of the ath task, C inven Represents the inventory holding cost, Kc a represents the inventory of the ath task, C iload represents the cost of unbalanced load on the production line, Fz g represents the load imbalance of the g-th production line; Step S504: Set constraints to solve the objective function of dynamic scheduling, obtain the demand for textile handicrafts in the supply chain, and produce according to the demand for textile handicrafts.

[0011] Specifically, the preprocessing in step S1 includes: data cleaning, data deduplication and data standardization; The data cleaning involves eliminating invalid, duplicate, erroneous or incomplete data records; The data deduplication, identifying and removing duplicate records or redundant information; The standardization converts data from different sources into a unified format and standard.

[0012] Specifically, an intelligent management system for textile handicrafts is used to implement the intelligent management method for textile handicrafts, including: a data acquisition module, a quality prediction module, a defect detection module, a parameter adjustment module and a dynamic optimization module; The data acquisition module is used to collect textile handicraft production-related data in real time and pre-process the collected textile handicraft production-related data; The quality prediction module is used to predict and analyze the quality of textiles and optimize the textile process based on the pre-processed textile handicraft production related data; The defect detection module is used to detect the quality and defects of textile handicrafts using image processing technology; The parameter adjustment module is used to automatically adjust the process parameters in textile production according to the quality and defect detection results of the textile handicrafts; The dynamic optimization module is used to dynamically adjust the supply chain according to market demand, production progress and inventory status, and optimize delivery routes, schedules and production.

[0013] Specifically, the quality prediction module includes: a quality prediction unit, a quality deviation loss unit and an adaptive adjustment unit; The quality prediction unit is used to establish a quality prediction function to predict the quality of the textile handicraft; The quality deviation loss unit is used to define a quality loss function based on the prediction result of the quality prediction function; The adaptive adjustment unit is used to automatically and adaptively adjust the parameters in the quality prediction function.

[0014] Specifically, the defect detection module includes: a data acquisition and preprocessing unit, a feature extraction unit and a defect analysis unit; The data acquisition and preprocessing unit is used to acquire images of the finished textile handicrafts and perform preprocessing, including image denoising and image enhancement; The feature extraction unit is used to extract a feature map of the preprocessed textile handicraft image using a multi-scale convolutional neural network; The defect analysis unit is used to classify defects on the weighted feature map of the preprocessed textile handicraft image.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention proposes an intelligent management method for textile handicrafts, which monitors various data in the production process (such as temperature and humidity, machine status, operating parameters, etc.) in real time, and performs real-time quality control and adjusts process parameters according to set standards.

[0016] 2. The present invention proposes an intelligent management method for textile handicrafts, which uses machine vision, image recognition and other technologies to detect defects on the fabric surface, and can immediately discover quality problems during the production process, reducing rework and scrap rates.

[0017] 3. The present invention proposes an intelligent management method for textile handicrafts. Through intelligent scheduling, according to changes in market demand and production tasks, it can automatically generate the optimal production schedule based on factors such as order type, production resources, and process requirements, thereby ensuring load balancing of the production line and optimization of the production cycle, and avoiding the inefficiency of manual operations in traditional production scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flow chart of an intelligent management method for textile handicrafts provided by the present invention; Figure 2 This is an architecture diagram of an intelligent management system for textile handicrafts provided by the present invention. DETAILED DESCRIPTION

[0019] The present application is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present application, but do not limit the present application in any form. It should be noted that, for those of ordinary skill in the art, several variations and improvements can also be made without departing from the concept of the present application. These all belong to the protection scope of the present application.

[0020] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0021] It should be noted that, if there is no conflict, the various features in the embodiments of the present application can be combined with each other and are all within the scope of protection of the present application. In addition, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flow chart, in some cases, the steps shown or described can be performed in a different order than the module division in the device or the order in the flow chart. In addition, the " The words "first", "second", "third", etc. do not limit the data and execution order, but only distinguish the same items or similar items with basically the same functions and effects.

[0022] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used in this specification and in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The term "and / or" used in this specification includes any and all combinations of one or more of the related listed items.

[0023] Example 1 See also Figure 1 , an embodiment provided by the present invention: an intelligent management method for textile handicrafts, comprising the following specific steps: Step S1: by installing intelligent sensors and cameras at key nodes of the production line (such as looms, dyeing machines, ironing machines, etc.), real-time data related to the production of textile handicrafts is collected, including but not limited to temperature, humidity, pressure, speed, color and size, etc., and the collected data related to the production of textile handicrafts is preprocessed; The preprocessing in step S1 includes: data cleaning, data deduplication and data standardization; the data cleaning involves eliminating invalid, duplicate, erroneous or incomplete data records. The cleaning process can be completed by setting rules, algorithms, manual intervention, etc., with the purpose of improving data quality and ensuring the accuracy of subsequent analysis; the data deduplication identifies and removes duplicate records or redundant information to ensure that each piece of data appears only once, thereby reducing redundant storage and computing burdens; the standardization has a wide range of data sources, and the formats and units may vary. The purpose of data standardization is to convert data from different sources into a unified format and standard, for example, a unified date format, a unified numerical unit, a unified classification variable coding, etc. The standardization process can help data between different systems work together, facilitating subsequent analysis and processing.

[0024] Step S2: predicting and analyzing the quality of textiles and optimizing the textile process according to the pre-processed textile handicraft production related data; The specific steps of step S2 are: Step S201: setting the pre-processed textile handicraft production related data at time t as D(t); Step S202: Establish a quality prediction function, the specific formula is: , Where Q(t) represents the quality prediction function, that is, the quality prediction value of textile handicrafts at time t, D i (t) represents the i-th production-related data of textile handicrafts after preprocessing at time t, n represents the number of types of textile handicraft production-related data, m represents the number of historical data features, D j (t) represents the jth production-related data of textile handicrafts after preprocessing at time t, Represents the weight coefficient of the i-th production-related data. This coefficient can be dynamically adjusted according to the production history data to reflect the influence of each parameter on the quality. represents the interaction effect coefficient between the i-th production-related data and the j-th production-related data, capturing the contribution of the joint impact of these two parameters to product quality. Represents the cross terms between process parameters and is used to model the nonlinear interaction of multiple process parameters. It represents the characteristic value of the kth historical data feature (such as the quality measurement value at a certain moment in the past, process adjustment record, etc.) at time t. Representing historical data characteristics The temporal weight coefficient describes the influence of the historical data on the current quality prediction. It represents the temporal dependency between the quality prediction value at time t and the historical process data. It is the output obtained by learning the historical quality sequence based on LSTM. The LSTM model predicts the quality evaluation at the current moment by receiving the process parameters (such as temperature, humidity, etc.) and quality measurement values ​​over a period of time in the past. The principle of the above formula: Based on multi-level feature extraction and dynamic spatiotemporal dependency modeling, the quality of textiles is predicted by capturing the complex nonlinear relationship between multiple key process parameters and historical data in the production process. Different from the linear regression and support vector regression methods of traditional models, this function makes full use of the advantages of deep learning and multimodal data fusion, and can more accurately model the spatiotemporal complexity of the process; Multi-parameter interaction: By introducing the interaction term , the model can capture the nonlinear coupling relationship between multiple process parameters, solving the problem that traditional models cannot handle complex interactions; Time series dependency: by combining the LSTM model, the system can automatically learn the quality change rules within a long span from historical data, avoiding the overfitting of short-term data by traditional methods; Step S203: Based on the prediction result of the quality prediction function, the quality deviation of the textile handicraft is minimized, and the quality loss function is defined by the error between the quality prediction function and the actual quality. The specific formula of the quality loss function is: , Where L(t) represents the quality loss function, and represents the weight coefficient, which describes the relative importance between the quality prediction error and the process parameter deviation, represents the quality prediction value at time t, represents the quality assessment value at time t, represents the absolute value function, Indicates the ideal process parameter value, which is determined by production goals or historical data; The principle of the above formula: This objective function contains two parts: one is the quality prediction error term, and the other is the deviation term between the process parameters and the ideal value. By minimizing the objective function, the process parameters are optimized to achieve optimal quality control; Step S204: Utilize the dynamic weight learning method of the feedback mechanism to solve the optimal process parameter configuration, and automatically and adaptively adjust the parameters in the quality prediction function after each production of the textile handicraft is completed.

[0025] The dynamic weight learning method can be a reinforcement learning method or a particle swarm optimization algorithm.

[0026] Step S3: using image processing technology to detect the quality and defects of textile handicrafts; The specific steps of step S3 are: Step S301: acquiring an image of a completed textile handicraft through an industrial camera and performing preprocessing, including image denoising and image enhancement; The image denoising uses Gaussian filtering or bilateral filtering to remove random noise in the textile handicraft image; the image enhancement uses contrast stretching or histogram equalization to enhance the features of the defective area; Step S302: extracting the feature map of the pre-processed textile handicraft image using a multi-scale convolutional neural network. The specific formula is: , in, represents the weighted feature map of the textile craft image after preprocessing, Represents the activation function, the output value is 0-1, W FC represents the weight of the fully connected layer, b represents the bias of the fully connected layer, and F global represents the global feature vector, p and q represent the size of the convolution kernel, represents the weight of the sth convolution kernel, (x, y) represents the coordinate position of the preprocessed textile handicraft image in the convolution operation, I(x, y) represents the preprocessed textile handicraft image, Represents the textile craft image after preprocessing in the convolution operation; In this embodiment, the global feature vector F global The specific formula is: , where U and V represent the height and width of the preprocessed textile crafts image, respectively. ,The principle of this formula: Global average pooling generates a vector representing the global features of the image by averaging the feature values ​​at each position of the feature map. This vector is used as input in the subsequent feature weighting process. By introducing weighted pooling, features of different scales can be combined to improve the perception of important defect information; The principle of the weighted feature map extraction formula is: the convolution kernel is used to perform a sliding window operation on the image I to extract the local features of the image, and the weighted coefficient of each feature is calculated through the fully connected layer. The weighted coefficient is: , which limits the output value to the range of [0,1], indicating the importance of each feature in the final classification; Step S303: Defect classification is performed on the weighted feature map of the pre-processed textile handicraft image. The specific formula is: , in, represents the predicted probability of the cth category, W c represents the weight vector of the cth category, T represents the transpose of the vector, and b crepresents the bias term; Step S304: Set the loss function to sum the errors of each category, optimize the multi-scale convolutional neural network, and select the category with the highest prediction probability as the defect category of the textile handicraft.

[0027] Step S4: automatically adjusting the process parameters in textile production according to the quality and defect detection results of the textile handicrafts; Step S5: Dynamically adjust the supply chain according to market demand, production progress and inventory status to optimize delivery routes, schedules and production.

[0028] The specific steps of step S5 are: Step S501: Based on the dynamic inventory management model of forecast error, the inventory level is adjusted in real time. The basic goal of inventory management is set to minimize inventory costs while ensuring timely delivery of orders. The target formula of the inventory management model is: ; Among them, min represents the minimization function, K represents the inventory of textile handicrafts at each moment, , represents the inventory of textile handicrafts at time t, Sc represents the total time, C hold represents the inventory holding cost, C short represents the out-of-stock cost, represents the shortage of textile handicrafts at time t; The principle of the above formula is to minimize costs by adjusting inventory in real time at different time points to reduce excess inventory and out-of-stock phenomena; Step S502: Optimizing the distribution path of textile handicrafts while satisfying the time window constraint; Specifically: In some delivery scenarios, delivery not only requires the shortest path, but also needs to meet time window constraints. For example, some customers can only receive delivery within a specific time period. In order to deal with this situation, the time window constrained shortest path problem is introduced. The goal is to minimize the total delivery time while ensuring that the delivery task is completed within the time window of each node. Step S503: Dynamically schedule the supply chain of textile handicrafts. The objective function of dynamic scheduling is: , Among them, A represents the number of production tasks, G represents the number of production lines, and C delay represents the cost of production delay, Dy a represents the delay time of the ath task, C inven Represents the inventory holding cost, Kc a represents the inventory of the ath task, C iloadrepresents the cost of unbalanced load on the production line, Fz g represents the load imbalance of the g-th production line; The principle of the above formula is as follows: Production delay cost: Each task delay will bring costs, and the delay time is usually related to the urgency of the production task and the proximity of the delivery date; Inventory cost: When production is delayed or scheduling is inappropriate, it will lead to inventory backlogs, which will increase inventory costs. The increase in inventory means that more storage space and capital are required; Production line load imbalance cost: If the load of the production line is unbalanced, it will affect production efficiency and lead to waste of production time. By optimizing the scheduling of production tasks and reducing load imbalance, the waste of production resources can be reduced; Step S504: Set constraints to solve the objective function of dynamic scheduling, obtain the demand for textile handicrafts in the supply chain, and produce according to the demand for textile handicrafts.

[0029] In this embodiment, the constraints include: production line capacity constraints, which ensure that the load of each production line does not exceed its maximum production capacity; time window constraints, each task must be completed within a predetermined time window, which can help ensure that production tasks do not miss customer demand or delivery dates due to delays; material supply constraints, which ensure that the material supply in the production process does not exceed the inventory, avoiding production interruptions due to material shortages.

[0030] Through the above steps, the textile handicraft supply chain is completed from production and processing to quality assessment, process parameter adjustment, to the final closed loop of the supply chain.

[0031] Example 2 See also Figure 2 , another embodiment provided by the present invention: an intelligent management system for textile handicrafts, comprising: a data acquisition module, a quality prediction module, a defect detection module, a parameter adjustment module and a dynamic optimization module; The data acquisition module is used to collect textile handicraft production-related data in real time and pre-process the collected textile handicraft production-related data; The quality prediction module is used to predict and analyze the quality of textiles and optimize the textile process based on the pre-processed textile handicraft production related data; The defect detection module is used to detect the quality and defects of textile handicrafts using image processing technology; The parameter adjustment module is used to automatically adjust the process parameters in textile production according to the quality and defect detection results of the textile handicrafts; The dynamic optimization module is used to dynamically adjust the supply chain according to market demand, production progress and inventory status, and optimize delivery routes, schedules and production.

[0032] The quality prediction module includes: a quality prediction unit, a quality deviation loss unit and an adaptive adjustment unit; The quality prediction unit is used to establish a quality prediction function to predict the quality of the textile handicraft; The quality deviation loss unit is used to define a quality loss function based on the prediction result of the quality prediction function; The adaptive adjustment unit is used to automatically and adaptively adjust the parameters in the quality prediction function.

[0033] The defect detection module includes: a data acquisition and preprocessing unit, a feature extraction unit and a defect analysis unit; The data acquisition and preprocessing unit is used to acquire images of the finished textile handicrafts and perform preprocessing, including image denoising and image enhancement; The feature extraction unit is used to extract a feature map of the preprocessed textile handicraft image using a multi-scale convolutional neural network; The defect analysis unit is used to classify defects on the weighted feature map of the preprocessed textile handicraft image.

[0034] In addition, the parts of the above-mentioned technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0035] The specific implementation modes as described above further describe the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation mode of the present invention and is not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent management method for textile handicrafts, characterized in that: The specific steps include: Step S1: collecting data related to the production of textile handicrafts in real time, and preprocessing the collected data related to the production of textile handicrafts; Step S2: predicting and analyzing the quality of textiles and optimizing the textile process according to the pre-processed textile handicraft production related data; Step S3: using image processing technology to detect the quality and defects of textile handicrafts; Step S4: automatically adjusting the process parameters in textile production according to the quality and defect detection results of the textile handicrafts; Step S5: Dynamically adjust the supply chain according to market demand, production progress and inventory status to optimize delivery routes, schedules and production.

2. The intelligent management method for textile handicrafts according to claim 1, characterized in that: The specific steps of step S2 are: Step S201: setting the pre-processed textile handicraft production related data at time t as D(t); Step S202: Establish a quality prediction function, the specific formula is: , Where Q(t) represents the quality prediction function, that is, the quality prediction value of textile handicrafts at time t, D i (t) represents the i-th production-related data of textile handicrafts after preprocessing at time t, n represents the number of types of textile handicraft production-related data, m represents the number of historical data features, D j (t) represents the jth production-related data of textile handicrafts after preprocessing at time t, represents the weight coefficient of the i-th production-related data, represents the interaction effect coefficient between the i-th production-related data and the j-th production-related data, represents the cross terms between process parameters, represents the characteristic value of the k-th historical data feature at time t, Representing historical data characteristics The timing weight coefficient of It represents the temporal dependency between the quality prediction value at time t and the historical process data.

3. The intelligent management method for textile handicrafts according to claim 2, characterized in that: The specific steps of step S2 also include: Step S203: Based on the prediction result of the quality prediction function, the quality deviation of the textile handicraft is minimized, and the quality loss function is defined by the error between the quality prediction function and the actual quality. The specific formula of the quality loss function is: ， Where L(t) represents the quality loss function, and represents the weight coefficient, represents the quality prediction value at time t, represents the quality assessment value at time t, represents the absolute value function, represents the ideal process parameter value; Step S204: Utilize the dynamic weight learning method of the feedback mechanism to solve the optimal process parameter configuration, and automatically and adaptively adjust the parameters in the quality prediction function after each production of the textile handicraft is completed.

4. The intelligent management method for textile handicrafts according to claim 3, characterized in that: The specific steps of step S3 are: The images of the finished textile crafts are acquired through industrial cameras and preprocessed, including image denoising and image enhancement; The multi-scale convolutional neural network is used to extract the feature map of the preprocessed textile craft image. The specific formula is: ， in, represents the weighted feature map of the textile craft image after preprocessing, represents the activation function, WFC represents the weight of the fully connected layer, b represents the bias of the fully connected layer, Fglobal represents the global feature vector, p and q represent the size of the convolution kernel, respectively. represents the weight of the sth convolution kernel, (x, y) represents the coordinate position of the preprocessed textile handicraft image in the convolution operation, I(x, y) represents the preprocessed textile handicraft image, Represents the textile craft image after preprocessing in the convolution operation; Step S303: Defect classification is performed on the weighted feature map of the pre-processed textile handicraft image. The specific formula is: ， in, represents the predicted probability of the cth category, W c represents the weight vector of the cth category, T represents the transpose of the vector, and b c represents the bias term; Step S304: Set the loss function to sum the errors of each category, optimize the multi-scale convolutional neural network, and select the category with the highest prediction probability as the defect category of the textile handicraft.

5. The intelligent management method for textile handicrafts according to claim 4, characterized in that: The specific steps of step S5 are: Step S501: Based on the dynamic inventory management model of forecast error, the inventory level is adjusted in real time. The basic goal of inventory management is set to minimize inventory costs while ensuring timely delivery of orders. The target formula of the inventory management model is: ； Among them, min represents the minimization function, K represents the inventory of textile handicrafts at each moment, , represents the inventory of textile handicrafts at time t, Sc represents the total time, C hold represents the inventory holding cost, C short represents the out-of-stock cost, represents the shortage of textile handicrafts at time t; Step S502: Optimizing the distribution path of textile handicrafts while satisfying the time window constraint; Step S503: Dynamically schedule the supply chain of textile handicrafts. The objective function of dynamic scheduling is: ， Among them, A represents the number of production tasks, G represents the number of production lines, and C delay represents the cost of production delay, Dy a represents the delay time of the ath task, C inven Represents the inventory holding cost, Kc a represents the inventory of the ath task, C iload represents the cost of unbalanced load on the production line, Fz g represents the load imbalance of the g-th production line; Step S504: Set constraints to solve the objective function of dynamic scheduling, obtain the demand for textile handicrafts in the supply chain, and produce according to the demand for textile handicrafts.

6. The intelligent management method for textile handicrafts according to claim 5, characterized in that: The preprocessing in step S1 includes: data cleaning, data deduplication and data standardization; The data cleaning involves eliminating invalid, duplicate, erroneous or incomplete data records; The data deduplication, identifying and removing duplicate records or redundant information; The standardization converts data from different sources into a unified format and standard.

7. An intelligent management system for textile handicrafts, used to implement an intelligent management method for textile handicrafts according to any one of claims 1 to 6, characterized in that: include: Data acquisition module, quality prediction module, defect detection module, parameter adjustment module and dynamic optimization module; The data acquisition module is used to collect textile handicraft production-related data in real time and pre-process the collected textile handicraft production-related data; The quality prediction module is used to predict and analyze the quality of textiles and optimize the textile process based on the pre-processed textile handicraft production related data; The defect detection module is used to detect the quality and defects of textile handicrafts using image processing technology; The parameter adjustment module is used to automatically adjust the process parameters in textile production according to the quality and defect detection results of the textile handicrafts; The dynamic optimization module is used to dynamically adjust the supply chain according to market demand, production progress and inventory status, and optimize delivery routes, schedules and production.

8. An intelligent management system for textile handicrafts according to claim 7, characterized in that: The quality prediction module includes: a quality prediction unit, a quality deviation loss unit and an adaptive adjustment unit; The quality prediction unit is used to establish a quality prediction function to predict the quality of the textile handicraft; The quality deviation loss unit is used to define a quality loss function based on the prediction result of the quality prediction function; The adaptive adjustment unit is used to automatically and adaptively adjust the parameters in the quality prediction function.

9. An intelligent management system for textile handicrafts according to claim 8, characterized in that: The defect detection module includes: a data acquisition and preprocessing unit, a feature extraction unit and a defect analysis unit; The data acquisition and preprocessing unit is used to acquire images of the finished textile handicrafts and perform preprocessing, including image denoising and image enhancement; The feature extraction unit is used to extract a feature map of the preprocessed textile handicraft image using a multi-scale convolutional neural network; The defect analysis unit is used to classify defects on the weighted feature map of the preprocessed textile handicraft image.

Images