Surface quality detection method for wool-like staple fiber yarn

Abstract

The invention provides a wool-like staple fiber yarn surface quality detection method, and relates to the technical field of quality detection.The method comprises the steps that surface quality indexes are traversed, correlation analysis is conducted on the surface quality indexes and production process variables, and a correlation variable set is generated; constructing a quality mapping function of the associated variable set and the surface quality index; calling an associated variable monitoring value set of the associated variable set, performing surface quality prediction through a quality mapping function, and generating a surface quality index prediction value; comparing the surface quality index predicted value with the surface quality index expected value to generate a surface quality deviation index; performing detection priority identification on the surface quality index to generate a detection priority label; and performing directional detection according to the detection priority label. The technical problem that in the prior art, due to the fact that the adjustment accuracy of production process parameters is low, the product surface quality is further affected can be solved, the technical aim of optimizing the production process is achieved, and the technical effect of improving the product surface quality is achieved.

Description

Technical Field

[0001] This application relates to the field of quality inspection technology, and in particular to a method for inspecting the surface quality of wool-like staple fiber yarn. Background Technology

[0002] Surface quality inspection of imitation wool staple fiber yarn refers to the evaluation and measurement of the appearance and surface characteristics of imitation wool staple fiber yarn to ensure that it meets the predetermined quality standards.

[0003] Currently, most existing surface quality inspection processes for imitation wool staple fiber yarns lack analytical indicators for surface quality. This makes it impossible to identify key process parameters, hindering accurate adjustment of these parameters and ultimately preventing effective optimization of surface quality. Furthermore, the lack of targeted adjustment strategies makes quality control even more difficult.

[0004] In summary, existing technologies often rely on correlation analysis between surface quality indicators and production process variables, making it difficult to accurately determine key process parameters and their impact on surface quality. This results in low accuracy in adjusting production process parameters, further affecting product surface quality. Summary of the Invention

[0005] The purpose of this application is to provide a surface quality testing method for wool-like staple fiber yarn, in order to solve the technical problem that in the prior art, most methods rely on correlation analysis between surface quality indicators and production process variable parameters, which makes it difficult to accurately determine key process parameters and their impact on surface quality, resulting in low accuracy in adjusting production process parameters and further affecting product surface quality.

[0006] In view of the above problems, this application provides a method for testing the surface quality of wool-like staple fiber yarn.

[0007] This application provides a surface quality detection method for imitation wool staple fiber yarn, wherein the method includes: traversing surface quality indicators and performing correlation analysis with the production process variables of the imitation wool staple fiber yarn to be tested to generate a set of correlation variables; constructing a quality mapping function between the set of correlation variables and the surface quality indicators based on big data; retrieving the monitoring value set of correlation variables of the set of correlation variables of the imitation wool staple fiber yarn to be tested, and performing surface quality prediction through the quality mapping function to generate predicted values ​​of surface quality indicators; comparing the predicted values ​​of surface quality indicators with the expected values ​​of surface quality indicators to generate a surface quality deviation index; assigning detection priority to the surface quality indicators according to the surface quality deviation index to generate detection priority labels; and performing targeted detection according to the detection priority labels.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: By traversing surface quality indicators and performing correlation analysis with the production process variables of the imitation wool staple fiber yarn to be tested, a set of correlation variables is generated. Based on big data, a quality mapping function is constructed between the set of correlation variables and the surface quality indicators. The monitoring value set of correlation variables of the set of correlation variables of the imitation wool staple fiber yarn to be tested is retrieved, and surface quality is predicted through the quality mapping function to generate predicted values ​​of surface quality indicators. The predicted values ​​of surface quality indicators are compared with the expected values ​​of surface quality indicators to generate a surface quality deviation index. According to the surface quality deviation index, the surface quality indicators are identified by detection priority, and detection priority labels are generated. Targeted detection is carried out according to the detection priority labels to realize the correlation analysis between surface quality indicators and production process variable parameters, determine the technical objectives of key process parameters and their impact on surface quality, optimize the production process, and achieve the technical effect of improving product surface quality.

[0009] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0011] Figure 1 This is a flowchart illustrating the surface quality testing method for a wool-like staple fiber yarn according to this application. Figure 2 This is a flowchart illustrating the process of generating a set of related variables in the surface quality detection method for imitation wool staple fiber yarn of this application. Detailed Implementation

[0012] This application provides a surface quality testing method for wool-like staple fiber yarn, solving the technical problem in existing technologies where the correlation analysis between surface quality indicators and production process variables makes it difficult to accurately determine key process parameters and their impact on surface quality. This results in low accuracy in adjusting production process parameters, further affecting product surface quality. The method achieves the technical objective of performing correlation analysis between surface quality indicators and production process variables to determine key process parameters and their impact on surface quality, thereby optimizing the production process and improving product surface quality.

[0013] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0014] Example Please see the appendix Figure 1 This application provides a method for detecting the surface quality of wool-like staple fiber yarn, wherein the method specifically includes the following steps: Step 1: Traverse the surface quality indicators and perform correlation analysis with the production process variables of the imitation wool staple fiber yarn to be tested to generate a set of correlation variables.

[0015] Specifically, all surface quality indicators are accessed sequentially. Correlation analysis is performed on production process variables related to these surface quality indicators to determine the degree of correlation. A set of correlated variables is generated based on this correlation, which is used to construct the subsequent quality mapping function.

[0016] Step 2: Based on big data, construct a quality mapping function between the set of related variables and the surface quality indicators.

[0017] Specifically, historical production data is obtained based on big data, and a quality mapping function is constructed between a set of correlated variables and surface quality indicators. This quality mapping function is trained using various machine learning algorithms. During training, the set of correlated variables is used as input features, and the surface quality indicators are used as output targets.

[0018] Step 3: Retrieve the monitoring value set of the correlation variables of the correlation variable set of the imitation wool staple fiber yarn to be tested, and perform surface quality prediction through the quality mapping function to generate predicted values ​​of surface quality indicators.

[0019] Specifically, the actual monitoring values ​​of the correlation variable set of the imitation wool staple fiber yarn to be tested are retrieved to obtain the correlation variable monitoring value set. The correlation variable monitoring value set is input into the quality mapping function, and the predicted values ​​of the surface quality index are generated through calculation by the quality mapping function.

[0020] Step 4: Compare the predicted values ​​of the surface quality indicators with the expected values ​​of the surface quality indicators to generate a surface quality deviation index.

[0021] Specifically, the predicted value of the surface quality index is compared with the expected value. If the predicted value does not meet the expected value, such as exceeding the allowable error range, the degree of deviation is calculated to generate a surface quality deviation index.

[0022] Step 5: Based on the surface quality deviation index, assign detection priority to the surface quality indicators and generate detection priority labels.

[0023] Specifically, surface quality indicators are sorted or graded according to the magnitude of the surface quality deviation index.

[0024] The higher the surface quality deviation index, the higher the detection priority of the surface quality indicator, and vice versa. A detection priority label is generated for each surface quality indicator.

[0025] Step 6: Perform targeted detection based on the detection priority label.

[0026] Specifically, surface quality indicators are targeted for testing based on testing priority labels. The higher the testing priority, the higher the ranking or efficiency of surface quality indicator testing, and vice versa.

[0027] The aforementioned surface quality testing method for imitation wool staple fiber yarn can perform correlation analysis between surface quality indicators and production process variable parameters, determine the technical objectives of key process parameters and their impact on surface quality, optimize the production process, and achieve the technical effect of improving product surface quality.

[0028] Furthermore, this application also includes: Obtain a first surface quality index, wherein the first surface quality index is any one of hairiness, crimp, gloss, elongation, and pilling amount over a preset time; obtain a set of production process variables for the imitation wool staple fiber yarn to be tested; collect the production log of the imitation wool staple fiber yarn based on the first surface quality index and the set of production process variables, wherein the production log of the imitation wool staple fiber yarn includes recorded values ​​of production process variables and recorded values ​​of the first surface quality index; perform correlation analysis on the recorded values ​​of the first surface quality index and the recorded values ​​of the production process variables to generate a correlation degree set; add production process variables whose correlation degree set is greater than or equal to the correlation degree threshold to the first correlation variable set of the first surface quality index; add the first correlation variable set to the correlation variable set.

[0029] Specifically, such as Figure 2 As shown, a first surface quality index is randomly obtained from the surface quality indices. This first surface quality index is any one of the following: fuzziness, curl, gloss, elongation, and pilling density after a preset time. The pilling density after a preset time refers to the pilling density reported by the user after a preset application time.

[0030] Then, the set of production process variables for the wool-like staple fiber yarn to be tested is obtained. For example, these include raw material type, spinning temperature, spinning speed, stretch ratio, heat setting temperature, etc.

[0031] Next, based on the first surface quality index and the set of production process variables, the actual values ​​of the production process variables are recorded in real time or periodically during the production process, and the surface quality index values ​​of the corresponding batch of imitation wool staple fiber yarn are recorded, which are then integrated to form an imitation wool staple fiber yarn production log. The imitation wool staple fiber yarn production log includes the recorded values ​​of the production process variables and the recorded values ​​of the first surface quality index.

[0032] Next, a correlation analysis is performed on the recorded values ​​of the first surface quality index and the recorded values ​​of the production process variables. The correlation degree between each production process variable and the first surface quality index is calculated, and a correlation degree set is generated.

[0033] Then, based on the results of statistical analysis, a correlation threshold is set, and production process variables that are greater than or equal to the correlation threshold in the correlation set are added to the first correlation variable set of the first surface quality index.

[0034] Next, the first set of associated variables is added to the total set of associated variables for subsequent quality prediction, control, or optimization.

[0035] By optimizing the correlation variables, surface quality can be improved or production costs can be reduced.

[0036] Furthermore, this application also includes: The first surface quality index record value and the production process variable record value are reconstructed from nearest neighbor to generate the first surface quality index feature value and the production process variable feature value; a correlation analysis benchmark sequence is constructed based on the first surface quality index feature value; a correlation analysis comparison sequence is constructed based on the production process variable feature value; and grey relational analysis is performed based on the correlation analysis benchmark sequence and the correlation analysis comparison sequence to generate the correlation degree set.

[0037] Specifically, the recorded values ​​of the first surface quality index and the recorded values ​​of the production process variables are reconstructed from nearest neighbor to extract more representative features, and feature values ​​of the first surface quality index and the production process variables are generated as new data that can represent the characteristics of the original data.

[0038] Then, the correlation analysis requires comparing the correlation strength of other sequences. The first surface quality index characteristic value is used as the baseline sequence for the correlation analysis. The baseline sequence for the correlation analysis should be an ordered set of data, which can be a time series, sequential series, etc., ensuring that each data point in the sequence represents a specific point in time or condition.

[0039] Next, the alignment sequence is used to compare with the baseline sequence. The characteristic values ​​of the production process variables are used as the alignment sequence for constructing the correlation analysis.

[0040] Next, grey relational analysis is a method used to analyze the degree of correlation between the baseline sequence and the comparison sequence in relational analysis. Since the information of the baseline sequence and the comparison sequence in relational analysis is incomplete or the data is inaccurate, grey relational analysis is used to generate a set of correlation degrees.

[0041] Feature values ​​are extracted by proximity reconstruction, and grey relational analysis is used to evaluate the degree of correlation between production process variables and surface quality indicators. This helps to identify production process variables that have a significant impact on surface quality indicators, thereby optimizing the production process and improving product quality.

[0042] Furthermore, this application also includes: Obtain a first production process variable record value, wherein the first production process variable record value has a unique first simultaneous surface quality index record value corresponding to the first surface quality index record value; obtain a second production process variable record value, wherein the second production process variable record value has a unique second simultaneous surface quality index record value corresponding to the first surface quality index record value; calculate the variable deviation modulus set of the first and second production process variable record values; statistically analyze the proportion of the variable deviation modulus set that does not satisfy the variable attribute deviation threshold set, and generate a first deviation coefficient; for the variable... The square root of the sum of the squares of the deviation modulus set is used to generate the second deviation coefficient. When the first deviation coefficient is less than or equal to the first deviation coefficient threshold and the second deviation coefficient is less than or equal to the second deviation coefficient threshold, either the first production process variable record value or the second production process variable record value is stored as the production process variable feature value of the first and second combined surface quality index record values. The analysis is iterated and iterated. When any two production process variable feature values ​​cannot be updated to be adjacent, the central tendency analysis is performed on the multiple combined surface quality index record values ​​of each production process variable feature value to generate the first surface quality index feature value.

[0043] Specifically, the first production process variable record value is randomly extracted based on the production process variable record value. The first production process variable record value is matched with the first surface quality index record value to obtain a uniquely corresponding first combined surface quality index record value, that is, to obtain the surface quality index parameter that changes with the production process.

[0044] Then, a second production process variable record value is randomly extracted from the production process variable record values. The second production process variable record value is matched with the first surface quality index record value to obtain a uniquely corresponding second combined surface quality index record value.

[0045] Next, the variable deviation modulus is used to quantify the difference between the recorded values ​​of the two variables. For example, it can be obtained through absolute difference, Euclidean distance, etc. The variable deviation modulus set for the first and second production process variable recorded values ​​is calculated.

[0046] Next, the variable attribute deviation threshold set refers to the deviation threshold of each variable attribute. The number of variables whose deviation modulus sets do not meet the variable attribute deviation threshold set is counted, and the ratio of the number of variables not meeting the variable attribute deviation threshold set to the total number of variables is calculated to generate the first deviation coefficient.

[0047] Then, the squares of all the variable deviation moduli in the variable deviation moduli set are summed, and then the square root is taken to obtain the comprehensive deviation metric, i.e., the second deviation coefficient.

[0048] Next, if the first deviation coefficient is less than or equal to the first deviation coefficient threshold, and the second deviation coefficient is less than or equal to the second deviation coefficient threshold, it indicates that the deviation between the first and second production process variable recorded values ​​is within an acceptable range, i.e., the deviation is small. Therefore, the first and second production process variable recorded values ​​belong to adjacent data. Either the first or second production process variable recorded value is selected and stored as a production process variable feature value associated with the first and second joint surface quality index recorded values, thus retaining one of the joint surface quality index recorded values. The first and second deviation coefficient thresholds are set by those skilled in the art based on actual conditions.

[0049] Next, the deviation modulus set is calculated for any two production process variable records, and the deviation coefficient is calculated for each. This process continues until the characteristic value of the production process variable for the combined surface quality index is generated. If any two characteristic values ​​of the production process variable cannot be updated with neighboring data, it means that there is no characteristic value of the production process variable for the combined surface quality index obtained by calculating the neighboring characteristic value of the production process variable. In this case, a central tendency analysis is performed on multiple combined surface quality index records for each characteristic value of the production process variable, such as calculating the mean and median, to generate the first characteristic value of the surface quality index and identify the main characteristics or trends of the surface quality index.

[0050] By generating the first surface quality index feature value and the production process variable feature value, meaningful feature values ​​can be extracted from the complex production process variable records and surface quality index data, providing data support for subsequent quality control, process optimization, etc.

[0051] Furthermore, this application also includes: Obtain the correlation set of the correlation variable set of the first surface quality index of the surface quality index, and configure the correlation variable weight set; according to the correlation variable set and the surface quality index, collect the same model of historical production data of the imitation wool staple fiber yarn to be tested; preprocess the historical value of the correlation variable of the same model of historical production data using the correlation variable weight set to generate the input dataset; set the historical value of the first surface quality index of the same model of historical production data as the output supervision data; configure the quality mapping function according to the output supervision data and the input dataset.

[0052] Specifically, a first surface quality index is randomly selected from the surface quality indices. A correlation set corresponding to the set of associated variables for this first surface quality index is then extracted. The correlation set refers to the degree of association between each associated variable and the surface quality index. Weights are assigned to each associated variable based on the correlation set, reflecting the importance of each associated variable in predicting or evaluating surface quality.

[0053] Then, historical production data of imitation wool staple fiber yarn of the same model as the imitation wool staple fiber yarn to be tested are found and collected.

[0054] Next, the historical values ​​of the related variables in the historical production data of the same model are weighted using the configured set of related variable weights. Other preprocessing steps, such as data cleaning, standardization, or normalization, can also be performed to improve data quality and model performance. From the preprocessed historical values ​​of the related variables, the historical values ​​of the first surface quality index corresponding to the first surface quality index are extracted as the input dataset for subsequent model training or evaluation.

[0055] Next, the historical value of the first surface quality index in the historical production data of the same model is set as the output supervision data to guide the training or evaluation of the model and ensure that the model can accurately predict or evaluate the surface quality.

[0056] Then, by configuring a quality mapping function using various machine learning algorithms, the quality mapping function is trained based on the output supervised data and the input dataset, mapping the associated variables to surface quality indicators.

[0057] By configuring a quality mapping function based on historical data to predict the surface quality of the imitation wool staple fiber yarn to be inspected, it is helpful to improve production efficiency and product quality stability.

[0058] Furthermore, this application also includes: A first quality mapping function is generated by configuring a support vector machine based on the output supervision data and the input dataset; a second quality mapping function is generated by configuring a random forest based on the output supervision data and the input dataset; a third quality mapping function is generated by configuring a backpropagation neural network based on the output supervision data and the input dataset; the ternary output feature values ​​of the first, second, and third quality mapping functions are set as the update input data, and a fourth quality mapping function is configured based on the output supervision data and the update input data; the output layers of the first, second, and third quality mapping functions and the input layer of the fourth quality mapping function are fused to generate the quality mapping function.

[0059] Specifically, the support vector machine is trained using the output supervision data and the input dataset. The support vector machine learns the mapping relationship from the input dataset to the output supervision data and generates the first quality mapping function.

[0060] Then, the same output supervision data and input dataset are used to train the random forest model, and multiple decision trees are built and integrated through the random forest to generate a second quality mapping function.

[0061] Next, the BP neural network is trained using the output supervision data and the input dataset. The network weights are adjusted through backpropagation of the BP neural network to learn the mapping relationship from the input dataset to the output supervision data and generate the third quality mapping function.

[0062] Next, the output values ​​of the first, second, and third quality mapping functions, i.e., the ternary output feature values, are used as new input data, i.e., updated input data. A new model is trained using the updated input data and the original output supervision data, such as by using a backpropagation neural network or linear regression again, to obtain the fourth quality mapping function. This learns the mapping relationship from the ternary output feature values ​​of the three basic quality mapping functions to the true quality index.

[0063] Then, the output layers of the first, second, and third quality mapping functions are fused with the input layer of the fourth quality mapping function. This means that during prediction or evaluation, the basic quality mapping functions, namely support vector machines, random forests, and backpropagation neural networks, are used to obtain three predicted values. These three predicted values ​​are then used as the input of the fourth quality mapping function to obtain the final prediction and evaluation result, which is the generation of the quality mapping function.

[0064] Integration methods can improve the accuracy and robustness of predictions and assessments, as different base models may capture different features or patterns in the data.

[0065] Furthermore, this application also includes: When the predicted value of the surface quality index does not meet the expected value of the surface quality index, the first-level surface quality deviation index is equal to 1; when the predicted value of the surface quality index does not meet the expected value of the surface quality index, the surface quality deviation index is equal to 0; wherein, when the first-level surface quality deviation index is equal to 1, the Euclidean distance between the predicted value of the surface quality index and the expected value of the surface quality index is calculated and set as the surface quality deviation index.

[0066] Specifically, the predicted and expected values ​​of surface quality indicators are compared. When the predicted value of a surface quality indicator does not meet the expected value, the first-level surface quality deviation index equals 1, indicating a deviation between the predicted and expected values.

[0067] Then, when the predicted value of the surface quality index meets the expected value of the surface quality index, the surface quality deviation index is equal to 0, indicating that the predicted value of the surface quality index meets the expected value of the surface quality index.

[0068] Next, when the first-level surface quality deviation index is equal to 1, the absolute value of the difference between the predicted value and the expected value of the surface quality index is calculated to obtain the Euclidean distance between the predicted value and the expected value of the surface quality index, and the Euclidean distance is set as the surface quality deviation index.

[0069] By measuring the distance between the predicted and expected values ​​of surface quality indicators, the accuracy of control over the predicted surface quality indicators can be improved.

[0070] Furthermore, this application also includes: Configure a priority identifier table for any of the surface quality indicators; fit the surface quality deviation index according to the priority identifier table to identify the detection priority, and generate the detection priority label.

[0071] Specifically, the priority identifier table is a pre-configured data table. For example, different priority levels can be defined, including high, medium, and low, or represented by numbers 1, 2, and 3. A threshold for the corresponding surface quality deviation index is set for each priority level. For example, a high-priority deviation index is greater than or equal to 10%, a medium-priority deviation index is greater than 5% and less than 10%, and a low-priority deviation index is less than or equal to 5%, thus creating a priority identifier table. A priority identifier table can be configured for any surface quality indicator.

[0072] Then, the surface quality deviation index is compared with the threshold in the priority label table to determine its priority level. Based on the comparison result, a corresponding detection priority label is generated for the surface quality index. For example, if the surface quality deviation index is 8%, it is classified as medium priority, and a corresponding detection priority label is generated.

[0073] The generated detection priority labels should be integrated with the actual detection process to ensure that high-priority surface quality indicators can be detected and processed in a timely and effective manner.

[0074] In summary, the surface quality testing method for wool-like staple fiber yarn provided in this application has the following technical advantages: By traversing surface quality indicators and performing correlation analysis with the production process variables of the imitation wool staple fiber yarn to be tested, a set of correlation variables is generated. Based on big data, a quality mapping function is constructed between the set of correlation variables and the surface quality indicators. The monitoring value set of correlation variables of the set of correlation variables of the imitation wool staple fiber yarn to be tested is retrieved, and surface quality is predicted through the quality mapping function to generate predicted values ​​of surface quality indicators. The predicted values ​​of surface quality indicators are compared with the expected values ​​of surface quality indicators to generate a surface quality deviation index. According to the surface quality deviation index, the surface quality indicators are identified by detection priority, and detection priority labels are generated. Targeted detection is carried out according to the detection priority labels to realize the correlation analysis between surface quality indicators and production process variable parameters, determine the technical objectives of key process parameters and their impact on surface quality, optimize the production process, and achieve the technical effect of improving product surface quality.

[0075] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not 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.

[0076] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for detecting the surface quality of wool-like staple fiber yarn, characterized in that, include: The surface quality indicators are traversed, and a correlation analysis is performed with the production process variables of the imitation wool staple fiber yarn to be tested to generate a set of correlation variables; Based on big data, a quality mapping function is constructed between the set of related variables and the surface quality indicators; The associated variable monitoring value set of the associated variable set of the wool-like short fiber yarn to be detected is retrieved, and the surface quality is predicted by the quality mapping function to generate the predicted value of the surface quality index. By comparing the predicted values ​​of the surface quality indicators with the expected values ​​of the surface quality indicators, a surface quality deviation index is generated. Based on the surface quality deviation index, the surface quality indicators are identified by detection priority, and detection priority labels are generated. Targeted detection is performed based on the detection priority label.

2. The method as described in claim 1, characterized in that, A correlation analysis was performed on the surface quality indicators and their correlation with the production process variables of the imitation wool staple fiber yarn to be tested, generating a set of correlation variables, including: A first surface quality index is obtained from the surface quality index, wherein the first surface quality index is any one of the following: hair amount, curl degree, gloss, elongation, and pilling amount over a preset time. Obtain the set of production process variables for the wool-like short fiber yarn to be tested; Based on the first surface quality index and the set of production process variables, collect the production log of the imitation wool staple fiber yarn, wherein the production log of the imitation wool staple fiber yarn includes the recorded values ​​of the production process variables and the recorded values ​​of the first surface quality index. A correlation analysis is performed between the recorded values ​​of the first surface quality index and the recorded values ​​of the production process variables to generate a correlation set; Add the production process variables whose correlation set is greater than or equal to the correlation threshold to the first correlation variable set of the first surface quality index; Add the first set of related variables to the set of related variables.

3. The method as described in claim 2, characterized in that, A correlation analysis is performed between the recorded values ​​of the first surface quality index and the recorded values ​​of the production process variables to generate a correlation set, including: The first surface quality index record value and the production process variable record value are reconstructed from nearest neighbor to generate the first surface quality index feature value and the production process variable feature value; Based on the characteristic values ​​of the first surface quality index, a correlation analysis benchmark sequence is constructed; Based on the characteristic values ​​of the production process variables, a correlation analysis comparison sequence is constructed; Grey relational analysis is performed on the baseline sequence of the correlation analysis and the comparison sequence of the correlation analysis to generate the correlation set.

4. The method as described in claim 3, characterized in that, The first surface quality index record value and the production process variable record value are reconstructed from nearest neighbor to generate the first surface quality index feature value and the production process variable feature value, including: Obtain the first production process variable record value of the production process variable record value, wherein the first production process variable record value has a unique first joint surface quality index record value corresponding to the first surface quality index record value. Obtain a second production process variable record value for the production process variable record value, wherein the second production process variable record value has a unique second combined surface quality index record value corresponding to the first surface quality index record value; Calculate the variable deviation modulus set of the first production process variable record value and the second production process variable record value; The percentage of variables whose deviation modulus set does not meet the variable attribute deviation threshold set is counted, and a first deviation coefficient is generated. The square root of the sum of the squares of the variable deviation modulus set is performed to generate the second deviation coefficient; When the first deviation coefficient is less than or equal to the first deviation coefficient threshold, and the second deviation coefficient is less than or equal to the second deviation coefficient threshold, either the first production process variable record value or the second production process variable record value is stored as the production process variable feature value of the first combined surface quality index record value and the second combined surface quality index record value. The analysis involves iteratively traversing the data. When any two production process variable feature values ​​cannot be updated in a neighboring manner, a central tendency analysis is performed on multiple joint surface quality index records for each production process variable feature value to generate the first surface quality index feature value.

5. The method as described in claim 1, characterized in that, Based on big data, a quality mapping function is constructed between the set of related variables and the surface quality indicators, including: Obtain the correlation set of the first surface quality index of the surface quality index, and configure the correlation variable weight set; Based on the set of related variables and the surface quality index, collect historical production data of the same model of the wool-like short fiber yarn to be tested; The historical records of the related variables of the same model's historical production data are preprocessed using the weight set of the related variables to generate the input dataset; Set the historical value of the first surface quality index of the same model's historical production data as the output monitoring data; Configure the quality mapping function based on the output supervision data and the input dataset.

6. The method as described in claim 5, characterized in that, Configure the quality mapping function based on the output supervision data and the input dataset, including: Configure a support vector machine based on the output supervision data and the input dataset, and generate a first quality mapping function; Based on the output supervision data and the input dataset, a random forest is configured to generate a second quality mapping function; Configure a BP neural network based on the output supervision data and the input dataset to generate a third quality mapping function; Set the ternary output feature values ​​of the first quality mapping function, the second quality mapping function, and the third quality mapping function as the update input data, and configure the fourth quality mapping function according to the output supervision data and the update input data; The output layers of the first, second, and third quality mapping functions are merged with the input layer of the fourth quality mapping function to generate the quality mapping function.

7. The method as described in claim 1, characterized in that, By comparing the predicted values ​​and expected values ​​of the surface quality indicators, a surface quality deviation index is generated, including: When the predicted value of the surface quality index does not meet the expected value of the surface quality index, the first-level surface quality deviation index is equal to 1. When the predicted value of the surface quality index meets the expected value of the surface quality index, the surface quality deviation index is equal to 0. When the first-level surface quality deviation index is equal to 1, the Euclidean distance between the predicted value of the surface quality index and the expected value of the surface quality index is calculated and set as the surface quality deviation index.

8. The method as described in claim 1, characterized in that, Based on the surface quality deviation index, the surface quality indicators are prioritized for detection, and detection priority labels are generated, including: Configure a priority identifier table for any of the aforementioned surface quality indicators; The detection priority is identified by fitting the surface quality deviation index to the priority identification table, and the detection priority label is generated.

Images