Textile printing and dyeing auxiliary proportioning optimization method and system based on digital twinning

Abstract

The application provides a textile printing and dyeing auxiliary proportioning optimization method and system based on digital twinning, relates to the technical field of textile printing and dyeing, and comprises the following steps: rendering an initial cloth printing and dyeing twinning model by using a cloth feature data set to obtain a cloth printing and dyeing twinning model; based on the cloth printing and dyeing twinning model, performing auxiliary proportioning scheme optimization by using a scheme evaluation function to meet the constraint of expected printing and dyeing quality, and determining an optimized proportioning scheme set; and respectively performing sampling printing and dyeing tests according to the optimized proportioning scheme set, and evaluating and determining an optimal proportioning scheme to perform printing and dyeing operations on target cloth. The application can solve the technical problem that the existing printing and dyeing auxiliary proportioning method has poor accuracy and precision of auxiliary proportioning scheme setting due to the inability to set a suitable auxiliary proportioning scheme according to the actual state of the cloth after pretreatment, and causes uneven overall printing and dyeing quality of cloth, and can improve the reliability and stability of cloth dyeing, and achieve the effect of improving the overall printing and dyeing quality of cloth.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of textile printing and dyeing, and particularly relates to a textile printing and dyeing auxiliary proportioning optimization method and system based on digital twinning. BACKGROUND

[0002] In the textile printing and dyeing industry, the pretreatment process is a key step to ensure that the cloth can uniformly and completely absorb dyes. However, due to the influence of many factors such as pretreatment process parameters, personnel operation process, working environment, etc., the quality of the cloth after each pretreatment may fluctuate, resulting in uneven cloth state. Such volatility may have a negative impact on the subsequent printing and dyeing process, affecting product quality.

[0003] Traditional printing and dyeing processes usually use uniformly proportioned printing and dyeing auxiliaries for printing and dyeing operations. Although this method is simple, due to the volatility of cloth quality, the overall printing and dyeing quality is uneven, affecting the printing and dyeing quality. For example, when the cloth quality fluctuates greatly, the use of uniformly proportioned printing and dyeing auxiliaries may cause uneven dyeing of some cloth, with obvious color difference, affecting the overall appearance and performance of the product.

[0004] In summary, the existing textile printing and dyeing auxiliary proportioning method cannot set an appropriate auxiliary proportioning scheme according to the actual state of the pretreated cloth, resulting in poor precision and accuracy of the auxiliary proportioning scheme setting, causing the technical problem of uneven overall printing and dyeing quality of the cloth. SUMMARY

[0005] The purpose of the present application is to provide a textile printing and dyeing auxiliary proportioning optimization method and system based on digital twinning, to solve the technical problem of uneven overall printing and dyeing quality of the cloth caused by the existing textile printing and dyeing auxiliary proportioning method, which cannot set an appropriate auxiliary proportioning scheme according to the actual state of the pretreated cloth, resulting in poor precision and accuracy of the auxiliary proportioning scheme setting.

[0006] In view of the above problems, the present application provides a textile printing and dyeing auxiliary proportioning optimization method and system based on digital twinning.

[0007] In a first aspect, the application provides a textile printing and dyeing auxiliary proportioning optimization method based on digital twinning. The method is implemented by a textile printing and dyeing auxiliary proportioning optimization system based on digital twinning, and includes the following steps: based on digital twinning, performing simulation modeling according to initial cloth attribute information, a predetermined printing and dyeing process, and a predetermined printing and dyeing equipment to generate an initial cloth printing and dyeing twin model; performing fixed-point sampling detection on a pretreated target cloth according to a predetermined detection index to obtain a cloth feature dataset, rendering the initial cloth printing and dyeing twin model using the cloth feature dataset to obtain a cloth printing and dyeing twin model; based on the cloth printing and dyeing twin model, performing auxiliary proportioning scheme optimization using a scheme evaluation function with the constraint of meeting an expected printing and dyeing quality to determine an optimized proportioning scheme set, wherein the auxiliary proportioning parameters include auxiliary types, auxiliary concentrations, auxiliary proportions, and auxiliary adding sequences; performing sampling printing and dyeing tests on the target cloth according to the optimized proportioning scheme set, evaluating multiple test results using the scheme evaluation function to determine an optimal proportioning scheme; performing printing and dyeing operations on the target cloth based on the optimal proportioning scheme, and adding the initial cloth attribute information, the predetermined printing and dyeing process, the predetermined printing and dyeing equipment, the cloth feature dataset, and the optimal proportioning scheme to a sample optimization database; training a scheme optimization model using the sample optimization database, and performing subsequent auxiliary proportioning scheme optimization after the scheme optimization model converges.

[0008] In a second aspect, the application also provides a textile printing and dyeing auxiliary proportioning optimization system based on digital twinning, which is used to execute the textile printing and dyeing auxiliary proportioning optimization method based on digital twinning as described in the first aspect, and comprises: a twin model construction module, which is used to simulate and model based on digital twinning according to initial cloth attribute information, a predetermined printing and dyeing process and a predetermined printing and dyeing equipment, to generate an initial cloth printing and dyeing twin model; a twin model rendering module, which is used to perform fixed-point sampling detection on a pretreated target cloth according to a predetermined detection index, to obtain a cloth feature dataset, to render the initial cloth printing and dyeing twin model by using the cloth feature dataset, and to obtain a cloth printing and dyeing twin model; an auxiliary proportioning scheme optimization module, which is used to perform auxiliary proportioning scheme optimization by using a scheme evaluation function based on the cloth printing and dyeing twin model with the constraint of meeting an expected printing and dyeing quality, to determine an optimized proportioning scheme set, wherein the auxiliary proportioning parameters include auxiliary types, auxiliary concentrations, auxiliary proportions and auxiliary adding sequences; a sampling printing and dyeing test module, which is used to perform sampling printing and dyeing tests on the target cloth according to the optimized proportioning scheme set, to evaluate a plurality of test results by using the scheme evaluation function, and to determine an optimal proportioning scheme; an optimization database construction module, which is used to perform printing and dyeing operations of the target cloth based on the optimal proportioning scheme, and to add the initial cloth attribute information, the predetermined printing and dyeing process, the predetermined printing and dyeing equipment, the cloth feature dataset and the optimal proportioning scheme to a sample optimization database; and a scheme optimization model training module, which is used to train a scheme optimization model by using the sample optimization database, and to perform subsequent auxiliary proportioning scheme optimization after the scheme optimization model converges.

[0009] The one or more technical solutions provided in the application have at least the following technical effects or advantages:

[0010] By constructing an initial cloth printing and dyeing twin model according to initial cloth attribute information, a predetermined printing and dyeing process and a predetermined printing and dyeing equipment, then performing fixed-point sampling detection on a pretreated target cloth to obtain a cloth feature dataset, rendering the initial cloth printing and dyeing twin model by using the cloth feature dataset to obtain a cloth printing and dyeing twin model, then performing auxiliary proportioning scheme optimization by using a scheme evaluation function based on the cloth printing and dyeing twin model with the constraint of meeting an expected printing and dyeing quality to determine an optimized proportioning scheme set, further performing sampling printing and dyeing tests according to the optimized proportioning scheme set to evaluate and determine an optimal proportioning scheme, then performing printing and dyeing operations of the target cloth based on the optimal proportioning scheme, and adding the initial cloth attribute information, the predetermined printing and dyeing process, the predetermined printing and dyeing equipment, the cloth feature dataset and the optimal proportioning scheme to a sample optimization database to construct the sample optimization database, finally training a scheme optimization model by using the sample optimization database, and performing subsequent auxiliary proportioning scheme optimization after the scheme optimization model converges, the adaptation of the auxiliary proportioning scheme to the state of the pretreated cloth can be improved, and thus the reliability and stability of cloth dyeing can be improved, and the technical effect of improving the overall printing and dyeing quality of the cloth can be achieved.

[0011] The above description is only a summary of the technical solutions of the present application. In order to enable the technical means of the present application to be more clearly understood, and to be implemented according to the content of the description, and in order to enable the above and other purposes, characteristics and advantages of the present application to be more apparent and easy to understand, the following specific embodiments of the present application are described. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only exemplary, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of the provided drawings.

[0013] Figure 1 Flowchart of the textile printing and dyeing auxiliary proportioning optimization method based on digital twinning of the present application;

[0014] Figure 2 Flowchart of the textile printing and dyeing auxiliary proportioning optimization method based on digital twinning of the present application for auxiliary proportioning scheme optimization;

[0015] Figure 3 Structure diagram of the textile printing and dyeing auxiliary proportioning optimization system based on digital twinning of the present application.

[0016] Explanation of reference signs:

[0017] Twinning model construction module 11, twinning model rendering module 12, auxiliary proportioning scheme optimization module 13, sampling printing and dyeing test module 14, optimization database construction module 15, scheme optimization model training module 16. DETAILED DESCRIPTION

[0018] The present application provides a textile printing and dyeing auxiliary proportioning optimization method and system based on digital twinning, which solves the technical problem of the existing textile printing and dyeing auxiliary proportioning method, i.e. the poor accuracy and fineness of the auxiliary proportioning scheme setting due to the inability to set an appropriate auxiliary proportioning scheme according to the actual state of the pretreated fabric, resulting in uneven overall printing and dyeing quality of the fabric. The adaptability of the auxiliary proportioning scheme to the state of the pretreated fabric can be improved, thereby improving the reliability and stability of fabric dyeing and achieving the technical effect of improving the overall printing and dyeing quality of the fabric.

[0019] Hereinafter, the technical solutions in the present application will be described clearly and completely with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application. In addition, it should be noted that, for the convenience of description, only parts related to the present application are shown in the drawings, not all.

[0020] Embodiment one, please refer to the attached Figure 1 The present application provides a textile printing and dyeing auxiliary proportioning optimization method based on digital twinning, which is applied to a textile printing and dyeing auxiliary proportioning optimization system based on digital twinning, and specifically includes the following steps:

[0021] Step one: based on digital twinning, according to initial cloth attribute information, predetermined printing and dyeing process and predetermined printing and dyeing equipment, simulation modeling is carried out to generate an initial cloth printing and dyeing twin model.

[0022] Specifically, digital twinning technology is a simulation technology based on virtual models. By creating a virtual copy corresponding to a physical entity, the monitoring, prediction and optimization of the physical entity can be realized. In the textile printing and dyeing industry, digital twinning technology can be applied to the printing and dyeing process of cloth. By establishing a digital twinning model of cloth printing and dyeing, the prediction of cloth printing and dyeing quality and the optimization of printing and dyeing process can be realized.

[0023] Firstly, the initial attribute information of the cloth is collected, including fiber type, weaving structure, size, weight, color and other data; then according to the final use and customer requirements of the cloth, the predetermined printing and dyeing process and the predetermined printing and dyeing equipment are determined, wherein the predetermined printing and dyeing process includes dye selection, process flow, etc., and the predetermined printing and dyeing equipment includes equipment specifications, operation parameters, performance indicators, etc. Further based on digital twinning technology, in a three-dimensional simulation platform, according to the initial cloth attribute information, the predetermined printing and dyeing process and the predetermined printing and dyeing equipment, simulation modeling is carried out to construct an initial cloth printing and dyeing twin model. This model will reflect the behavior and state of the cloth in the actual printing and dyeing process. By constructing the initial cloth printing and dyeing twin model based on digital twinning technology, the authenticity and accuracy of cloth printing and dyeing simulation can be improved.

[0024] Step two: according to the predetermined detection index, the pretreated target cloth is sampled and detected at fixed points to obtain a cloth feature data set, and the initial cloth printing and dyeing twin model is rendered using the cloth feature data set to obtain a cloth printing and dyeing twin model.

[0025] Specifically, a predetermined detection index is obtained, wherein the predetermined detection index at least includes pH value, moisture content, whiteness, fiber strength, surface flatness, and thickness uniformity, and a person skilled in the art can also add additional detection indexes according to actual conditions. Then, under the predetermined detection node, the pretreated target fabric is sampled and detected according to the predetermined detection index, wherein multiple points on the fabric can be selected for sampling to ensure that the sampling points can represent the overall state of the fabric, and then professional detection equipment and methods are used to detect the sampling points to collect characteristic data of the fabric, such as pH value, moisture content, whiteness, fiber strength, surface flatness, and thickness uniformity, to obtain a fabric characteristic data set. Further, the initial fabric printing and dyeing twin model is rendered using the fabric characteristic data set to obtain a fabric printing and dyeing twin model, wherein the fabric printing and dyeing twin model can reflect the actual state of the pretreated fabric.

[0026] By constructing an initial fabric printing and dyeing twin model and rendering the initial fabric printing and dyeing twin model using the obtained fabric characteristic data set, the fabric printing and dyeing twin model can be obtained, which can reduce unnecessary simulation modeling time and improve the efficiency of the fabric printing and dyeing twin model.

[0027] Step three: based on the fabric printing and dyeing twin model, using a scheme evaluation function to optimize the auxiliary agent ratio scheme to determine an optimized ratio scheme set, wherein the auxiliary agent ratio parameters include auxiliary agent type, auxiliary agent concentration, auxiliary agent proportion, and auxiliary agent addition sequence.

[0028] Specifically, first, multiple auxiliary agent ratio parameters in the auxiliary agent ratio scheme are obtained, wherein the auxiliary agent ratio parameters include auxiliary agent type, auxiliary agent concentration, auxiliary agent proportion, and auxiliary agent addition sequence, such as selecting suitable auxiliary agent types according to the requirements of the printing and dyeing process and the characteristics of the fabric, including dyes, fixing agents, leveling agents, thickening agents, emulsifiers, dispersants, etc. Then, based on the fabric printing and dyeing twin model and the scheme evaluation function, the auxiliary agent ratio scheme is optimized using an optimization algorithm to obtain an optimized ratio scheme set, with the expected printing and dyeing quality as the optimization constraint condition; wherein the expected printing and dyeing quality can be set according to actual quality requirements, including color fastness index, color index, texture index, and size deviation index; the scheme evaluation function is used to comprehensively evaluate multiple auxiliary agent ratio schemes in the optimization process, wherein the evaluation index includes printing and dyeing quality, auxiliary agent resource consumption, and printing and dyeing efficiency; and the optimized ratio scheme set includes multiple optimal ratio schemes.

[0029] Step four: according to the optimized ratio scheme set, respectively sample and print and dye the target fabric, and use the scheme evaluation function to evaluate multiple test results to determine the optimal ratio scheme.

[0030] Specifically, according to the set of optimized matching schemes, the target fabric is sampled and tested for printing and dyeing, that is, in the testing process, actual printing and dyeing operations are performed according to the selected auxiliary agent matching scheme; then after the test is completed, the printing and dyeing results of the fabric are collected, including color fastness, color, texture, dimensional deviation and other indicators; further, the collected test results are evaluated using a scheme evaluation function, which considers multiple factors such as printing and dyeing quality, auxiliary agent resource consumption and printing and dyeing efficiency to determine the score of each matching scheme; finally, according to the evaluation results, the matching scheme with the highest score is selected as the optimal matching scheme.

[0031] Since the auxiliary agent matching scheme is based on the simulation optimization of the fabric printing and dyeing twin model, it has certain deviation, by obtaining multiple optimal matching schemes and performing subsequent actual tests according to the multiple optimal matching schemes, the uncertainty caused by simulation deviation can be reduced, thereby the accuracy and reliability of the optimal matching scheme setting can be improved without wasting test resources.

[0032] Step five: perform the printing and dyeing operation of the target fabric based on the optimal matching scheme, and add the initial fabric attribute information, the predetermined printing and dyeing process, the predetermined printing and dyeing equipment, the fabric feature dataset and the optimal matching scheme to the sample optimization database.

[0033] Specifically, then according to the optimal auxiliary agent matching scheme, the printing and dyeing operation of the target fabric is performed on the predetermined printing and dyeing process and the predetermined printing and dyeing equipment to ensure that all process parameters and equipment settings meet the requirements of the optimal matching scheme. Further, the initial fabric attribute information, the predetermined printing and dyeing process, the predetermined printing and dyeing equipment, the fabric feature dataset and the optimal matching scheme are added to the sample optimization database as a group of sample data. By constructing the sample optimization database, training data support is provided for subsequent construction of the scheme optimization model.

[0034] Step six: train the scheme optimization model using the sample optimization database, and perform subsequent auxiliary agent matching scheme optimization after the scheme optimization model converges.

[0035] Specifically, the BP neural network is a feedforward artificial neural network that can be iteratively optimized in machine learning, mainly composed of an input layer, a hidden layer and an output layer, which are connected to each other through connection weights, and can be used in the field of prediction analysis. Then, a scheme optimization model is constructed based on the BP neural network, wherein the input data of the input layer of the scheme optimization model is the initial cloth attribute information, the predetermined printing and dyeing process, the predetermined printing and dyeing equipment and the cloth feature data set, and the output data of the output layer is the optimal matching scheme. Then, the sample optimization database is used as sample training data to supervise the training of the scheme optimization model, and a scheme optimization model that meets the expected convergence constraint is obtained, wherein the expected convergence constraint is the model convergence precision, which can be set according to actual needs. Finally, after the scheme optimization model converges, the subsequent auxiliary agent matching scheme optimization is performed. By constructing a scheme optimization model based on a BP neural network to perform subsequent auxiliary agent matching scheme optimization, since the model training data is sample data obtained by optimization, the accuracy is high, so the optimization efficiency of the subsequent auxiliary agent matching scheme can be improved without affecting the accuracy of the auxiliary agent matching scheme optimization.

[0036] The textile printing and dyeing auxiliary agent matching optimization method based on digital twinning is applied to a textile printing and dyeing auxiliary agent matching optimization system based on digital twinning, which can solve the technical problem of the existing textile printing and dyeing auxiliary agent matching method, i.e., the poor accuracy and fineness of the auxiliary agent matching scheme setting due to the inability to set a matching auxiliary agent matching scheme according to the actual state of the pretreated cloth, resulting in uneven overall printing and dyeing quality of the cloth. First, based on digital twinning, an initial cloth printing and dyeing twin model is generated by simulating modeling according to the initial cloth attribute information, the predetermined printing and dyeing process and the predetermined printing and dyeing equipment. Then, the target cloth after pretreatment is detected by point sampling according to the predetermined detection index, and a cloth feature data set is obtained. The cloth feature data set is used to render the initial cloth printing and dyeing twin model to obtain a cloth printing and dyeing twin model. Next, based on the cloth printing and dyeing twin model, an optimal matching scheme set is determined by using a scheme evaluation function for auxiliary agent matching scheme optimization with the expected printing and dyeing quality as a constraint, wherein the auxiliary agent matching parameters include the type, concentration, proportion and addition sequence of the auxiliary agent. Next, the target cloth is sampled and printed and dyed according to the optimal matching scheme set, and the optimal matching scheme is determined by evaluating multiple test results using the scheme evaluation function. Further, the printing and dyeing operation of the target cloth is performed based on the optimal matching scheme, and the initial cloth attribute information, the predetermined printing and dyeing process, the predetermined printing and dyeing equipment, the cloth feature data set and the optimal matching scheme are added to the sample optimization database. Finally, the scheme optimization model is trained using the sample optimization database, and subsequent auxiliary agent matching scheme optimization is performed after the scheme optimization model converges. The adaptability of the auxiliary agent matching scheme to the state of the pretreated cloth can be improved, and thus the reliability and stability of the cloth printing and dyeing can be improved, achieving the technical effect of improving the overall printing and dyeing quality of the cloth.

[0037] Further, the step two of the present application comprises:

[0038] The predetermined detection indexes at least include pH value, moisture content, whiteness, fiber strength, surface flatness and thickness uniformity.

[0039] Specifically, the predetermined detection indexes at least include pH value, moisture content, whiteness, fiber strength, surface flatness and thickness uniformity, and a person skilled in the art can also add additional detection indexes according to the actual situation. For example, the pH value of the cloth is crucial for the use of dyes and auxiliaries, different dyes and auxiliaries have different requirements for pH value, and too high or too low pH value may affect the adsorption of dyes and color fastness; the moisture content of the cloth affects the absorption and coloring effect of the dye, and too high or too low moisture content may cause uneven dyeing.

[0040] Further, the optimization of the auxiliary ratio scheme is performed, as shown in the accompanying Figure 2 The step three of the present application comprises:

[0041] obtaining an auxiliary ratio parameter threshold value, randomly selecting any parameter in the auxiliary ratio parameter threshold value without replacement for combination to generate a first initial auxiliary ratio scheme, iteratively selecting multiple times to obtain multiple initial auxiliary ratio schemes; inputting the multiple initial auxiliary ratio schemes into the cloth printing and dyeing twin model for simulation printing and dyeing to obtain multiple simulation printing and dyeing results; screening the multiple simulation printing and dyeing results according to the expected printing and dyeing quality to determine a set of qualified initial auxiliary ratio schemes, wherein the expected printing and dyeing quality includes color fastness index, color index, texture index and size deviation index, and the texture index includes texture definition, texture alignment accuracy, texture symmetry and texture uniformity; based on the scheme evaluation function and the set of qualified initial auxiliary ratio schemes, the optimization of the auxiliary ratio scheme is performed, and an optimized ratio scheme set is output.

[0042] Specifically, first, an auxiliary ratio parameter threshold value is obtained, wherein the auxiliary ratio parameter threshold value refers to the adjustment range of the auxiliary ratio parameters, including auxiliary type threshold value, auxiliary concentration threshold value, auxiliary proportion threshold value and auxiliary addition sequence threshold value. Then, any parameter in the auxiliary type threshold value, auxiliary concentration threshold value, auxiliary proportion threshold value and auxiliary addition sequence threshold value is randomly selected without replacement for combination to generate a first initial auxiliary ratio scheme; and the same method is iteratively selected multiple times to obtain multiple initial auxiliary ratio schemes that meet the predetermined quantity constraint.

[0043] Then the multiple initial auxiliary agent matching schemes are input into the fabric printing and dyeing twin model to simulate printing and dyeing, the fabric printing and dyeing twin model is used to simulate the physical and chemical changes of the fabric in the printing and dyeing process, including the processes of dye adsorption, diffusion, fixation and the like, and multiple simulation printing and dyeing results are obtained. Further, the multiple simulation printing and dyeing results are screened according to the expected printing and dyeing quality, wherein the expected printing and dyeing quality includes color fastness index, color index, texture index and size deviation index, and the texture index includes texture definition, texture alignment accuracy, texture symmetry and texture uniformity; the initial auxiliary agent matching scheme corresponding to the simulation printing and dyeing result meeting the expected printing and dyeing quality is extracted and set as a qualified initial auxiliary agent matching scheme, and a qualified initial auxiliary agent matching scheme set is obtained. Finally, based on the scheme evaluation function and the qualified initial auxiliary agent matching scheme set, an optimization algorithm is used to optimize the auxiliary agent matching scheme, and an optimized matching scheme set is obtained.

[0044] Further, based on the scheme evaluation function and the qualified initial auxiliary agent matching scheme set, the auxiliary agent matching scheme is optimized, and the application further includes the following steps:

[0045] The qualified initial auxiliary agent matching scheme set is evaluated by using the scheme evaluation function, and multiple fitnesses are output; the qualified initial auxiliary agent matching scheme is regarded as an initial solution, the multiple initial solutions are arranged in descending order of fitness, the first N initial solutions in the initial solution sequence are selected and set as first solutions, and the last M initial solutions are set as tail solutions, N regions are obtained by clustering the M tail solutions with the N first solutions as the center, wherein M is an integral multiple of N; in the N regions, the tail solutions in the region are mutated once in the optimization direction of the first solution according to a predetermined mutation step, and N updated regions are obtained, wherein if the mutated tail solution does not meet the auxiliary agent matching parameter threshold and / or the expected printing and dyeing quality, the mutation is not performed; the N updated regions are identified, and if there is a tail solution whose fitness is greater than that of the first solution in the same updated region, the first solution is replaced by the tail solution; iterative optimization is performed until a predetermined optimization number of times is met, N current regions are output, the N current regions are evaluated based on the scheme evaluation function, a region with the maximum sum of fitnesses is selected as an optimal region, and the optimal region is set as the optimized matching scheme set.

[0046] Specifically, first, the multiple qualified initial auxiliary agent matching schemes in the qualified initial auxiliary agent matching scheme set are evaluated by using the scheme evaluation function to obtain multiple fitness, wherein the higher the fitness, the better the comprehensive effect of the scheme; then the qualified initial auxiliary agent matching scheme is regarded as an initial solution, the multiple initial solutions are arranged in descending order of fitness to generate an initial solution sequence; further, the first N initial solutions of the initial solution sequence are set as first solutions and the last M initial solutions are set as tail solutions, wherein M is an integral multiple of N; then, the M tail solutions are randomly equal-value clustered around the N first solutions to obtain N fields, wherein the number of tail solutions in each field is the same.

[0047] Then, in the N fields, the tail solutions in the fields are mutated once in the optimization direction of the first solution according to a predetermined mutation step, wherein the predetermined mutation step includes an auxiliary agent type adjustment step, an auxiliary agent concentration adjustment step, an auxiliary agent proportion adjustment step and an auxiliary agent addition order adjustment step, wherein in the process of mutating the tail solutions, if the mutated tail solution does not satisfy the auxiliary agent matching parameter threshold and / or the expected printing and dyeing quality, the tail solution is not mutated this time, and N updated fields are obtained. Then, the N updated fields are identified, and if there is a tail solution in the same updated field whose fitness is greater than that of the first solution, the first solution is replaced by the tail solution; the iterative optimization is continuously performed until a predetermined optimization number is reached, wherein the predetermined optimization number can be set according to the expected optimization accuracy, and N current fields are output. Then, the N current fields are evaluated based on the scheme evaluation function, the field with the maximum sum of fitness is selected as the optimal field, and the optimal field is set as the optimized matching scheme set.

[0048] By using the above optimization algorithm for matching scheme optimization, since the global convergence ability of the algorithm is strong, it can avoid falling into local optimum, improve the globality and rationality of optimization, and at the same time, the algorithm can retain multiple optimal results, providing support for subsequent actual test.

[0049] Further, a scheme evaluation function is constructed, and the application further includes the following steps:

[0050] The expression of the scheme evaluation function is: ; wherein, is the evaluation fitness of the i-th scheme, is the printing and dyeing quality weight, is the auxiliary agent resource consumption weight, is the printing and dyeing efficiency weight, is the color fastness weight, is the color weight, is the texture weight, is the size deviation weight, is the color fastness evaluation coefficient of the i-th scheme, is a standard color fastness coefficient, is a color gray value of the i-th scheme, is a standard color gray value, is a texture feature of the i-th scheme, is a standard texture feature, is size data of the i-th scheme, is standard size data, is a total amount of auxiliary resource consumption of the i-th scheme, is a printing and dyeing time length of the i-th scheme.

[0051] Specifically, in the scheme evaluation function, is an evaluation fitness of the i-th scheme, and the higher the evaluation fitness, the better the scheme effect is represented; is a printing and dyeing quality weight, is an auxiliary resource consumption weight, is a printing and dyeing efficiency weight, is a color fastness weight, is a color weight, is a texture weight, is a size deviation weight, is a color fastness evaluation coefficient of the i-th scheme, is a standard color fastness coefficient, is a color gray value of the i-th scheme, is a standard color gray value, is a texture feature of the i-th scheme, is a standard texture feature, is size data of the i-th scheme, is standard size data, is a total amount of auxiliary resource consumption of the i-th scheme, is a printing and dyeing time length of the i-th scheme; wherein the higher the color fastness evaluation coefficient, the better the cloth printing and dyeing quality is represented, the smaller the color deviation is, the better the cloth printing and dyeing quality is represented, the smaller the texture feature deviation is, the better the cloth printing and dyeing quality is represented, wherein the texture feature includes texture definition, texture alignment accuracy, texture symmetry and texture uniformity; the smaller the size deviation is, the better the cloth printing and dyeing quality is represented; the weight can be set according to the influence degree of the index on the whole, and the higher the influence degree is, the greater the weight is. By constructing the scheme evaluation function, a basis is provided for scheme effect evaluation, and meanwhile the accuracy and reliability of scheme evaluation can be improved.

[0052] Further, a plurality of test results are obtained, and the step four of the application includes:

[0053] According to the set of optimized matching schemes, sample printing and dyeing tests are performed on target fabrics to obtain a plurality of fabric printing and dyeing results; a first fabric printing and dyeing result is randomly selected from the plurality of fabric printing and dyeing results, image acquisition and grayscale processing are performed on the first fabric printing and dyeing result, deviation comparison is performed based on the first fabric grayscale and the expected color grayscale to obtain a first color test result; feature extraction is performed on the first fabric grayscale image through a texture feature extraction channel, and deviation comparison is performed based on the first texture feature extraction result and the expected texture feature to obtain a first texture test result, wherein the texture feature extraction channel is constructed based on a convolutional neural network; the first test result is constructed based on the first color test result and the first texture test result and is added to the plurality of test results, wherein the first test result further includes a first color fastness test result and a first size deviation test result.

[0054] Specifically, further according to the set of optimized matching schemes, sample printing and dyeing tests are performed on target fabrics to obtain a plurality of fabric printing and dyeing results; then a first fabric printing and dyeing result is randomly selected from the plurality of fabric printing and dyeing results, and the first fabric printing and dyeing result is any one of the plurality of fabric printing and dyeing results. Then image acquisition and grayscale processing are performed on the first fabric printing and dyeing result, such as using a high-resolution camera or a scanning device to acquire images of the fabric printing and dyeing result, and converting the acquired color images into grayscale images. Through grayscale processing, color information can be eliminated, only brightness information is retained, which helps to more accurately compare color deviation and improve the accuracy of color difference analysis; further deviation comparison is performed based on the grayscale image of the first fabric and the grayscale image of the expected color, which is realized by calculating the difference between the two grayscale images. The smaller the difference is, the smaller the color deviation is, and the first color test result is obtained.

[0055] A texture feature extraction channel is constructed based on a convolutional neural network, wherein the convolutional neural network is a deep learning model used for texture feature extraction from fabric grayscale images. The input data of the texture feature extraction channel is a fabric grayscale image, and the output data is a fabric texture feature. The texture feature extraction channel is trained to a convergent state through sample data. Through texture feature analysis based on the texture feature extraction channel constructed based on the convolutional neural network, the accuracy and efficiency of fabric texture feature analysis can be improved.

[0056] Then feature extraction is performed on the first fabric grayscale image through the texture feature extraction channel to obtain a first texture feature extraction result; further deviation comparison is performed based on the first texture feature extraction result and the expected texture feature to obtain a first texture test result, wherein the smaller the texture feature deviation is, the better the fabric printing and dyeing quality is.

[0057] The first color test result, the first texture test result, the first color fastness test result, and the first size deviation test result are used to construct a first test result, which is added to the plurality of test results.

[0058] Further, the training scheme optimization model is optimized, and the step five of the present application comprises:

[0059] After the number of sample optimization data in the sample optimization database meets the predetermined constraint, the sample optimization database is equally divided into Q parts, and Q times are selected with replacement to construct a first sample set. Q times are iteratively selected to obtain Q sample sets. The sample initial fabric attribute information, the sample printing and dyeing process, the sample printing and dyeing equipment, and the sample fabric feature data are used as inputs, and the sample optimal matching scheme is used as supervision. The Q sample sets are used to supervise the training of the BP neural network to obtain Q convergent scheme optimization units. The scheme optimization model is constructed based on the Q convergent scheme optimization units, wherein the output of the scheme optimization model is the mode of the output results of the Q convergent scheme optimization units.

[0060] Specifically, a predetermined number constraint of sample data is configured, and after the number of sample optimization data in the sample optimization database meets the predetermined number constraint, the sample optimization database is equally divided into Q parts, and Q times are selected with replacement in the Q data sets to construct a first sample set. Q times are iteratively selected using the same method to obtain Q sample sets.

[0061] The initial scheme optimization unit is constructed based on the BP neural network. Then, the sample initial fabric attribute information, the sample printing and dyeing process, the sample printing and dyeing equipment, and the sample fabric feature data are used as input data, and the sample optimal matching scheme is used as supervision data. The Q sample sets are used to supervise the training of the BP neural network. In the process of supervision training, the output result is calculated by forward propagation, and then the connection weight and threshold value are adjusted by error back propagation algorithm. The forward propagation and error back propagation process are repeatedly performed until the deviation between the output result of the network and the sample optimal matching scheme reaches a preset threshold value or the iteration number reaches a preset upper limit, to obtain Q convergent scheme optimization units. Finally, the scheme optimization model is constructed based on the Q convergent scheme optimization units, wherein the output of the scheme optimization model is the mode of the output results of the Q convergent scheme optimization units. By constructing the scheme optimization model based on the Q convergent scheme optimization units, the prediction error of a single convergent scheme optimization unit can be reduced, and the accuracy and reliability of the output of the scheme optimization model can be improved.

[0062] In summary, the textile printing and dyeing auxiliary matching optimization method based on digital twinning provided by the present application has the following technical effects:

[0063] 1. Constructing an initial cloth printing and dyeing twin model according to initial cloth attribute information, a predetermined printing and dyeing process and a predetermined printing and dyeing equipment; then performing fixed-point sampling detection on the pretreated target cloth to obtain a cloth feature data set, and rendering the initial cloth printing and dyeing twin model using the cloth feature data set to obtain a cloth printing and dyeing twin model; then based on the cloth printing and dyeing twin model, using a scheme evaluation function to optimize the auxiliary agent ratio scheme to determine an optimized ratio scheme set while satisfying the expected printing and dyeing quality as a constraint; further performing sampling printing and dyeing tests according to the optimized ratio scheme set to determine an optimal ratio scheme; then performing printing and dyeing operations on the target cloth based on the optimal ratio scheme, and adding the initial cloth attribute information, the predetermined printing and dyeing process, the predetermined printing and dyeing equipment, the cloth feature data set and the optimal ratio scheme to a sample optimization database to construct the sample optimization database; finally training a scheme optimization model using the sample optimization database, and performing subsequent auxiliary agent ratio scheme optimization after the scheme optimization model converges; the adaptation of the auxiliary agent ratio scheme to the state of the pretreated cloth can be improved, thereby improving the reliability and stability of cloth dyeing and achieving the technical effect of improving the overall printing and dyeing quality of the cloth.

[0064] 2. Since the auxiliary agent ratio scheme is a simulated optimization based on the cloth printing and dyeing twin model, it has a certain deviation, and by obtaining multiple optimal ratio schemes and performing subsequent actual tests according to the multiple optimal ratio schemes, the uncertainty caused by the simulation deviation can be reduced, thereby improving the accuracy and reliability of the optimal ratio scheme setting without wasting test resources.

[0065] 3. By constructing a scheme optimization model based on a BP neural network to perform subsequent auxiliary agent ratio scheme optimization, since the model training data is sample data obtained by optimization, the accuracy is high, and therefore the optimization efficiency of the subsequent auxiliary agent ratio scheme can be improved without affecting the accuracy of the auxiliary agent ratio scheme optimization.

[0066] 4. By using the above optimization algorithm to optimize the ratio scheme, since the algorithm has strong global convergence ability, it can avoid falling into local optimum and improve the globality and rationality of the optimization, and the algorithm can retain multiple optimal results to support subsequent actual tests.

[0067] Embodiment two, based on the textile printing and dyeing auxiliary agent ratio optimization method based on digital twinning in the foregoing embodiments, the same invention concept, the present application also provides a textile printing and dyeing auxiliary agent ratio optimization system based on digital twinning, please refer to the attached Figure 3 , including:

[0068] The twin model construction module 11 is configured to simulate and model based on digital twinning according to initial cloth attribute information, a predetermined printing and dyeing process and a predetermined printing and dyeing equipment to generate an initial cloth printing and dyeing twin model.

[0069] The twin model rendering module 12 is configured to perform spot sampling detection on the preprocessed target fabric according to a predetermined detection index, obtain a fabric feature dataset, render the initial fabric printing and dyeing twin model by using the fabric feature dataset, and obtain a fabric printing and dyeing twin model.

[0070] The auxiliary agent ratio scheme optimization module 13 is configured to perform auxiliary agent ratio scheme optimization by using a scheme evaluation function based on the fabric printing and dyeing twin model and with the expected printing and dyeing quality as a constraint, and determine an optimized ratio scheme set, wherein the auxiliary agent ratio parameters include auxiliary agent type, auxiliary agent concentration, auxiliary agent proportion, and auxiliary agent addition sequence.

[0071] The sampling printing and dyeing test module 14 is configured to perform sampling printing and dyeing tests on the target fabric according to the optimized ratio scheme set, respectively, evaluate a plurality of test results by using the scheme evaluation function, and determine an optimal ratio scheme.

[0072] The optimization database construction module 15 is configured to perform printing and dyeing operations on the target fabric based on the optimal ratio scheme, and add the initial fabric attribute information, the predetermined printing and dyeing process, the predetermined printing and dyeing equipment, the fabric feature dataset, and the optimal ratio scheme to a sample optimization database.

[0073] The scheme optimization model training module 16 is configured to train a scheme optimization model by using the sample optimization database, and perform subsequent auxiliary agent ratio scheme optimization after the scheme optimization model converges.

[0074] Further, the twin model rendering module 12 in the digital-twin-based textile printing and dyeing auxiliary agent ratio optimization system is further configured to: the predetermined detection index at least includes pH value, water content, whiteness, fiber strength, surface flatness, and thickness uniformity.

[0075] Further, the auxiliary agent ratio scheme optimization module 13 in the digital-twin-based textile printing and dyeing auxiliary agent ratio optimization system is further configured to: obtain auxiliary agent ratio parameter threshold values, randomly select any parameter in the auxiliary agent ratio parameter threshold values without replacement for combination to generate a first initial auxiliary agent ratio scheme, iteratively select multiple times to obtain a plurality of initial auxiliary agent ratio schemes; input the plurality of initial auxiliary agent ratio schemes into the fabric printing and dyeing twin model to perform simulation printing and dyeing, and obtain a plurality of simulation printing and dyeing results; screen the plurality of simulation printing and dyeing results according to the expected printing and dyeing quality, and determine a set of qualified initial auxiliary agent ratio schemes, wherein the expected printing and dyeing quality includes color fastness index, color index, texture index, and size deviation index, the texture index includes texture definition, texture alignment accuracy, texture symmetry, and texture uniformity; perform auxiliary agent ratio scheme optimization based on the scheme evaluation function and the set of qualified initial auxiliary agent ratio schemes, and output an optimized ratio scheme set.

[0076] Further, the assistant ratio scheme optimization module 13 in the textile printing and dyeing auxiliary ratio optimization system based on digital twinning is further configured to: evaluate the qualified initial auxiliary ratio scheme set by using the scheme evaluation function, and output multiple fitness; regard the qualified initial auxiliary ratio scheme as an initial solution, arrange the multiple initial solutions in descending order of fitness, select the first N initial solutions as the first solutions and the last M initial solutions as the last solutions, cluster the M last solutions around the N first solutions, and obtain N fields, wherein M is an integer multiple of N; in the N fields, take the first solution as the optimization direction, perform one mutation on the last solution in the field according to a predetermined mutation step, and obtain N updated fields, wherein if the mutated last solution does not satisfy the auxiliary ratio parameter threshold and / or the expected printing and dyeing quality, the mutation is not performed; identify the N updated fields, if in the same updated field, the fitness of the last solution is greater than that of the first solution, replace the first solution with the last solution; perform iterative optimization until a predetermined optimization number of times is reached, output N current fields, evaluate the N current fields based on the scheme evaluation function, select a field with the maximum sum of fitness as the optimal field, and set the optimal field as the optimized ratio scheme set.

[0077] Further, the assistant ratio scheme optimization module 13 in the textile printing and dyeing auxiliary ratio optimization system based on digital twinning is further configured to: the expression of the scheme evaluation function is: ; wherein, is the evaluation fitness of the i-th scheme, is the printing and dyeing quality weight, is the auxiliary resource consumption weight, is the printing and dyeing efficiency weight, is the color fastness weight, is the color weight, is the texture weight, is the size deviation weight, is the color fastness evaluation coefficient of the i-th scheme, is the standard color fastness coefficient, and is the color gray value of the i-th scheme, is the standard color gray value, is the texture feature of the i-th scheme, is the standard texture feature, is the size data of the i-th scheme, is the standard size data, is the total amount of auxiliary resource consumption of the i-th scheme, is the printing and dyeing time length of the i-th scheme.

[0078] Further, the sampling printing and dyeing test module 14 in the textile printing and dyeing auxiliary proportioning optimization system based on digital twinning is further configured to: perform sampling printing and dyeing tests on the target fabric according to the set of optimized proportioning schemes respectively to obtain a plurality of fabric printing and dyeing results; randomly select a first fabric printing and dyeing result from the plurality of fabric printing and dyeing results, perform image acquisition and grayscale processing on the first fabric printing and dyeing result, perform deviation comparison based on the first fabric grayscale and the expected color grayscale to obtain a first color test result; perform feature extraction on the first fabric grayscale image through a texture feature extraction channel, perform deviation comparison based on the first texture feature extraction result and the expected texture feature to obtain a first texture test result, wherein the texture feature extraction channel is constructed based on a convolutional neural network; and construct a first test result based on the first color test result and the first texture test result, and add the first test result to the plurality of test results, wherein the first test result further includes a first color fastness test result and a first size deviation test result.

[0079] Further, the scheme optimization model training module 16 in the textile printing and dyeing auxiliary proportioning optimization system based on digital twinning is further configured to: after the number of sample optimization data in the sample optimization database meets a predetermined constraint, divide the sample optimization database into Q parts, and randomly select Q times with replacement to construct a first sample set, and iteratively select Q times to obtain Q sample sets; take the sample initial fabric attribute information, the sample printing and dyeing process, the sample printing and dyeing equipment, and the sample fabric feature data as inputs, take the sample optimal proportioning scheme as supervision, and use the Q sample sets to supervise the training of a BP neural network to obtain Q convergent scheme optimization units; and construct the scheme optimization model based on the Q convergent scheme optimization units, wherein the output of the scheme optimization model is the mode of the output results of the Q convergent scheme optimization units.

[0080] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The textile printing and dyeing auxiliary proportioning optimization system based on digital twinning in the foregoing embodiment one is also applicable to the textile printing and dyeing auxiliary proportioning optimization system based on digital twinning in the present embodiment. Based on the foregoing detailed description of the textile printing and dyeing auxiliary proportioning optimization method based on digital twinning, those skilled in the art can clearly understand the textile printing and dyeing auxiliary proportioning optimization system based on digital twinning in the present embodiment. Therefore, for the sake of brevity of the specification, the textile printing and dyeing auxiliary proportioning optimization system based on digital twinning in the present embodiment will not be described in detail. For the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the relevant part is described in the method part.

[0081] The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Numerous modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without the use of the inventive faculty. Therefore, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0082] It will be readily apparent to one skilled in the art that varying substitutions and modifications can be made to the application disclosed herein without departing from the scope and spirit of the application. Moreover, it is the intent that all such variations and modifications be considered as falling within the scope of the application, and that the application be limited only by the definitions contained in the appended claims.

Claims

1. A method for optimizing the formulation of textile printing and dyeing auxiliaries based on digital twins, characterized in that, include: Based on digital twins, simulation modeling is performed according to the initial fabric attribute information, the predetermined printing and dyeing process and the predetermined printing and dyeing equipment to generate the initial fabric printing and dyeing twin model; The pre-processed target fabric is sampled and tested at fixed points according to the predetermined detection indicators to obtain the fabric feature dataset. The fabric feature dataset is then used to render the initial fabric printing and dyeing twin model to obtain the fabric printing and dyeing twin model. Based on the fabric printing and dyeing twin model, and with the expectation of printing and dyeing quality as a constraint, the scheme evaluation function is used to optimize the auxiliary agent ratio scheme and determine the set of optimized ratio schemes. The auxiliary agent ratio parameters include auxiliary agent type, auxiliary agent concentration, auxiliary agent ratio and auxiliary agent addition order. Based on the set of optimized proportioning schemes, sampling dyeing tests were conducted on the target fabrics. The scheme evaluation function was used to evaluate the multiple test results and determine the optimal proportioning scheme. The target fabric is dyed and printed based on the optimal ratio scheme, and the initial fabric attribute information, the predetermined dyeing and printing process, the predetermined dyeing and printing equipment, the fabric feature dataset and the optimal ratio scheme are added to the sample optimization database. The sample optimization database is used to train a scheme optimization model, and after the scheme optimization model converges, subsequent adjuvant ratio scheme optimization is performed. Construct a solution evaluation function, including: The expression for the scheme evaluation function is: ； in, The fitness of the i-th scheme is evaluated. For the weight of printing and dyeing quality, As a weight for the consumption of auxiliary agent resources, Weighted by dyeing and printing efficiency. As the weight for color fastness, For color weights, For texture weights, As the weight for size deviation, Let be the colorfastness evaluation coefficient for scheme i. The standard color fastness factor is... Let i be the grayscale value of the i-th scheme. Standard color grayscale value, For the texture features of the i-th scheme, For standard texture features, For the size data of the i-th scheme, For standard size data, Let be the total amount of auxiliary resources consumed in the i-th scheme. Let be the dyeing duration for scheme i.

2. The method for optimizing the formulation of textile printing and dyeing auxiliaries based on digital twins according to claim 1, characterized in that, The predetermined testing indicators include at least pH value, moisture content, whiteness, fiber strength, surface smoothness, and thickness uniformity.

3. The method for optimizing the formulation of textile printing and dyeing auxiliaries based on digital twins according to claim 1, characterized in that, Optimizing the formulation of auxiliary agents includes: Obtain the threshold of the auxiliary agent ratio parameter, and randomly select any parameter from the auxiliary agent ratio parameter threshold without replacement to generate the first initial auxiliary agent ratio scheme. Iterate and select multiple times to obtain multiple initial auxiliary agent ratio schemes. The multiple initial auxiliary agent ratio schemes are input into the fabric printing and dyeing twin model for simulated printing and dyeing, resulting in multiple simulated printing and dyeing results. Based on the expected dyeing quality, the multiple simulated dyeing results are screened to determine a set of qualified initial auxiliary agent ratio schemes. The expected dyeing quality includes color fastness index, color index, texture index and size deviation index. The texture index includes texture clarity, texture alignment accuracy, texture symmetry and texture uniformity. Based on the evaluation function of the proposed scheme and the set of qualified initial adjuvant ratio schemes, the adjuvant ratio scheme is optimized, and the set of optimized ratio schemes is output.

4. The method for optimizing the formulation of textile printing and dyeing auxiliaries based on digital twins according to claim 3, characterized in that, Based on the aforementioned scheme evaluation function and the set of qualified initial adjuvant ratio schemes, the adjuvant ratio scheme optimization is performed, including: The qualified initial adjuvant formulation scheme set is evaluated using the scheme evaluation function, and multiple fitness scores are output. The qualified initial additive ratio scheme is regarded as the initial solution. Multiple initial solutions are arranged in descending order of fitness. The first N initial solutions of the initial solution sequence are set as the first solution and the last M initial solutions are set as the last solutions. The M last solutions are clustered with the N first solutions as the center to obtain N neighborhoods, where M is an integer multiple of N. Within N domains, taking the first solution as the optimization direction, the tail solutions within the domain are mutated once according to the predetermined variable asynchronous length to obtain N updated domains. If the mutated tail solutions do not meet the threshold of the auxiliary agent ratio parameter and / or the expected printing and dyeing quality, then this mutation is not performed. Identify N update neighborhoods. If, within the same update neighborhood, the fitness of the tail solution is greater than that of the first solution, then the tail solution is used to replace the first solution. Perform iterative optimization until the predetermined number of optimization attempts is met, output N current domains, evaluate the N current domains based on the scheme evaluation function, select the domain with the largest sum of fitness as the optimal domain, and set the optimal domain as the set of optimized allocation schemes.

5. The method for optimizing the formulation of textile printing and dyeing auxiliaries based on digital twins according to claim 1, characterized in that, Multiple test results were obtained, including: Based on the optimized formula set, sampling dyeing tests were conducted on the target fabrics to obtain multiple fabric dyeing results. Randomly select the first fabric printing result from the multiple fabric printing results, perform image acquisition and grayscale processing on the first fabric printing result, and compare the deviation between the first fabric grayscale and the expected color grayscale to obtain the first color test result. The texture feature extraction channel is used to extract features from the grayscale image of the first fabric. The deviation between the first texture feature extraction result and the expected texture feature is compared to obtain the first texture test result. The texture feature extraction channel is constructed based on a convolutional neural network. A first test result is constructed based on the first color test result and the first texture test result, and added to the plurality of test results. The first test result also includes a first color fastness test result and a first size deviation test result.

6. The method for optimizing the formulation of textile printing and dyeing auxiliaries based on digital twins according to claim 1, characterized in that, The training scheme optimizes the model, including: After the number of sample optimization data in the sample optimization database meets the predetermined constraints, the sample optimization database is divided into Q equal parts, and the first sample set is constructed by selecting Q times with replacement. The first sample set is constructed by iteratively selecting Q times to obtain Q sample sets. Using the initial fabric attribute information, dyeing process, dyeing equipment and fabric feature data as inputs, and the optimal ratio scheme as supervision, the BP neural network is trained by Q sample sets to obtain Q convergent scheme optimization units. The scheme optimization model is constructed based on the Q convergent scheme optimization units, wherein the output of the scheme optimization model is the mode of the output results of the Q convergent scheme optimization units.

7. A textile printing and dyeing auxiliary ratio optimization system based on digital twins, characterized in that, The steps for implementing the textile printing and dyeing auxiliary ratio optimization method based on digital twins as described in any one of claims 1 to 6 include: The twin model construction module is used to perform simulation modeling based on digital twins, according to the initial fabric attribute information, the predetermined dyeing process and the predetermined dyeing equipment, to generate the initial fabric dyeing twin model. The twin model rendering module is used to perform fixed-point sampling and detection on the preprocessed target fabric according to predetermined detection indicators, obtain fabric feature dataset, and use the fabric feature dataset to render the initial fabric printing and dyeing twin model to obtain the fabric printing and dyeing twin model. The auxiliary agent ratio optimization module is used to optimize the auxiliary agent ratio scheme based on the fabric printing and dyeing twin model, with the expected printing and dyeing quality as a constraint, and to determine the set of optimized ratio schemes using the scheme evaluation function. The auxiliary agent ratio parameters include auxiliary agent type, auxiliary agent concentration, auxiliary agent ratio and auxiliary agent addition order. The sampling dyeing test module is used to conduct sampling dyeing tests on the target fabric according to the set of optimized ratio schemes, evaluate the multiple test results using the scheme evaluation function, and determine the optimal ratio scheme. An optimized database construction module is used to perform dyeing and printing operations on the target fabric based on the optimal ratio scheme, and to add the initial fabric attribute information, the predetermined dyeing and printing process, the predetermined dyeing and printing equipment, the fabric feature dataset, and the optimal ratio scheme to the sample optimization database. The scheme optimization model training module is used to train the scheme optimization model using the sample optimization database, and to perform subsequent auxiliary agent ratio scheme optimization after the scheme optimization model converges.

Images