A production process optimization method for non-iron shirts

Abstract

The present invention relates to the technical field related to fabric production control, and specifically includes a production process optimization method for wrinkle-free shirts, comprising: performing basic detection on a medicated shirt, and if there is no defect, obtaining medicated process information; determining a wrinkle-free qualified probability through backtracking detection, and if it is lower than a threshold, optimizing process parameters, and controlling the shirts by immersion according to the optimization result, thereby solving the technical problem that the optimization in the production process of wrinkle-free shirts focuses on the adjustment of fixed parameters, ignores the complex interaction between multiple factors, and has limited optimization effect, and realizes the study of the relationship between various links, selects high-quality, high-elastic yarns to control the density and twist of the yarns to achieve the wrinkle-free effect, optimizes the components of a chemical treatment agent, finds the most suitable component ratio, and realizes the technical effect of multi-dimensional precise optimization.

Description

Technical Field

[0001] The invention relates to the technical field related to fabric production control, and in particular to a production process optimization method for wrinkle-free shirts. Background Art

[0002] In the production process of conventional wrinkle-free shirts, there is a contradiction between the durability and wearing comfort of wrinkle-free shirts. Although early wrinkle-free shirts are easy to care for, their anti-wrinkle effect will be significantly reduced after multiple washings. At the same time, resin finishing will reduce the strength and wear resistance of the fabric. Secondly, during the processing, the seams of the shirts are prone to become uneven, affecting the appearance. In addition, the wrinkle-free treatment in the prior art has the probability of damaging the natural properties of cotton fibers, such as breathability and skin-friendliness, or introducing harmful substances such as excessive formaldehyde during the processing.

[0003] In summary, the optimization of the production process of wrinkle-free shirts in the prior art focuses on the adjustment of fixed parameters, ignoring the complex interactions among multiple factors, resulting in a technical problem of limited optimization effect. Summary of the invention

[0004] The present application provides a method for optimizing the production process of wrinkle-free shirts, aiming to solve the technical problem in the prior art that the optimization of the production process of wrinkle-free shirts focuses on the adjustment of fixed parameters, ignores the complex interactions among multiple factors, and has limited optimization effect.

[0005] In view of the above problems, the technical solution to implement this application is:

[0006] The present application provides a method for optimizing the production process of wrinkle-free shirts, wherein the method comprises: performing multi-dimensional basic testing on the shirts to be dipped in medicine, and generating pre-production test results;

[0007] When the number of defect identifications included in the pre-production inspection result is equal to 0, basic information of the drug soaking process is obtained, wherein the basic information of the drug soaking process includes the proportion of chemical additives, primary drug soaking control parameters and secondary drug soaking control parameters;

[0008] Perform retrospective testing based on the chemical additive component ratio, the primary immersion control parameters, and the secondary immersion control parameters to generate a probability of no-ironing pass;

[0009] When the non-ironing qualified probability is less than or equal to the qualified probability threshold, the chemical auxiliary agent component ratio, the primary immersion control parameters and the secondary immersion control parameters are adjusted to generate an immersion process optimization result;

[0010] The drug soaking of the shirt to be soaked in drugs is controlled according to the optimization result of the drug soaking process.

[0011] In summary, one or more technical solutions provided in the present application realize the study of the relationship between various links, select high-quality, highly elastic yarns to control the density and twist of the yarns to achieve a wrinkle-free effect, and by soaking the clothes in chemical treatment agents, a protective film is produced on the surface of the clothes to prevent the clothes from wrinkling. The composition of the chemical treatment agent is optimized to find the most suitable ratio of ingredients, thereby achieving a multi-dimensional precise optimization technical effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A schematic flow chart of a method for optimizing the production process of wrinkle-free shirts is provided for this application;

[0013] Figure 2 The present application provides a flow chart for generating a wrinkle-free shirt pass probability in a method for optimizing the production process of wrinkle-free shirts. DETAILED DESCRIPTION

[0014] The present application is described in detail below with reference to the accompanying drawings. Figure 1 As shown, the present application provides a method for optimizing the production process of wrinkle-free shirts, wherein the method comprises:

[0015] S1: Perform multi-dimensional basic inspection on the shirts to be soaked in chemicals and generate pre-production inspection results; S2: When the number of defect identifications included in the pre-production inspection results is equal to 0, obtain basic information on the soaking process, wherein the basic information on the soaking process includes the proportion of chemical additives, primary soaking control parameters and secondary soaking control parameters.

[0016] Long-term monitoring of the production process of wrinkle-free shirts revealed that the production process includes multiple key control points. Specifically, in the selection of fabrics, high-quality materials and advanced textile technology are crucial to the wrinkle-free performance of shirts. Commonly, the pretreatment process of fabrics, such as soaking (immersing the fabric in a chemical treatment agent to create a protective film on the surface of the fabric to prevent the fabric from wrinkling), dyeing and pre-shrinking, are key control points to ensure the stability of fabric quality. Based on this, this application effectively improves the wrinkle resistance of the fabric and maintains the durability of the appearance by reasonably optimizing the chemical formula and its soaking process.

[0017] Specifically, the testing items should be established, and further, the multiple dimensions that need to be tested should be determined. The multiple dimensions should cover the key quality indicators in the early stage of shirt production, including: fabric water absorption test (to evaluate the speed at which the fabric absorbs water, affecting the penetration efficiency of subsequent chemical treatment agents), warp and weft breaking strength test (to measure the force required to stretch the fabric to break in the warp and weft directions, to ensure the durability of the shirt), fabric density test (to measure the fabric density in the warp and weft directions separately, which helps to control the shrinkage rate after immersion in chemicals), seam flatness test (to check whether the seams of the shirt are flat, to ensure appearance quality and wearing comfort).

[0018] Fabric water absorption test, further, drop a standard volume of water drops on a specified position of the shirt, and use a timer to record the time required for complete absorption; warp and weft breaking strength test, further, use a material tester to apply gradually increasing force to the shirt fabric sample until it breaks, and record the breaking force; fabric density test, further, use an electronic fabric density mirror or a microscope to count the number of warp and weft yarns per unit area; seam flatness test, further, visual inspection or use a flatness measuring device to quantify the flatness of the seam area.

[0019] After completing the fabric water absorption test, warp and weft breaking strength test, fabric density test, and seam flatness test of the shirt to be impregnated with drugs, the following also includes: for each test item, if the test result does not meet the established basic test constraints (such as too long water absorption time, lower than the standard breaking strength, abnormal density or uneven seams), it is marked as a defect mark; the number of defect marks is counted, and the defect marks of all test items are summarized. If the total number is 0, it indicates that the shirt has no defects at the basic test level.

[0020] Then, basic information of the immersion process is obtained, and the basic information of the immersion process includes the proportion of chemical additives, primary immersion control parameters and secondary immersion control parameters, wherein the proportion of chemical additives: refer to GB / T 2660-2017 "Shirts" or other relevant standards, set the standard chemical additive formula suitable for the corresponding type of shirts, and ensure that the additive ingredients can effectively form a protective film and meet environmental protection requirements; the primary immersion control parameters: determine the immersion time and immersion temperature, which directly affect the penetration and reaction effect of the chemical additives; the secondary immersion control parameters: set the spray flow rate and atomization time, and the spray flow rate and atomization time control the uniform distribution and curing process of the chemical additives on the fabric surface.

[0021] In addition, if the number of defect marks included in the pre-production inspection results is not equal to 0, it is necessary to adjust the process parameters or pre-treat the shirt according to the specific problem until it reaches the defect-free standard before the immersion process is carried out. Ensure that the shirt to be immersing meets the basic quality requirements before entering the chemical treatment, and improve the qualification rate and quality of the non-ironing shirt products.

[0022] S3: Perform back-testing based on the chemical additive component ratio, the primary immersion control parameters and the secondary immersion control parameters to generate a wrinkle-free pass probability; S4: When the wrinkle-free pass probability is less than or equal to the pass probability threshold, perform process adjustments on the chemical additive component ratio, the primary immersion control parameters and the secondary immersion control parameters to generate an immersion process optimization result; S5: Perform immersion control on the shirt to be immersed based on the immersion process optimization result.

[0023] Conduct retrospective testing on the chemical additive ratio, primary immersion control parameters and secondary immersion control parameters, and generate the probability of passing the non-ironing procedure based on this, and then make process adjustments as needed to optimize the immersion process. Specifically, collect the currently used chemical additive ratio, primary immersion control parameters (including immersion time and temperature) and secondary immersion control parameters (spray flow rate and atomization time);

[0024] Retrospective detection and calculation of the probability of free-ironing qualification, further, spatial coordinate construction, including, according to the proportion of chemical auxiliary ingredients, primary and secondary dipping control parameters to construct corresponding multi-dimensional spatial coordinates, so as to match with user feedback information; characteristic distance calculation, including, calculating the characteristic distance between the current process parameters and the spatial coordinates of successful cases in user feedback, to ensure that all parameters are within an acceptable range; free-ironing qualification probability setting, including, based on the number of user feedbacks that meet a specific threshold for characteristic distance, calculating the free-ironing qualification probability, that is, the probability that the shirt will achieve the expected free-ironing effect under specific conditions.

[0025] Qualified probability evaluation, preferably, if the calculated qualified probability of the non-ironing product is less than or equal to the preset qualified probability threshold, it indicates that the current dipping process needs to be optimized; on the basis that the qualified probability of the non-ironing product is less than or equal to the qualified probability threshold, the process adjustment is started, preferably, it is decided to adjust the proportion of chemical additives, the primary dipping control parameters and the secondary dipping control parameters.

[0026] In one feasible manner, the chemical additive component ratio, the primary drug soaking control parameters and the secondary drug soaking control parameters are process adjusted. Specifically, the drug soaking history log is analyzed, the data is divided into a training set, and a machine learning algorithm (such as a multi-head network and a fitting network) is used to model the relationship between the chemical additive ratio and the drug soaking control parameters and the non-ironing qualified probability; the model is used to predict the non-ironing qualified probability under different parameter combinations, and the optimal parameter combination is found: when there are enough user backtracking data, the user feedback data is directly used; otherwise, the model prediction is used for supplementary analysis; if the non-ironing qualified probability still does not meet the standard after the first adjustment, the parameter fine-tuning is continued in a cycle until the qualified probability threshold is reached or exceeded.

[0027] Then, according to the finally optimized soaking process parameters, the soaking shirts to be soaked are subjected to precisely controlled soaking treatment to ensure that the wrinkle-free effect of the shirts is optimized, and the production process of wrinkle-free shirts is comprehensively optimized to improve product quality and production efficiency.

[0028] Furthermore, the shirt to be dipped in chemicals is subjected to multi-dimensional basic testing to generate pre-production test results. The present application method includes:

[0029] The fabric water absorption of the shirt to be immersed in medicine is tested to obtain the water absorption time of the fabric; the warp and weft breaking strength of the shirt to be immersed in medicine is tested to obtain the warp breaking strength and the weft breaking strength; the fabric density of the shirt to be immersed in medicine is tested to obtain the warp fabric density and the weft fabric density; the seam flatness of the shirt to be immersed in medicine is tested to obtain the seam flatness; the detection indicators of the water absorption time of the fabric, the warp breaking strength, the weft breaking strength, the warp fabric density, the weft fabric density and the seam flatness that do not meet the basic detection constraint conditions are extracted, and a defect identification indicator is generated, which is added to the pre-production detection result.

[0030] Carry out multi-dimensional basic testing on the shirts to be soaked in chemicals and generate pre-production test results. In the preparation stage, refer to industry standards to determine the specific testing standards and constraints for fabric water absorption, warp and weft breaking strength, fabric density, and seam flatness; ensure that all testing equipment such as water absorption testers, tensile testing machines, density measuring instruments, and visual inspection equipment are calibrated to ensure the accuracy of test data.

[0031] Fabric water absorption test: Specifically, several samples are evenly taken from different parts of the shirt to be soaked to ensure that the samples are representative; a water absorption tester is used to record the time required for the fabric to completely absorb a specified amount of water, and the water absorption time of the fabric of each sample is recorded.

[0032] The warp and weft breaking strength test is specifically conducted by using a tensile testing machine to apply gradually increasing forces along the warp and weft directions of the shirt respectively until the fabric breaks, and the maximum force value at the time of break is recorded to obtain the warp and weft breaking strengths.

[0033] Fabric density detection, specifically, uses a fabric density meter or a microscope to measure and calculate the number of warp yarns and weft yarns per square centimeter of the fabric, thereby obtaining the warp fabric density and the weft fabric density.

[0034] Seam flatness detection: Specifically, use a flatness detection instrument to check whether the seams of the shirt are straight and wrinkle-free, and record the seam flatness.

[0035] Then, the test results are evaluated and defects are marked, and the test data is compared with the pre-set basic test constraints, such as the upper limit of water absorption time, the lower limit of breaking strength, the density range and the seam flatness standard; the indicators that do not meet the constraints are marked as defect identification indicators, such as the fabric absorbing water for too long, the warp and weft breaking strength is lower than the standard, the fabric density deviates from the normal range or the seams are uneven, etc.; all defect identification indicators are summarized to form a complete pre-production test result report.

[0036] In actual applications, it also includes deciding whether to rework unqualified shirts, replace raw materials or adjust subsequent processing parameters based on the results of pre-production inspections; if there is no defect mark or the defect is within an acceptable range, the next step of the soaking process can be entered; if the defect is serious, appropriate corrective measures must be taken.

[0037] Ensure that the quality of the wrinkle-free shirts before dipping meets the production requirements, thereby improving the quality stability of the final product.

[0038] Furthermore, the shirt to be dipped in chemicals is subjected to multi-dimensional basic testing to generate pre-production test results. Prior to this, the present application method includes:

[0039] The characteristic values ​​of the pre-production process factors of the shirts to be immersed in drugs are weighted to generate a weight distribution result of the production process factors; the shirts to be immersed in drugs are grouped according to the weight distribution result of the production process factors and the characteristic values ​​of the pre-production process factors to generate a grouping result of the shirts to be immersed in drugs; any group of the grouping results of the shirts to be immersed in drugs is sampled and a preset number is traversed to perform multi-dimensional basic inspections, and the proportion of the inspection results with the number of defect marks equal to 0 is obtained, which is set as the pre-production qualified probability; when the pre-production qualified probability is greater than or equal to the pre-production qualified probability threshold, the number of defect marks of the pre-production inspection results of all the shirts to be immersed in drugs in the corresponding group is configured to 0.

[0040] Multi-dimensional basic testing is carried out on the shirts to be soaked in chemicals to generate pre-production testing results. Before that, group sampling and weight configuration are carried out before the multi-dimensional basic testing of the shirts to be soaked in chemicals. Specifically, for each shirt to be soaked in chemicals, characteristic values ​​reflecting the differences in its pre-production process are collected. The characteristic values ​​of the pre-production process factors include yarn type, weaving process, dyeing uniformity, sewing accuracy, etc.

[0041] Determine the weights of production process factors. Specifically, assign corresponding weights to each collected characteristic value of production process factors based on process experience and historical data analysis. For example, if experience has shown that the weaving process has a greater impact on the final wrinkle-free effect of the shirt, then the weight of the corresponding feature will be higher; then, based on multiple production process factor weights, calculate the comprehensive score of the pre-production process factors of each shirt to form the weight distribution results corresponding to the production process factors, which will help identify the corresponding deviation distribution of the shirt in the process.

[0042] Using cluster analysis or other algorithms, the shirts to be soaked in chemicals are divided into several groups based on the weight distribution results of the production process factors. The shirts in each group have similar pre-production process characteristics, while there are significant differences between the groups.

[0043] In the actual operation process, it also includes configuring the preset number of samplings and setting the preset number of samplings for each group to ensure the statistical validity of the sampling. The sampling number should be set based on the total number of shirts in the group and the variation (variability: measuring the distribution differences of the defect identification corresponding to the shirts in each group).

[0044] For each group, shirt samples are randomly selected according to a preset number, and multi-dimensional basic tests are carried out under the limitations of fabric water absorption, warp and weft breaking strength, fabric density, and seam flatness, and defect identification is recorded.

[0045] The proportion of defective marks of 0 in each group of sampling test results is calculated, which is the pre-production qualified probability corresponding to the group. The pre-production qualified probability reflects the overall consistency of the production process of each group of shirts.

[0046] If the pre-production qualified probability of a group reaches or exceeds the preset qualified probability threshold, it means that the production process of this group of shirts is stable and its production process deviation is small, and the quality of the entire group of shirts can be considered relatively reliable; for the qualified groups, the number of defect identifications in the pre-production inspection results of all shirts in the group is uniformly set to 0, which means that this batch of shirts can directly enter the subsequent drug soaking processing without the need for additional single-piece re-inspection.

[0047] Preferably, a group sampling method is adopted. Even for the same batch of shirts to be dipped in medicine, there are still certain production process deviations. Commonly, sampling is only conducted on the same batch, which is not representative enough. Comparatively, on this basis, the present application uses process deviation group sampling to obtain a more representative shirt group, thereby achieving more refined management of the same batch of shirts, improving the pertinence and efficiency of detection, and ensuring the quality and efficiency of the production of wrinkle-free shirts.

[0048] Furthermore, according to the weight distribution result of the production process factor and the characteristic value of the pre-production process factor, the shirts to be dipped are grouped to generate the grouping result of the shirts to be dipped. The method of the present application includes:

[0049] Obtain a first pre-production process factor characteristic value of a first shirt to be soaked in medicine and a second pre-production process factor characteristic value of a second shirt to be soaked in medicine; normalize the first pre-production process factor characteristic value, and weight it according to the weight distribution result of the production process factor to obtain the pre-production high-dimensional coordinates of the first shirt to be soaked in medicine; normalize the second pre-production process factor characteristic value, and weight it according to the weight distribution result of the production process factor to obtain the pre-production high-dimensional coordinates of the second shirt to be soaked in medicine; when the Euclidean distance between the high-dimensional coordinates of the pre-production of the first shirt to be soaked in medicine and the high-dimensional coordinates of the pre-production of the second shirt to be soaked in medicine is less than or equal to the Euclidean distance threshold, add the first shirt to be soaked in medicine and the second shirt to be soaked in medicine as the same group.

[0050] Collect characteristic values ​​of pre-production process factors. Specifically, obtain samples of shirts to be dipped in chemicals from the production line, and record the characteristic values ​​of pre-production process factors of the first shirt (recorded as shirt A) and the second shirt (recorded as shirt B), respectively, including but not limited to yarn quality, weaving density, pretreatment conditions, etc., which directly or indirectly affect the wrinkle-free performance of the shirts.

[0051] Then, the characteristic values ​​of the pre-production process factors of shirt A and shirt B are normalized. The purpose of normalization is to eliminate the differences in dimensions and numerical ranges between different features and ensure that each feature has the same weight when compared. For example, if a certain characteristic value range is 1-50, it will be converted into a value between 0 and 1 after normalization.

[0052] According to the previously determined production process factor weight distribution results, the normalized eigenvalues ​​of shirt A and shirt B are weighted, which means that each eigenvalue will be multiplied by the corresponding weight coefficient, thereby emphasizing or weakening the influence of certain features on the grouping decision; then, the weighted multiple eigenvalues ​​are combined to form the shirt's pre-production high-dimensional coordinates. For example, if three features are considered and each feature has a weight, then the high-dimensional coordinates of the shirt are a point in the three-dimensional space.

[0053] The Euclidean distance formula is used to calculate the distance between the pre-production high-dimensional coordinates of shirt A and shirt B. Euclidean distance is a method to measure the straight-line distance between two multidimensional points. The formula is: d=sqrt((x_1-x_2)^n+(y_1-y_2)^n+…+(i_1-i_2)^n), where sqrt() represents the square root function, x_1 and x_2 are the values ​​corresponding to the first dimension of the pre-production high-dimensional coordinates of the first shirt to be soaked and the second dimension of the pre-production high-dimensional coordinates of the second shirt to be soaked, y_1 and y_2 are the values ​​corresponding to the first dimension of the pre-production high-dimensional coordinates of the first shirt to be soaked and the second dimension of the pre-production high-dimensional coordinates of the second shirt to be soaked, …, i_1 and i_2 are the values ​​corresponding to the first dimension of the pre-production high-dimensional coordinates of the first shirt to be soaked and the second dimension of the pre-production high-dimensional coordinates of the second shirt to be soaked, and n is the number of feature dimensions.

[0054] If the Euclidean distance between the high-dimensional coordinates of the previous production of shirt A and shirt B is less than or equal to the preset Euclidean distance threshold, it indicates that the similarity of the two shirts in the key production process characteristics meets the grouping requirements, and they are classified into the same group; based on this, repeat until all shirts are grouped. Specifically, repeat for all shirts to be dipped in chemicals on the production line, constantly compare and classify until all shirts are reasonably allocated to each group. In this way, even shirts from the same batch can be carefully distinguished according to slight differences in production processes, thereby improving the pertinence and efficiency of subsequent processing.

[0055] Realize refined grouping of shirts to be dipped in chemicals, ensure more precise adjustment of process parameters, and thus improve the overall quality and production efficiency of wrinkle-free shirts.

[0056] Furthermore, if Figure 2 As shown, according to the chemical additive composition ratio, the primary immersion control parameters and the secondary immersion control parameters, retrospective detection is performed to generate a probability of no-ironing pass. The present application method includes:

[0057] Wherein, the first-level immersion control parameters include immersion time information and immersion temperature information; wherein, the second-level immersion control parameters include spray flow information and atomization time information; according to the chemical auxiliary agent component ratio, the immersion time information, the immersion temperature information, the spray flow information and the atomization time information, user feedback information of the same model of the shirt to be immerged is collected, wherein the user feedback information of the same model of the shirt to be immerged includes the number of wash tolerances; the proportion of information whose number of wash tolerances is greater than or equal to the threshold of the number of wash tolerances is set as the wrinkle-free qualified probability.

[0058] Retrospective testing is performed to generate the probability of passing the wrinkle-free product. Specifically, for the same model of shirts, data on the proportion of chemical additives, primary immersion control parameters (including immersion time and immersion temperature), and secondary immersion control parameters (including spray flow rate and atomization time) of the wrinkle-free shirts that have been produced are collected; actual usage feedback from users is collected through sales records, customer service feedback, or online surveys for the above-mentioned specific models of shirts, with a focus on collecting user-reported tolerance to washes for the model of shirts that still maintains the wrinkle-free effect after multiple washes, to ensure that the collected information covers a sufficiently large user base so that the statistical analysis is representative.

[0059] The chemical additive composition ratio, primary and secondary immersion control parameters of each shirt collected are matched with the corresponding user feedback tolerance number of washes to ensure that each data record is complete; a threshold of tolerance number of washes is determined based on industry standards or set quality goals. The threshold of tolerance number of washes represents the minimum requirement for the shirt to maintain a good wrinkle-free effect after this number of washes.

[0060] Calculate the pass probability of wrinkle-free. Specifically, for each set of parameter settings, count the amount of user feedback information on shirts that can withstand washing times that reach or exceed the threshold. Then, define the ratio of the amount of information that meets the conditions to the total amount of feedback information as the pass probability of wrinkle-free for this set of parameter settings. For example, if, under a certain parameter combination, 80% of users report that their shirts can withstand at least a specified number of washes without wrinkling, then the pass probability of wrinkle-free for the corresponding parameter combination is 80%.

[0061] In practical applications, it also includes evaluating the effects of different chemical auxiliary agent ratios and immersion control parameters based on the calculated probability of wrinkle-free pass. If a parameter combination with a higher probability means that it is easier to achieve the ideal wrinkle-free effect in practical applications, then it is decided whether the current production process needs to be adjusted and optimized, or the best parameter combination to be used in future production is determined. The production process of wrinkle-free shirts can be scientifically evaluated and optimized to ensure that the product meets the expected wrinkle-free performance requirements.

[0062] Furthermore, according to the chemical additive component ratio, the soaking time information, the soaking temperature information, the spray flow information and the atomization time information, user feedback information of the same model of the shirt to be soaked is collected, and the application method includes:

[0063] Obtain the first chemical additive component ratio, first immersion time information, first immersion temperature information, first spray flow rate information and first atomization time information of the first user-uploaded information of the same model of the first shirt to be immersed; construct a first spatial coordinate according to the first chemical additive component ratio, construct a second spatial coordinate according to the first immersion time information and the first immersion temperature information, and construct a third spatial coordinate according to the first spray flow rate information and the first atomization time information; construct a fourth spatial coordinate according to the chemical additive component ratio, construct a fifth spatial coordinate according to the immersion time information and the immersion temperature information, and construct a sixth spatial coordinate according to the spray flow rate information and the atomization time information; calculate a first characteristic distance between the first spatial coordinate and the fourth spatial coordinate, calculate a second characteristic distance between the second spatial coordinate and the fifth spatial coordinate, and calculate a third characteristic distance between the third spatial coordinate and the sixth spatial coordinate; when the first characteristic distance is less than or equal to a first distance threshold, and the second characteristic distance is less than or equal to a second distance threshold, and the third characteristic distance is less than or equal to a third distance threshold, add the first user-uploaded information of the same model of the shirt to be immersed into the feedback information of the same model of the shirt to be immersed.

[0064] Collect information about specific models of wrinkle-free shirts from user feedback channels or databases, including key parameters such as the first chemical additive ingredient ratio, the first soaking time, the first soaking temperature, the first spray flow rate and the first atomization time. This information usually comes from user evaluations after use, questionnaires or directly uploaded usage experience data; organize and classify the collected user data to ensure that each feedback information corresponds to the corresponding shirt model and that all parameters are accurate.

[0065] Each parameter is mapped one by one to construct spatial coordinates. Specifically, based on the information of the first shirt to be soaked in chemicals uploaded by the user, three coordinates are constructed, representing the combination of chemical additive composition and chemical additive ratio (first spatial coordinate), the combination of soaking time and temperature (second spatial coordinate), and the combination of spray flow rate and atomization time (third spatial coordinate); then, referring to the currently used chemical additive ratio, soaking time and temperature, spray flow rate and atomization time as the adjustment starting point, the corresponding coordinates are constructed (the currently used chemical additive ratio corresponds to the fourth spatial coordinate, the currently used soaking time and temperature corresponds to the fifth spatial coordinate, and the currently used spray flow rate and atomization time corresponds to the sixth spatial coordinate) as a comparison benchmark.

[0066] Calculate the characteristic distance, using Euclidean distance or other appropriate distance measurement methods to calculate the distance between the spatial coordinates (first to third) of the user uploaded information and the reference coordinates (fourth to sixth), including the first characteristic distance (the distance between the chemical additive ratio coordinates), the second characteristic distance (the distance between the immersion time and temperature coordinates), and the third characteristic distance (the distance between the spray flow rate and the atomization time coordinates);

[0067] According to the pre-set distance thresholds, such as the first distance threshold, the second distance threshold and the third distance threshold, the degree of closeness of the user data to the standard or optimized parameters is judged: if all feature distances do not exceed their corresponding thresholds, it means that the shirt processing parameters fed back by the user are very close to or consistent with the ideal parameters, reflecting that the stability and consistency of the processing process are good; then, the user feedback information that meets the conditions (the first feature distance is less than or equal to the first distance threshold, and the second feature distance is less than or equal to the second distance threshold, and the third feature distance is less than or equal to the third distance threshold) is used as the user feedback information of the same model of the shirt to be soaked in medicine. This part of the data is considered to be valid and representative, and can be used to further analyze or optimize the process parameters.

[0068] In actual applications, it also includes using the collected and screened user feedback information to analyze the relationship between the wrinkle-free effect of the shirt and the processing parameters, identifying which parameter combinations can best improve the wrinkle-free performance, or discovering potential process improvement points, and making necessary adjustments to the production process of wrinkle-free shirts based on the analysis results, such as adjusting the proportion of chemical additives, immersion time, temperature, spray flow rate and atomization time, etc., in order to achieve a higher wrinkle-free pass rate and user satisfaction.

[0069] A closed-loop management system from user feedback to process optimization is implemented, ensuring continuous improvement of the production process of wrinkle-free shirts and continuous improvement of product quality.

[0070] Furthermore, when the pass probability of the non-ironing product is less than or equal to the pass probability threshold, the chemical additive component ratio, the primary immersion control parameters and the secondary immersion control parameters are adjusted to generate an immersion process optimization result. The method of the present application includes:

[0071] Based on the model of the shirt to be dipped, the dip history log is collected and divided into a first data set, a second data set and a third data set, wherein the dip history log includes a record of the proportion of chemical additive components, a first-level dip control record parameter, a second-level dip control record parameter and wrinkle-free pass probability identification information; according to the first data set, the second data set and the third data set, with the record of the proportion of chemical additive components, the first-level dip control record parameter and the second-level dip control record parameter as input, and with the wrinkle-free pass probability identification information as supervision configuration of the first head network, the second head network and the third head network; with the outputs of the first head network, the second head network and the third head network as input, and with the wrinkle-free pass probability identification information as supervision configuration, the first head network, the second head network and the third head network are configured. The first fitting network is configured; the output layers of the first head network, the second head network and the third head network are merged with the input layer of the first fitting network to generate a free-ironing qualified probability identification model; the chemical auxiliary agent component ratio, the first-level immersion control parameters and the second-level immersion control parameters are adjusted to generate a immersion process adjustment result; the immersion process adjustment result is back-tested: when the amount of back-tested data is greater than or equal to the back-tested data amount threshold, the adjustment result free-ironing qualified probability of the back-test is output; when the amount of back-tested data is less than the back-tested data amount threshold, the adjustment result free-ironing qualified probability is generated through the free-ironing qualified probability identification model; when the adjustment result free-ironing qualified probability is greater than or equal to the qualified probability threshold, the immersion process optimization result is output.

[0072] The process adjustment process when the pass probability of wrinkle-free products does not meet the requirements. Specifically, for the shirt model to be optimized, collect the past operation records of the dipping process to form a dipping history log, including the proportion of chemical additives in each dipping process, the control parameters of the primary dipping (such as dipping time and temperature) and the secondary dipping (such as spray flow and atomization time), and the pass probability of wrinkle-free products of wrinkle-free shirts.

[0073] The drug soaking history log is randomly divided into three independent data sets (the first data set, the second data set and the third data set) for model training and verification; then, the supervision configuration is performed with the non-ironing qualified probability identification information, and the first head network, the second head network and the third head network are constructed respectively, and the input of each network is the chemical auxiliary agent ratio, the first-level drug soaking control parameter and the second-level drug soaking control parameter; the output layers of the first head network, the second head network and the third head network and the input layer of the first fitting network are merged to determine the non-ironing qualified probability identification model, and the non-ironing qualified probability identification model is used to predict the non-ironing qualified probability, and to make supplementary predictions on the adjustment values ​​that are insufficient or not covered by the retrospective user data, so as to improve the comprehensiveness and accuracy of the decision.

[0074] Relatively speaking, since the accuracy of the free-ironing qualified probability identification model is slightly lower than the data accuracy of actual applications, when the number of retrospective users is large enough, user retrospective testing is performed; however, there are also some adjustment values ​​that have not appeared, resulting in a small number of retrospective user samples and non-representativeness. At this time, the model is used for supplementary analysis.

[0075] Then, the process parameters are adjusted according to the situation that needs to be optimized, the proportion of chemical additives, and the primary and secondary soaking control parameters are adjusted to generate a new soaking process adjustment plan.

[0076] Preferably, retrospective detection and evaluation, specifically, if the amount of user feedback collected about the new adjustment plan is sufficient (i.e., retrospective data ≥ retrospective data volume threshold), the corresponding actual data is directly used to evaluate the probability of free-ironing qualification after adjustment; if the amount of retrospective data is insufficient, the free-ironing qualification probability model constructed previously is used to predict the free-ironing qualification probability under the corresponding adjustment plan.

[0077] In the actual application process, if the predicted or actually measured probability of non-ironing qualification after adjustment reaches or exceeds the qualification probability threshold, the obtained process adjustment result is regarded as an effective optimization solution and output as the final optimization result of the dipping process. Combining actual operation data with the predictive ability of machine learning models, it can flexibly respond to various process adjustment scenarios to ensure the continuous optimization and improvement of the non-ironing shirt production process.

[0078] Furthermore, the present application method also includes:

[0079] When the non-ironing qualified probability of the adjustment result is less than the qualified probability threshold, the chemical auxiliary agent component ratio, the primary immersion control parameters and the secondary immersion control parameters are cyclically adjusted to generate the immersion process optimization result.

[0080] When the non-ironing qualified probability of the adjustment result is less than the qualified probability threshold, it is necessary to enter the cyclic process adjustment stage to further optimize the drug soaking process. Specifically, based on the current chemical auxiliary agent component ratio, the current primary drug soaking control parameters and the current secondary drug soaking control parameters, the first round of process adjustment is carried out to determine the first round of drug soaking process optimization results; based on the data review of the first round of drug soaking process optimization results, the chemical auxiliary agent component ratio and the fine-tuning range of drug soaking control parameters at all levels are determined; then, it is necessary to expand the adjustment range, explore parameter combinations that have not been tested before, and make new adjustment attempts one by one or in combination on the chemical auxiliary agent component ratio, the primary drug soaking control parameters and the secondary drug soaking control parameters; each adjustment is regarded as a new cycle;

[0081] At the same time, the wrinkle-free performance of the shirts produced after each new process adjustment is tested, and relevant data is collected, including but not limited to indicators such as the number of washes and the degree of shrinkage; based on the collected data, the wrinkle-free pass probability is recalculated to evaluate whether it has reached or exceeded the pass probability threshold; further, if the new wrinkle-free pass probability is still less than the pass probability threshold, return to the step to continue parameter fine-tuning and strategy adjustment; if the threshold is reached or exceeded, it is considered that a more optimal combination of immersion process parameters has been found, and the cyclic process adjustment is ended.

[0082] In actual application, it also includes recording the final determined chemical additive composition ratio, primary immersion control parameters and secondary immersion control parameters as the default process parameters for producing this model of shirts. Furthermore, because changes in raw materials, environment and other factors will affect the quality of the final product, even if the qualification standards are met, it is necessary to continuously monitor the production data and restart the optimization process when necessary.

[0083] In summary, the beneficial effects of the embodiments of the present application are:

[0084] 1. Through refined process parameter adjustment, the wrinkle resistance and durability of shirts are significantly improved. At the same time, multi-dimensional basic testing is carried out to ensure that each batch of shirts can be adjusted according to its characteristics.

[0085] 2. Introduce user feedback information and machine learning models to monitor the wrinkle-free effect and durability in real time, quickly feedback to the production process, realize immediate process adjustment, and reduce ineffective production.

[0086] 3. Through group sampling, production process factor weight configuration and multi-space coordinate modeling, the interaction of multiple process parameters is comprehensively considered, and multi-parameter collaborative optimization is adopted to achieve multi-dimensional precise optimization and significantly improve the wrinkle-free effect of shirts.

[0087] 4. By adopting the method of obtaining the first pre-production process factor characteristic value of the first shirt to be soaked and the second pre-production process factor characteristic value of the second shirt to be soaked; normalizing the characteristic value of the first pre-production process factor and weighting it by the weight distribution result of the production process factor to obtain the pre-production high-dimensional coordinates of the first shirt to be soaked; normalizing the characteristic value of the second pre-production process factor and weighting it by the weight distribution result of the production process factor to obtain the pre-production high-dimensional coordinates of the second shirt to be soaked; when the Euclidean distance between the pre-production high-dimensional coordinates of the first shirt to be soaked and the pre-production high-dimensional coordinates of the second shirt to be soaked is less than or equal to the Euclidean distance threshold, the first shirt to be soaked and the second shirt to be soaked are added to the same group. Refined grouping of shirts to be soaked is achieved, ensuring more accurate adjustment of process parameters, thereby improving the overall quality and production efficiency of non-ironing shirts.

[0088] In summary, any step can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor, and no unnecessary restrictions are made here.

[0089] Furthermore, the above-mentioned technical scheme only reflects the preferred technical scheme of the technical scheme of the embodiment of the present application. Some changes that may be made to some parts thereof by technicians in this technical field all reflect the novel principles of the embodiment of the present application. Obviously, technicians in this field can make various changes and modifications to the present application without departing from the scope of the present application.

Claims

1. A method for optimizing the production process of non-ironing shirts, characterized in that: include: Conduct multi-dimensional basic testing on the shirts to be soaked in chemicals and generate pre-production testing results; When the number of defect identifications included in the pre-production inspection result is equal to 0, basic information of the drug soaking process is obtained, wherein the basic information of the drug soaking process includes the proportion of chemical additives, primary drug soaking control parameters and secondary drug soaking control parameters; Perform retrospective testing based on the chemical additive component ratio, the primary immersion control parameters, and the secondary immersion control parameters to generate a probability of no-ironing pass; When the non-ironing qualified probability is less than or equal to the qualified probability threshold, the chemical auxiliary agent component ratio, the primary immersion control parameters and the secondary immersion control parameters are adjusted to generate an immersion process optimization result; Controlling the soaking of the shirt to be soaked in chemicals according to the optimization result of the soaking process; Among them, the retrospective detection is performed according to the chemical auxiliary agent component ratio, the primary immersion control parameters and the secondary immersion control parameters to generate the probability of non-ironing pass, including: Wherein, the first-level soaking control parameters include soaking time information and soaking temperature information; Wherein, the secondary immersion control parameters include spray flow information and atomization time information; Collect user feedback information of the same model of the shirt to be soaked according to the chemical additive component ratio, the soaking time information, the soaking temperature information, the spray flow information and the atomization time information, wherein the user feedback information of the same model of the shirt to be soaked includes the number of washing tolerances; The percentage of information that the number of tolerable washing times is greater than or equal to the tolerable washing times threshold is set as the non-ironing qualified probability; Wherein, collecting user feedback information of the same model of shirts to be soaked in medicine according to the chemical auxiliary agent component ratio, the soaking time information, the soaking temperature information, the spray flow information and the atomization time information includes: Obtaining the first chemical additive component ratio, first soaking time information, first soaking temperature information, first spray flow information and first atomization time information of the same model of the first shirt to be soaked in chemicals uploaded by the user; According to the first chemical auxiliary agent component ratio, a first spatial coordinate is constructed, according to the first soaking time information and the first soaking temperature information, a second spatial coordinate is constructed, and according to the first spray flow rate information and the first atomization time information, a third spatial coordinate is constructed; According to the chemical auxiliary agent component ratio, a fourth spatial coordinate is constructed, according to the soaking time information and the soaking temperature information, a fifth spatial coordinate is constructed, and according to the spray flow information and the atomization time information, a sixth spatial coordinate is constructed; Calculating a first characteristic distance between the first spatial coordinate and the fourth spatial coordinate, calculating a second characteristic distance between the second spatial coordinate and the fifth spatial coordinate, and calculating a third characteristic distance between the third spatial coordinate and the sixth spatial coordinate; When the first characteristic distance is less than or equal to a first distance threshold, and the second characteristic distance is less than or equal to a second distance threshold, and the third characteristic distance is less than or equal to a third distance threshold, the user uploaded information of the same model of the first shirt to be soaked in medicine is added to the user feedback information of the same model of the shirt to be soaked in medicine.

2. The method according to claim 1, characterized in that Perform multi-dimensional basic testing on the shirts to be dipped in chemicals and generate pre-production test results, including: Conducting a water absorption test on the shirt to be dipped in chemicals to obtain the water absorption time of the fabric; Performing warp and weft breaking strength tests on the shirt to be dipped in chemicals to obtain warp breaking strength and weft breaking strength; Conducting fabric density testing on the shirt to be dipped in chemicals to obtain warp fabric density and weft fabric density; Performing seam flatness testing on the shirt to be dipped in chemicals to obtain seam flatness; The detection indicators of the fabric water absorption time, the warp breaking strength, the weft breaking strength, the warp fabric density, the weft fabric density and the seam flatness that do not meet the basic detection constraint conditions are extracted, and defect identification indicators are generated and added to the pre-production detection results.

3. The method according to claim 1, characterized in that Perform multi-dimensional basic testing on the shirts to be dipped in chemicals and generate pre-production test results, which previously included: Performing weight configuration on the characteristic values ​​of the pre-production process factors of the shirt to be dipped in chemicals to generate a weight distribution result of the production process factors; According to the weight distribution result of the production process factor and the characteristic value of the pre-production process factor, the shirts to be dipped in chemicals are grouped to generate a grouping result of the shirts to be dipped in chemicals; Traversing any group of the grouping results of the shirts to be dipped in chemicals, sampling a preset number of samples for multi-dimensional basic testing, and obtaining the proportion of the test results with the number of defect marks equal to 0, which is set as the pre-production qualified probability; When the pre-production qualified probability is greater than or equal to the pre-production qualified probability threshold, the number of defect identifications of the pre-production inspection results of all shirts to be dipped in chemicals in the corresponding group is configured to be 0.

4. The method according to claim 3, characterized in that According to the weight distribution result of the production process factor and the characteristic value of the pre-production process factor, the shirts to be dipped are grouped to generate a grouping result of the shirts to be dipped, including: Obtain a first pre-production process factor characteristic value of the first shirt to be dipped in chemicals and a second pre-production process factor characteristic value of the second shirt to be dipped in chemicals; Normalizing the characteristic value of the first pre-production process factor, and weighting it by the weight distribution result of the production process factor, to obtain the pre-production high-dimensional coordinates of the first shirt to be dipped in medicine; Normalizing the characteristic value of the second pre-production process factor, and weighting it according to the weight distribution result of the production process factor, to obtain the pre-production high-dimensional coordinates of the second shirt to be dipped in medicine; When the Euclidean distance between the pre-production high-dimensional coordinates of the first shirt to be soaked in medicine and the pre-production high-dimensional coordinates of the second shirt to be soaked in medicine is less than or equal to the Euclidean distance threshold, the first shirt to be soaked in medicine and the second shirt to be soaked in medicine are added as the same group.

5. The method according to claim 1, characterized in that When the pass probability of the non-ironing product is less than or equal to the pass probability threshold, the chemical additive component ratio, the primary immersion control parameters and the secondary immersion control parameters are adjusted to generate an immersion process optimization result, including: Based on the model of the shirt to be dipped, the dip history log is collected and divided into a first data set, a second data set and a third data set, wherein the dip history log includes chemical additive component ratio records, primary dip control record parameters, secondary dip control record parameters and free-ironing qualified probability identification information; According to the first data set, the second data set and the third data set, the chemical additive component ratio record, the first-level immersion control record parameter and the second-level immersion control record parameter are used as input, and the non-ironing qualified probability identification information is used to supervise the configuration of the first head network, the second head network and the third head network; Taking the outputs of the first head network, the second head network and the third head network as inputs, and supervising the configuration of the first fitting network with the free-iron qualified probability identification information; Merging the output layers of the first head network, the second head network and the third head network with the input layer of the first fitting network to generate a free-iron qualified probability identification model; Performing process adjustment on the chemical auxiliary agent component ratio, the primary immersion control parameters and the secondary immersion control parameters to generate immersion process adjustment results; The adjustment results of the soaking process are retrospectively tested: When the amount of backtracking data is greater than or equal to the backtracking data amount threshold, the probability of qualified free-iron adjustment result of the backtracking detection is output; When the amount of backtracking data is less than the threshold of the amount of backtracking data, the free-iron qualified probability of the adjustment result is generated by the free-iron qualified probability identification model; When the non-ironing qualified probability of the adjustment result is greater than or equal to the qualified probability threshold, the optimization result of the dipping process is output.

6. The method according to claim 5, characterized in that Also includes: When the non-ironing qualified probability of the adjustment result is less than the qualified probability threshold, the chemical auxiliary agent component ratio, the primary immersion control parameters and the secondary immersion control parameters are cyclically adjusted to generate the immersion process optimization result.

Images