Production quality evaluation method for polyester fiber preparation

Abstract

The invention provides a production quality evaluation method for polyester fiber preparation, and relates to the technical field of quality evaluation, and the method comprises the steps: carrying out the quality influence analysis through a plurality of production process parameters, and generating a plurality of process influence characteristics; carrying out production correlation analysis according to the plurality of real-time production monitoring data and the plurality of process influence characteristics, and constructing a production process influence data network; carrying out production correlation analysis according to the plurality of real-time production monitoring data and the production environment parameter set, and constructing a production environment influence data network; and performing multiple regression analysis on the two influence data networks, constructing a plurality of quality indexes to perform stage quality fluctuation evaluation, generating a plurality of stage quality scores, and performing quality warning on preparation of polyester fibers. The technical problem that the real-time performance and the accuracy of production quality evaluation are limited due to the lack of a comprehensive consideration mechanism for production process and environmental factors in the prior art is solved, and the accuracy, the comprehensiveness and the real-time performance of polyester fiber production quality evaluation are improved.

Description

Technical Field

[0001] This application relates to the technical field of quality assessment, and particularly to a production quality assessment method for polyester fiber preparation. Background Art

[0002] As an important synthetic fiber, polyester fiber is widely used in industries such as textiles, clothing, and home furnishings. In the existing production process of polyester fiber, the production quality assessment method mainly focuses on conventional indicators such as fiber strength and elongation, and most of them are centrally detected after the production process ends. This method focuses on static detection and conventional indicator monitoring, and fails to fully consider the changes in environmental conditions, fluctuations in process parameters, and the complex relationships between these factors during the production process. At the same time, there is a lack of real-time monitoring and stage assessment mechanisms, which cannot capture the quality fluctuations occurring during the production process in real time, and it is difficult to identify potential quality problems in the production process in a timely manner. This may lead to the discovery of problems only after a large number of unqualified products are produced, thus delaying the best opportunity for production adjustment. Summary of the Invention

[0003] This application provides a production quality assessment method for polyester fiber preparation, which solves the technical problem in the prior art that due to the lack of a comprehensive consideration mechanism for production processes and environmental factors, the production quality of polyester fiber cannot be comprehensively evaluated and the quality fluctuations during the production process cannot be warned in real time, and achieves the technical effect of improving the accuracy, comprehensiveness, and real-time nature of the production quality assessment of polyester fiber, thereby enhancing the production quality stability of polyester fiber.

[0004] In view of the above problems, this application provides a production quality assessment method for polyester fiber preparation, and the method includes: traversing and collecting based on the production process information of polyester fiber to obtain a plurality of real-time production monitoring data, and extracting a plurality of production process parameters according to the production process information of polyester fiber; performing quality impact analysis on the preparation of polyester fiber based on the plurality of production process parameters to generate a plurality of process impact characteristics; performing production correlation analysis according to the plurality of real-time production monitoring data combined with the plurality of process impact characteristics according to the production process information to construct a production process impact data network; performing production correlation analysis according to the plurality of real-time production monitoring data combined with the production environment parameter set according to the production process information to construct a production environment impact data network; performing multiple regression analysis on the production process impact data network and the production environment impact data network to construct a plurality of quality indicators, performing stage quality fluctuation assessment on the plurality of real-time production monitoring data according to the plurality of quality indicators to generate a plurality of stage quality scores; performing quality warning on the preparation of polyester fiber based on the plurality of stage quality scores to generate a production warning instruction, and optimizing the production control of polyester fiber preparation through the production warning instruction.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages: By traversing the production process information of polyester fibers, multiple real-time production monitoring data are collected and multiple production process parameters are extracted, providing comprehensive raw data for subsequent analysis. Based on the multiple production process parameters, an analysis of the quality impact on the preparation of polyester fibers is carried out, generating multiple process impact characteristics, finding out the relationship pattern between process parameters and product quality, and further clarifying the key directions of quality assessment and control. According to the multiple real-time production monitoring data combined with the multiple process impact characteristics, production correlation analysis is carried out according to the production process information to construct a production process impact data network; intuitively and systematically presenting the mutual relationship between process factors and actual production conditions in the production process, providing a networked analysis framework for comprehensive quality assessment. According to the multiple real-time production monitoring data combined with the production environment parameter set, production correlation analysis is carried out according to the production process information to construct a production environment impact data network, revealing the comprehensive impact of the interaction between production processes and the production environment on quality. A multiple regression analysis is performed on the production process impact data network and the production environment impact data network to construct multiple quality indicators, making quality assessment more scientific, comprehensive, and accurate. According to the multiple quality indicators, a stage quality fluctuation assessment is carried out on the multiple real-time production monitoring data to generate multiple stage quality scores, which can quantitatively evaluate quality fluctuations from multiple dimensions. Through these stage quality scores, the quality fluctuation characteristics of different production stages are revealed, helping to identify potential quality problems in each stage of production. Based on the multiple stage quality scores, a quality warning is issued for the preparation of polyester fibers to generate a production warning instruction, timely identifying abnormal fluctuations in the production process, and optimizing the production control of polyester fiber preparation through the production warning instruction to avoid the production of unqualified products.

[0006] In summary, this application starts from basic data collection and gradually analyzes the impact of production processes and environmental factors on the quality of polyester fiber preparation. By constructing data networks, performing multiple regression analysis to construct quality indicators, and conducting stage quality assessment, quality warning and production control optimization are achieved. This solution comprehensively considers process and environmental factors in the production process, realizes timely, accurate, and comprehensive assessment of the production quality of polyester fibers, can timely detect quality fluctuations in each stage of production and make adjustments, thereby effectively improving the production quality stability of polyester fibers and optimizing the production process.

[0007] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the description. And in order to make the above and other purposes, features, and advantages of this application more obvious and understandable, the following specifically exemplifies the specific implementation manners of this application. Brief Description of the Drawings

[0008] Figure 1 Schematic flow chart of the production quality assessment method for polyester fiber preparation provided by the embodiment of the present application.

[0009] Figure 2 Schematic flow chart of constructing a production process impact data network in the production quality assessment method for polyester fiber preparation provided by the embodiment of the present application.

[0010] Figure 3 Schematic flow chart of constructing a production environment impact data network in the production quality assessment method for polyester fiber preparation provided by the embodiment of the present application. Detailed implementation manners

[0011] The embodiment of the present application provides a production quality assessment method for polyester fiber preparation. By traversing and collecting the production process information of polyester fibers, multi-dimensional data analysis is carried out in combination with production process parameters and production environment parameters, and a production process impact data network and a production environment impact data network are constructed. The production process impact data network and the production environment impact data network are combined for multiple regression analysis, thereby constructing multiple quality indicators, evaluating the stage quality fluctuations, and generating multiple stage quality scores. Quality warnings are given to the polyester fiber preparation process through the stage quality scores, and production warning instructions are generated. The embodiment of the present application solves the technical problems in the prior art that due to the lack of a comprehensive consideration mechanism for production processes and environmental factors, the production quality of polyester fibers cannot be comprehensively evaluated and the quality fluctuations in the production process cannot be warned in real time, and achieves the technical effects of improving the accuracy, comprehensiveness and real-time nature of the production quality assessment of polyester fibers, thereby enhancing the production quality stability of polyester fibers.

[0012] As Figure 1 shown, the embodiment of the present application provides a production quality assessment method for polyester fiber preparation, and the method includes: Step S1: Traverse and collect based on the production process information of polyester fibers to obtain a plurality of real-time production monitoring data, and extract a plurality of production process parameters according to the production process information of polyester fibers.

[0013] Specifically, production process information refers to the detailed data of each production stage from raw materials to the final product in the production of polyester fibers, including polymerization, melt spinning, drawing, etc. Each stage has different process parameters and production control standards. Real-time production monitoring data are the data collected in real time through various sensors or monitoring devices during the production of polyester fibers, such as temperature, pressure, flow rate, rotational speed, etc. These data can reflect the actual state during the production process. Production process parameters refer to various parameters set according to the production process requirements during the production of polyester fibers, such as spinning speed, draw ratio, heat setting temperature, etc. These parameters are the key factors to ensure product quality, and different process parameters will affect the performance of polyester fibers. For example, too fast spinning speed may lead to a decrease in fiber strength.

[0014] Obtain the complete production process information of polyester fibers from the production management system, and traverse and collect the production process through various sensors (such as temperature sensors, pressure sensors, speed sensors, etc.) installed on the production equipment to obtain real-time production monitoring data such as temperature, pressure, and speed. At the same time, based on the production process information, extract production process parameters such as spinning speed and draw ratio. For example, in the melt spinning stage, it is necessary to collect in real time parameters such as melt temperature, spinneret temperature, and spinning speed. At the same time, extract the set melt spinning process parameters, such as spinning temperature and spinning speed. By capturing real-time production monitoring data and extracting production process parameters, it provides raw data support for subsequent quality analysis and ensures the comprehensiveness and timeliness of the data.

[0015] Step S2: Analyze the quality impact on the preparation of polyester fibers based on the multiple production process parameters to generate multiple process impact characteristics.

[0016] Specifically, process impact characteristics are the characteristics extracted from production process parameters that can significantly affect fiber quality. Such as temperature fluctuations, pressure fluctuations, etc.

[0017] Based on the obtained multiple production process parameters, use quality analysis methods to analyze the quality impact on the preparation of polyester fibers. Through statistical analysis methods, identify the relationship patterns between different process parameters (such as temperature, pressure, drawing speed, etc.) and product quality, and generate multiple process impact characteristics. For example, if the temperature is too high, it may lead to a decrease in fiber strength; if the drawing speed is inconsistent, it may lead to poor fiber uniformity. Through quality impact analysis, it is possible to determine the impact of different production process parameters on quality during the production of polyester fibers, providing a basis for subsequent quality prediction and adjustment.

[0018] Step S3: Perform production correlation analysis according to the multiple real-time production monitoring data in combination with the multiple process impact characteristics according to the production process information to construct a production process impact data network.

[0019] Specifically, the production process impact data network is a data network structure constructed by production correlation analysis, which combines real-time production monitoring data and process impact characteristics according to production process information. This network can visually display the relationship between process factors and actual production data during the production process.

[0020] According to production process information, production correlation analysis is performed on multiple real-time production monitoring data and multiple process impact characteristics using data mining algorithms (such as association rule mining algorithms) to determine the impact relationships between different real-time production monitoring data. A production process impact data network is constructed through production correlation analysis. The nodes in this data network can be different process parameters, monitoring data, etc., and the edges represent the impact relationships between the nodes. The production process impact data network reflects the impact of production process parameters on fiber quality and provides a basis for subsequent quality prediction and adjustment.

[0021] Step S4: Perform production correlation analysis according to the multiple real-time production monitoring data combined with the production environment parameter set according to the production process information to construct a production environment impact data network.

[0022] Specifically, the production environment parameter set refers to various parameters in the polyester fiber production environment, such as the temperature, humidity, and air cleanliness in the workshop. These environmental factors will affect the production quality of polyester fibers. For example, too high humidity may affect the drying degree of the fibers and thus affect their quality. The production environment impact data network is a data network structure constructed by production correlation analysis, which combines real-time production monitoring data and the production environment parameter set according to production process information and is used to display the relationship between environmental factors and actual production data.

[0023] Obtain the production environment parameter set, combine multiple real-time production monitoring data with the production environment parameter set (which can be obtained by installing environmental sensors such as temperature and humidity sensors in the workshop), and perform production correlation analysis according to production process information. For example, analyze the impact relationship between the workshop humidity and the water content of the fibers during the spinning process. Construct a production environment impact data network according to the analysis results. This network shows the correlation relationship between production environment parameters and real-time production monitoring data and reveals the impact of environmental factors on production quality during the production process.

[0024] Step S5: Perform multiple regression analysis on the production process impact data network and the production environment impact data network to construct multiple quality indicators, and evaluate the stage quality fluctuations of the multiple real-time production monitoring data according to the multiple quality indicators to generate multiple stage quality scores.

[0025] Specifically, quality indicators are quantitative indicators used to measure the quality of polyester fibers, such as fiber strength, elongation at break, fineness, etc. These indicators can objectively reflect the quality status of the product. The stage quality score is a quantitative score obtained after quality assessment of each production stage, which can intuitively reflect the quality of the product in that stage.

[0026] Using the multiple regression analysis method, various factors in the production process impact data network and the production environment impact data network are used as independent variables to analyze the quality indicators of polyester fibers (such as strength, elongation at break, etc.). For example, factors such as spinning speed, workshop humidity, spinning temperature, etc. are used as independent variables, and fiber strength is used as the dependent variable to establish a regression equation. Through this equation, the influence coefficient of each factor on the quality indicator is calculated, thereby constructing multiple quality indicators. Then, based on these quality indicators, stage quality fluctuation assessment is carried out on the real-time production monitoring data. By comparing the quality indicator data of different stages, it is determined whether the quality is stable, the quality score of each stage is calculated, and multiple stage quality scores corresponding to multiple production stages are obtained.

[0027] This step constructs multiple quality indicators through multivariate analysis, accurately quantifies the quality fluctuations in the complex production process, and provides a scientific basis for quality early warning and optimization control.

[0028] Step S6: Based on the multiple stage quality scores, conduct quality warning for the preparation of polyester fibers, generate a production warning instruction, and optimize the production control of polyester fiber preparation through the production warning instruction.

[0029] Specifically, the production warning instruction is a warning message issued when the stage quality score does not meet the requirements or shows abnormal fluctuations, and is used to guide the direction of production control adjustment in the process of polyester fiber preparation.

[0030] According to the multiple stage quality scores obtained in step S5, set the threshold for quality warning. When a certain stage quality score is lower than or higher than this threshold, it is determined that a quality problem has occurred, and thus a quality warning is issued to generate a production warning instruction. This instruction clearly indicates which links in the production process need to be adjusted, such as adjusting the spinning temperature, changing the draw ratio, etc. By executing these production warning instructions, the production process of polyester fiber preparation is optimized.

[0031] Furthermore, step S2 of the embodiment of the present application further includes: Step S21: Based on the multiple production process parameters, conduct multi-dimensional quality impact analysis on the preparation of polyester fibers to generate the quality influence of the multiple production process parameters.

[0032] Step S22: Sort the quality influence of the multiple production process parameters in descending order to generate a quality influence sequence.

[0033] Step S23: performing feature mining analysis on the multiple production process parameters according to the quality impact sequence to determine the multiple process impact features.

[0034] Specifically, the quality influence is a quantitative value that measures the degree of influence of each production process parameter on the quality of polyester fiber preparation. The quality influence sequence is a sequence obtained by arranging multiple production process parameters from large to small (in descending order) according to their quality influence.

[0035] For each production process parameter, the preparation quality of polyester fiber is analyzed from multiple dimensions to generate the quality influence of multiple production process parameters. For example, for the production process parameter of spinning speed, the analysis is performed from multiple quality dimensions such as fiber strength, fineness, elasticity, etc. This can be achieved through experimental design, data collection and analysis. For example, different spinning speeds are set in the laboratory to produce multiple groups of polyester fiber samples, and then the quality indicators such as strength, fineness, elasticity, etc. of each group of samples are measured and analyzed to determine the degree of influence of spinning speed on these quality indicators, and then the quality influence of spinning speed as a production process parameter is generated. The quality influences of the multiple production process parameters obtained are compared and arranged in order from large to small (descending order) to generate a quality influence sequence. For example, if the quality influence of spinning speed is 0.8, the quality influence of stretching multiple is 0.6, and the quality influence of heat setting temperature is 0.5, then the quality influence sequence is spinning speed, stretching multiple, and heat setting temperature. Through multidimensional quality impact analysis, the influence of each production process parameter on the quality of polyester fiber can be fully quantified, and key process parameters can be identified through quality influence sorting. This provides a scientific basis for subsequent production quality optimization.

[0036] Based on the quality influence sequence, feature mining analysis is performed on multiple production process parameters. Starting from the production process parameter with the greatest impact on quality, the relationship between it and the quality of polyester fiber is deeply analyzed, and the features that can represent this relationship, namely the process influence features, are mined. For example, for the spinning speed that ranks first in the quality influence sequence, the microstructure, molecular arrangement and other characteristics of the fiber at different spinning speeds are analyzed to determine the process influence features related to fiber quality, such as the relationship characteristics between spinning speed and fiber crystallinity. Feature mining analysis further reveals the specific impact mechanism of each key process parameter on quality, such as fluctuations in melting temperature will cause fluctuations in the breaking strength of the fiber, changes in stretching temperature mainly affect the ductility of the fiber, and the stretching speed has a relatively stable impact on the fineness of the fiber. According to these process influence features, key parameters can be adjusted and controlled more accurately during the production process to ensure the stability and consistency of polyester fiber quality.

[0037] Further, such as Figure 2As shown, step S3 of the embodiment of the present application further includes: Step S31: Match the multiple real-time production monitoring data with the multiple process influence characteristics to generate a production influence matching result.

[0038] Step S32: Perform influence identification on the production influence matching result according to the production process information to obtain multiple production influence variables.

[0039] Step S33: Use the multiple production influence variables as nodes, and connect the multiple nodes according to the production influence matching result to construct the production process influence data network.

[0040] Specifically, obtain the multiple real-time production monitoring data in step S1 and the multiple process influence characteristics generated in step S2. For each real-time production monitoring data, find the corresponding or relevant characteristics in the multiple process influence characteristics for matching. For example, the spinning temperature data in the real-time production monitoring data is matched with the characteristic regarding the influence of spinning temperature on fiber quality in the process influence characteristics. If the real-time spinning temperature is within the ideal temperature range set by the process influence characteristics, the matching result may be "good match"; otherwise, the matching result may be "poor match". By matching all the real-time production monitoring data and process influence characteristics, a production influence matching result is generated. This production influence matching result reflects the matching degree between the real-time monitored data and the previously analyzed process influence characteristics in actual production.

[0041] According to the sequence and production characteristics of each link in the production process information, analyze the influence of each matching result on the production process, so as to determine the factors that affect the quality of the final product in the production process, that is, the production influence variables. These production influence variables can be the operating parameters of a certain device, the environmental data of a certain production link, etc. For example, in the spinning process, according to the production process information, it is determined that factors such as spinning temperature and speed have an important impact on the final fiber quality. If the production influence matching result shows that the matching of the spinning temperature is poor, then the spinning temperature is identified as a production influence variable. Perform influence identification on all the production influence matching results to obtain multiple production influence variables. These production influence variables are the key variables that affect production efficiency, quality, and cost in the actual production process of polyester fibers.

[0042] Take multiple production impact variables as nodes, with each node representing the data of a production impact variable in the production process. Then, determine the association relationships between the nodes according to the production impact matching results and perform association connections. For example, if the production impact matching results show that there is an interaction between the spinning temperature and the spinning pressure (such as an increase in temperature will cause a change in pressure), then establish an association connection between the nodes representing the spinning temperature and the spinning pressure, forming an edge connecting the nodes. These edges represent the relationships or degrees of influence between the nodes, which can be unidirectional or bidirectional, indicating the influence or dependence of one node on another. By establishing the association connections between each node and other nodes, construct the entire production process impact data network. This data network reveals the complex interactions of various variables in the production process by visualizing the relationships between each node (production impact variable).

[0043] Furthermore, the construction method of the production environment parameter set described in step S4 of the embodiment of the present application includes: Step S4-1: Perform a correlation analysis on the production environment based on the production process information of polyester fibers in combination with the multiple production process parameters to generate multiple production environment correlation coefficients.

[0044] Step S4-2: Sort and align the multiple production environment correlation coefficients according to the production time sequence to generate a production environment correlation sequence.

[0045] Step S4-3: Perform an impact analysis based on the production environment correlation sequence in combination with the multiple production process parameters to set target environment impact parameters.

[0046] Step S4-4: Perform a sensing feedback on the production environment for the preparation of polyester fibers according to the target environment impact parameters to construct the production environment parameter set.

[0047] Specifically, the production environment correlation coefficient is a quantitative index used to represent the degree of correlation between the production environment and the production process parameters of polyester fibers. The production environment correlation sequence is a sequence obtained by sorting and aligning multiple production environment correlation coefficients according to the production time sequence (the sequence of precedence in the production process). For example, in the production process of polyester fibers, raw material preparation is carried out first, and then spinning and other processes are carried out. According to this time sequence, the production environment correlation coefficients corresponding to different processes are arranged to form a production environment correlation sequence. The target environment impact parameters refer to the environmental factors that are determined to have a greater impact on the quality of polyester fibers and need to be key controlled after analyzing the production environment correlation sequence. These parameters can be temperature, humidity, etc., and need to be particularly concerned about and optimized during the production process.

[0048] According to the production process information of polyester fibers, for each production link, find the correlation between production environment factors (such as humidity, temperature, air velocity, etc.) and production process parameters (such as melting temperature, drawing speed, etc.). Statistical methods (such as Pearson correlation coefficient) can be used to quantify the relationship between these environmental factors and process parameters, and calculate the production environment correlation coefficients between each production environment factor and production process information and production process parameters. For example, for the spinning process in the production process, analyze the correlation between the workshop temperature and the spinning speed.

[0049] According to the production time sequence, sort multiple production environment correlation coefficients. For example, in the production of polyester fibers, if the raw material preparation process precedes the spinning process, then align and sort the production environment correlation coefficients corresponding to the raw material preparation process and the production environment correlation coefficients corresponding to the spinning process in this order to form a production environment correlation sequence.

[0050] Since production environment data includes various data types, such as temperature, humidity, wind speed, gas concentration, dust content, etc. The impact degree of each type of data on production quality is different. In order to save computing resources and data collection costs, it is necessary to select the data types that have a significant impact on production quality from various production environment data for collection.

[0051] Based on the obtained production environment correlation sequence, conduct an impact analysis in combination with multiple production process parameters. Analyze the impact degree of the production environment factors corresponding to each correlation coefficient in the production environment correlation sequence on production process parameters and the quality of the final product. According to the impact degree of each environmental factor on production quality, select the production environment factors with a greater impact degree in descending order of impact degree and set them as target environmental impact parameters to ensure that environmental factors with a significant impact on the quality of polyester fibers are given priority attention. A threshold of impact degree can be preset during the collection process, and the production environment factors with an impact degree greater than or equal to this impact degree threshold are extracted and set as target environmental impact parameters. For example, if in the production environment correlation sequence, the humidity correlation coefficient corresponding to the spinning process indicates that humidity has a greater impact on spinning quality, and it is found through analysis in combination with production process parameters that humidity changes will significantly affect the strength of the fiber, then set humidity as the target environmental impact parameter.

[0052] According to the set target environmental impact parameters, install corresponding sensors in the production environment of polyester fiber preparation. Through these sensors, monitor the production environment in real time and transmit the monitoring data back to construct a production environment parameter set. This production environment parameter set covers all environmental factors that have a significant impact on quality, such as humidity, temperature, etc. For example, if the target environmental impact parameters for the spinning link include temperature and humidity, then install temperature and humidity sensors in the spinning workshop, collect temperature and humidity data in real time and transmit them back as part of the production environment parameter set.

[0053] Through the correlation analysis and ranking of production environment parameters, the above steps can effectively identify environmental factors that have a significant impact on the quality fluctuations in the production process. By setting target environmental impact parameters and performing sensing feedback, the changes in the production environment can be monitored in real time, thereby providing accurate data for evaluating the impact of production environment factors on production quality, helping to reduce the impact of environmental fluctuations on product quality, and improving the quality stability of polyester fibers.

[0054] Furthermore, as Figure 3 shown, step S4 of the embodiment of the present application further includes: Step S41: Determine multiple production stages according to the production process information, and perform correlation analysis based on the multiple production stages to formulate association rules.

[0055] Step S42: Perform data correlation analysis on the multiple real-time production monitoring data and the production environment parameter set according to the association rules to generate multiple data correlation impact coefficients.

[0056] Step S43: Perform impact identification based on the multiple data correlation impact coefficients to construct multiple impact factors.

[0057] Step S44: Perform network modeling according to the multiple impact factors in combination with the multiple data correlation impact coefficients to construct the production environment impact data network.

[0058] Specifically, the association rules are rules formulated based on the association relationships between multiple production stages determined according to the production process information. For example, a certain parameter in the raw material preparation stage (such as raw material purity) has a specific relationship with the product quality in the spinning stage, and this specific relationship can be represented by the association rules.

[0059] Determine multiple production stages according to the production process information of polyester fibers. Each production stage involves different process operations and controls. For example, the melt spinning stage involves temperature control, and the stretching stage involves the control of stretching speed and temperature. According to the production characteristics of these production stages, perform correlation analysis to reveal the mutual relationships between different stages and between different stages and environmental parameters, and formulate association rules. In the production of polyester fibers, the association rules can help identify the relationships between different production stages and production environment parameters. For example, if the humidity and temperature in a certain stage reach specific values, it may have an impact on the stretching rate or strength in the next stage.

[0060] According to the established association rules, multiple real-time production monitoring data are associated and analyzed with the production environment parameter set. At this time, the relationship between the real-time production monitoring data (such as temperature, pressure, stretching rate, etc.) and the production environment parameters (such as humidity, air flow rate, etc.) will be quantified into multiple data association influence coefficients through data association analysis. These data association influence coefficients are a quantitative index used to represent the degree of influence of the data association between multiple real-time production monitoring data and the production environment parameter set. The correlation analysis function in statistical analysis software can be used to calculate the data association influence coefficients. Input the data of the real-time production monitoring data and the production environment parameter set, and calculate the Pearson correlation coefficient between the two as the data association influence coefficient through the correlation analysis module.

[0061] Influence identification is carried out based on multiple data association influence coefficients. For example, if a certain data association influence coefficient indicates that the association between a certain production monitoring data and the environmental parameter has a greater impact on product quality, then the relevant factors in this association relationship are constructed as influence factors. By analyzing all the data association influence coefficients, multiple influence factors are constructed. Exemplarily, an influence threshold can be set. When the data association influence coefficient exceeds this threshold, the corresponding factor is constructed as an influence factor. These influence factors are various environmental factors that have an important impact on the final quality during the production process of polyester fibers.

[0062] By performing network modeling on the influence factors and their data association influence coefficients, a production environment influence data network is constructed. In this network, each influence factor serves as a node, and the nodes are connected through the data association influence coefficients. Through these nodes and connections, a production environment influence data network is formed, clearly showing the interaction between the production environment and process parameters. This data network can reflect the influence paths of different factors on quality, thus providing strong support for the optimization of the production process. The construction of this production environment influence data network and the identification of influence factors enable each key environmental factor in the production process to be accurately monitored, thereby effectively improving the accuracy and real-time nature of polyester fiber quality assessment.

[0063] Furthermore, step S5 of the embodiment of the present application further includes: Step S51: Perform multiple linear regression on the multiple production process parameters and the multiple process influence characteristics based on the production process influence data network to generate a first influence variable set, and the first influence variable set includes a first independent variable and a first dependent variable.

[0064] Step S52: Perform multiple linear regression on the multiple real-time production monitoring data and the production environment parameter set based on the production environment influence data network to generate a second influence variable set, and the second influence variable set includes a second independent variable and a second dependent variable.

[0065] Step S53: Perform weighted summation on the first independent variable and the second independent variable according to the first dependent variable and the second dependent variable to generate a weighted calculation result.

[0066] Step S54: Perform comprehensive analysis on the first independent variable and the second independent variable according to the weighted calculation result to construct the multiple quality indicators.

[0067] Specifically, based on the production process impact data network, perform multiple linear regression on multiple production process parameters and multiple process impact characteristics. Multiple linear regression can analyze the relationship between process parameters and process impact characteristics. The result of the regression analysis will generate a first set of impact variables, which includes a first independent variable (production process parameters and process impact characteristics) and a first dependent variable (quality impact variables related to the process, such as certain quality characteristics of fibers).

[0068] Similarly, based on the production environment impact data network, perform multiple linear regression on multiple real-time production monitoring data and the production environment parameter set. Through regression analysis, it can be found out how environmental factors affect the real-time monitoring data during the production process. The regression result generates a second set of impact variables, including a second independent variable (real-time production monitoring data and the production environment parameter set) and a second dependent variable (quality impact variables related to the environment).

[0069] Determine the weights according to the first dependent variable and the second dependent variable, and perform weighted summation on the first independent variable and the second independent variable. Perform comprehensive analysis on the weighted calculation result. Analyze the comprehensive impact of the weighted first independent variable and second independent variable on the quality of polyester fibers. According to the analysis result, construct multiple quality indicators.

[0070] The above steps can generate corresponding quality indicators for each production stage and environmental condition through multiple linear regression analysis and weighted calculation. These quality indicators provide evaluation criteria for the monitoring and optimization of quality fluctuations during the production process, so as to be able to quantify the quality fluctuations during the production process and provide guidance for the optimization of production parameters.

[0071] Further, step S54 further includes: Step S541: Perform a strong correlation determination on the first independent variable and the second independent variable based on the weighted calculation result to generate a determination result.

[0072] Step S542: If the determination result is that there is a strong correlation between the first independent variable and the second independent variable, generate a dimensionality reduction instruction.

[0073] Step S543: Reduce the dimension of the first independent variable and / or the second independent variable through the dimensionality reduction instruction to generate a first reduced-dimensional independent variable and / or a second reduced-dimensional independent variable.

[0074] Step S544: Conduct quality risk assessment based on the first dimension-reduced independent variable and / or the second dimension-reduced independent variable, and generate a first quality risk contribution value and / or a second quality risk contribution value. There is a corresponding relationship between the first quality risk contribution value and the first dimension-reduced independent variable, and there is a corresponding relationship between the second quality risk contribution value and the second dimension-reduced independent variable.

[0075] Step S545: Construct the multiple quality indicators according to the first quality risk contribution value and / or the second quality risk contribution value.

[0076] Specifically, perform a correlation analysis on the first independent variable and the second independent variable in the weighted calculation result to determine whether there is a strong correlation between the first independent variable and the second independent variable, and generate a judgment result. This strong correlation indicates that these independent variables have a synergistic effect or an interdependent relationship in the process of influencing the dependent variable (such as product quality-related indicators). Statistical methods, such as the Pearson correlation coefficient or regression analysis, can be used to quantify the correlation between these two variables. If the correlation coefficient of the two variables exceeds a certain threshold, it is considered that there is a strong correlation between them. Exemplarily, a strong correlation determination is performed by calculating the Pearson correlation coefficient. If the absolute value of the Pearson correlation coefficient is close to 1 (for example, greater than 0.8 or 0.9, and the specific threshold can be determined according to the actual situation), it is determined that there is a strong correlation.

[0077] If the judgment result indicates that there is a strong correlation between the first independent variable and the second independent variable, generate a dimension reduction instruction to merge or simplify these two variables into one variable. Through the dimension reduction operation, redundant features can be reduced, thereby simplifying the analysis process of quality assessment. For example, if the effects of temperature and humidity on fiber strength are almost overlapping, it can be considered to merge them into a new variable. According to the dimension reduction instruction, perform dimension reduction processing on the first independent variable and / or the second independent variable. Dimension reduction methods can adopt techniques such as principal component analysis (PCA) or factor analysis. Through these methods, variables with strong correlation are merged into new variables, which are called the first dimension-reduced independent variable and / or the second dimension-reduced independent variable.

[0078] Conduct quality risk assessment based on the first dimension-reduced independent variable and / or the second dimension-reduced independent variable. This can be achieved by establishing a risk assessment model. For example, use the first dimension-reduced independent variable and / or the second dimension-reduced independent variable as inputs, and the defect rate or quality fluctuation range of product quality, etc. as outputs, and perform model training through historical data. According to the relationship between the values of these dimension-reduced independent variables and product quality problems (such as the defective rate) in historical production data, calculate the contribution value of each first dimension-reduced independent variable to the quality risk, that is, the first quality risk contribution value. Similarly, calculate the contribution value of each second dimension-reduced independent variable to the quality risk, that is, the second quality risk contribution value.

[0079] Construct multiple quality indicators based on the first quality risk contribution value and / or the second quality risk contribution value. Exemplarily, a weighted average method can be used to construct quality indicators. Multiply the impact factor corresponding to different quality dimensions by the corresponding quality risk contribution value as the quality indicator for that quality dimension.

[0080] The above steps, combined with strong correlation determination and dimensionality reduction techniques, can effectively reduce redundant variables and improve the simplicity and accuracy of the quality assessment process. The independent variables after dimensionality reduction can better reveal the impact of key factors in the production process on quality, thereby achieving more accurate quality assessment and control. By constructing quality indicators based on risk contribution values, the sources of quality fluctuations can be identified more efficiently, and then targeted measures can be taken to optimize the production process and improve the product quality of polyester fibers.

[0081] Further, step S5 of the embodiment of the present application further includes: Step S55: Based on the multiple production stages, perform tolerance analysis on the multiple real-time production monitoring data according to the multiple quality indicators, and set the expected tolerance ranges for the multiple production stages.

[0082] Step S56: Slice the multiple real-time production monitoring data according to the multiple production stages to generate multi-stage real-time production monitoring data.

[0083] Step S57: Match and align the multi-stage real-time production monitoring data with the expected tolerance ranges of the multiple production stages, and determine whether the multi-stage real-time production monitoring data is within the expected tolerance ranges of the multiple production stages.

[0084] Step S58: If the multi-stage real-time production monitoring data is within the expected tolerance ranges of the multiple production stages, generate a data recording instruction, perform production recording on the multiple real-time production monitoring data through the data recording instruction to generate a normal production record log, and perform quality scoring according to the normal production record log to generate the quality scores for the multiple stages.

[0085] Step S59: If the multi-stage real-time production monitoring data is not within the expected tolerance ranges of the multiple production stages, generate a data analysis instruction, perform quality fluctuation analysis on the multi-stage real-time production monitoring data through the data analysis instruction, draw a quality fluctuation trend chart, and perform quality scoring according to the quality fluctuation trend chart to generate the quality scores for the multiple stages.

[0086] Specifically, different production stages have different quality indicators. Perform tolerance analysis on the multiple real-time production monitoring data for each production stage, and set the expected tolerance ranges for each production stage. These ranges represent the quality fluctuations allowed by the quality indicators in each stage.

[0087] Slice the multiple real-time production monitoring data according to different production stages, and divide them into multi-stage real-time production monitoring data corresponding to multiple production stages.

[0088] Match and align the multi-stage real-time production monitoring data with the expected tolerance ranges of multiple production stages. Determine the expected tolerance range corresponding to the real-time production monitoring data of each production stage. Then compare the real-time monitoring data with the set expected tolerance range to judge whether the multi-stage real-time production monitoring data is within the expected tolerance ranges of multiple production stages.

[0089] If the multi-stage real-time production monitoring data is within the expected tolerance ranges of multiple production stages, generate a data recording instruction. This instruction guides the production management system to record the data within the normal range. Generate a normal production record log. This log contains all production data within the expected tolerance range. According to the normal production record log, combined with the MES (Manufacturing Execution System) or SCADA system, automatically generate multiple stage quality scores corresponding to multiple production stages according to the preset quality standards, and quantitatively evaluate the product quality during the production process.

[0090] If the multi-stage real-time production monitoring data is not within the expected tolerance ranges of multiple production stages, generate a data analysis instruction. Execute this instruction to analyze the data beyond the tolerance range to determine the cause of quality fluctuations. By performing statistical analysis, trend analysis, etc. on the data beyond the tolerance range, determine the pattern and cause of quality fluctuations, and draw a quality fluctuation trend chart. This quality fluctuation trend chart can visually display the change trend of product quality over time or production stages, and identify the root cause of quality problems. According to the quality fluctuation trend chart, combined with the MES (Manufacturing Execution System) or SCADA system, automatically generate multiple stage quality scores according to the preset quality standards.

[0091] Through the above steps, it is possible to judge and record the quality status of production in real time. If the production data exceeds the tolerance range, quality fluctuation analysis will be initiated, and a quality fluctuation trend chart will be drawn, thereby providing more in-depth quality analysis, quantifying the degree of quality fluctuation, generating quality scores, and providing direction guidance for production process adjustment.

[0092] In summary, the production quality assessment method provided by the embodiment of the present application for polyester fiber preparation has the following technical effects: Through multi-dimensional data collection and analysis, the embodiments of this application comprehensively improve the quality control ability of polyester fiber production. First, by collecting production data in real time and extracting key process parameters, the comprehensiveness and timeliness of production quality data are ensured; secondly, by constructing a production process and environmental impact data network, the impacts of the process and environment on quality are comprehensively analyzed, and the quality fluctuations are quantified through multiple regression analysis, providing data support for quality early warning. Finally, the stage quality scoring and warning mechanism ensures the timely identification and intelligent control of quality fluctuations during the production process, avoiding the lag response of quality problems and resource waste. Overall, through real-time and accurate quality fluctuation monitoring and early warning, the embodiments of this application not only achieve timely, accurate, and comprehensive assessment of the production quality of polyester fibers, but also can timely detect quality fluctuations and make adjustments at all stages of production, improving the stability of production quality and providing decision-making support for the intelligent optimization of the production process.

[0093] The above description of the disclosed embodiments enables those skilled in the art to implement or use this application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A production quality assessment method for the preparation of polyester fibers, characterized in that, The method includes: Traversing and collecting based on the production process information of polyester fibers to obtain multiple real-time production monitoring data, and extracting multiple production process parameters according to the production process information of polyester fibers; Conducting a quality impact analysis on the preparation of polyester fibers based on the multiple production process parameters to generate multiple process impact characteristics; Conducting a production correlation analysis according to the multiple real-time production monitoring data in combination with the multiple process impact characteristics according to the production process information to construct a production process impact data network; Conducting a production correlation analysis according to the multiple real-time production monitoring data in combination with the production environment parameter set according to the production process information to construct a production environment impact data network; Performing a multiple regression analysis on the production process impact data network and the production environment impact data network to construct multiple quality indicators, and performing a stage quality fluctuation assessment on the multiple real-time production monitoring data according to the multiple quality indicators to generate multiple stage quality scores; Conducting a quality warning on the preparation of polyester fibers based on the multiple stage quality scores to generate a production warning instruction, and optimizing the production control of the preparation of polyester fibers through the production warning instruction.

2. The production quality evaluation method for polyester fiber preparation according to claim 1, characterized in that Conducting a quality impact analysis on the preparation of polyester fibers based on the multiple production process parameters to generate multiple process impact characteristics. The method includes: Conducting a multi-dimensional quality impact analysis on the preparation of polyester fibers based on the multiple production process parameters to generate the quality influence of the multiple production process parameters; Sorting the quality influence of the multiple production process parameters in descending order to generate a quality influence sequence; Conducting a feature mining analysis on the multiple production process parameters according to the quality influence sequence to determine the multiple process impact characteristics.

3. The production quality evaluation method for polyester fiber preparation according to claim 1, characterized in that, Conducting a production correlation analysis according to the multiple real-time production monitoring data in combination with the multiple process impact characteristics according to the production process information to construct a production process impact data network. The method includes: Matching the multiple real-time production monitoring data with the multiple process impact characteristics to generate a production impact matching result; Conducting an impact identification on the production impact matching result according to the production process information to obtain multiple production impact variables; Using the multiple production impact variables as nodes, and associating and connecting the multiple nodes according to the production impact matching result to construct the production process impact data network.

4. The production quality evaluation method for polyester fiber preparation according to claim 1, characterized in that, The construction method of the production environment parameter set includes: Conducting a correlation analysis on the production environment based on the production process information of polyester fibers in combination with the multiple production process parameters to generate multiple production environment correlation coefficients; Sorting and aligning the multiple production environment correlation coefficients according to the production time sequence to generate a production environment correlation sequence; Conducting an impact analysis based on the production environment correlation sequence in combination with the multiple production process parameters to set target environment impact parameters; Performing a sensing feedback on the production environment of the preparation of polyester fibers according to the target environment impact parameters to construct the production environment parameter set.

5. The production quality evaluation method for polyester fiber preparation according to claim 4, characterized in that Conducting a production correlation analysis according to the multiple real-time production monitoring data in combination with the production environment parameter set according to the production process information to construct a production environment impact data network. The method includes: Determine multiple production stages based on the production process information, conduct correlation analysis based on the multiple production stages, and formulate correlation rules; According to the correlation rules, perform data correlation analysis on the multiple real-time production monitoring data and the production environment parameter set to generate multiple data correlation influence coefficients; Based on the multiple data correlation influence coefficients, conduct influence identification to construct multiple influence factors; According to the multiple influence factors and in combination with the multiple data correlation influence coefficients, perform network modeling to construct the production environment impact data network.

6. The production quality evaluation method for polyester fiber preparation according to claim 1, characterized in that, Conduct multiple regression analysis on the production process impact data network and the production environment impact data network to construct multiple quality indicators. The method includes: Based on the production process impact data network, perform multiple linear regression on the multiple production process parameters and the multiple process impact characteristics to generate a first set of influence variables, where the first set of influence variables includes a first independent variable and a first dependent variable; Based on the production environment impact data network, perform multiple linear regression on the multiple real-time production monitoring data and the production environment parameter set to generate a second set of influence variables, where the second set of influence variables includes a second independent variable and a second dependent variable; Perform weighted summation on the first independent variable and the second independent variable according to the first dependent variable and the second dependent variable to generate a weighted calculation result; Based on the weighted calculation result, conduct comprehensive analysis on the first independent variable and the second independent variable to construct the multiple quality indicators.

7. The production quality evaluation method for polyester fiber preparation according to claim 6, characterized in that Based on the weighted calculation result, conduct comprehensive analysis on the first independent variable and the second independent variable to construct the multiple quality indicators. The method includes: Based on the weighted calculation result, conduct a strong correlation determination on the first independent variable and the second independent variable to generate a determination result; If the determination result is that there is a strong correlation between the first independent variable and the second independent variable, generate a dimensionality reduction instruction; Reduce the dimension of the first independent variable and / or the second independent variable through the dimensionality reduction instruction to generate a first reduced-dimensional independent variable and / or a second reduced-dimensional independent variable; Based on the first reduced-dimensional independent variable and / or the second reduced-dimensional independent variable, conduct quality risk assessment to generate a first quality risk contribution value and / or a second quality risk contribution value. The first quality risk contribution value has a corresponding relationship with the first reduced-dimensional independent variable, and the second quality risk contribution value has a corresponding relationship with the second reduced-dimensional independent variable; Construct the multiple quality indicators according to the first quality risk contribution value and / or the second quality risk contribution value.

8. The production quality evaluation method for polyester fiber preparation according to claim 5, wherein According to the multiple quality indicators, conduct stage quality fluctuation assessment on the multiple real-time production monitoring data to generate multiple stage quality scores. The method includes: Based on the multiple production stages, conduct tolerance analysis on the multiple real-time production monitoring data according to the multiple quality indicators to set the expected tolerance range for the multiple production stages; Slice the multiple real-time production monitoring data according to the multiple production stages to generate multi-stage real-time production monitoring data; Match and align the multi-stage real-time production monitoring data with the expected tolerance ranges of the multiple production stages, and determine whether the multi-stage real-time production monitoring data is within the expected tolerance ranges of the multiple production stages; If the multi-stage real-time production monitoring data is within the expected tolerance ranges of the multiple production stages, generate a data recording instruction, use the data recording instruction to perform production recording on the multiple real-time production monitoring data, generate a normal production record log, perform quality scoring based on the normal production record log, and generate the quality scores for the multiple stages; If the multi-stage real-time production monitoring data is not within the expected tolerance ranges of the multiple production stages, generate a data analysis instruction, use the data analysis instruction to perform quality fluctuation analysis on the multi-stage real-time production monitoring data, draw a quality fluctuation trend chart, perform quality scoring based on the quality fluctuation trend chart, and generate the quality scores for the multiple stages.

Images