Methods, apparatus, equipment and media for determining parameters affecting wire drying outlet temperature

Abstract

This invention discloses a method, apparatus, equipment, and medium for determining the influencing parameters of the drying outlet temperature. The method includes: performing parameter correlation analysis on the steady-state drying data set corresponding to the target production batch in the target yarn-making workshop; determining basic influencing parameters based on the correlation analysis results of each parameter; calculating the standard deviation of the outlet temperature in the steady-state drying data set based on a preset sliding window; dividing the steady-state drying data set into a drying outlet temperature fluctuation dataset based on the standard deviation calculation results; calculating the degree of influence of the basic influencing parameters on the global drying outlet temperature fluctuation dataset to determine candidate influencing parameters; dividing the steady-state drying data set based on the candidate influencing parameter data; predicting the outlet temperature of the data subset after data division based on a target machine learning model; and determining the target influencing parameter based on the prediction accuracy. Through the technical solution of this invention, key factors affecting the fluctuation of the drying outlet temperature can be identified more accurately and stably.

Description

Technical Field

[0001] This invention relates to the field of tobacco processing technology, and in particular to a method, apparatus, equipment, and medium for determining the parameters affecting the outlet temperature of tobacco drying. Background Technology

[0002] Cigarette manufacturing is a complex and precise industrial production process, in which quality control in the tobacco processing stage plays a decisive role in the quality of the final finished tobacco product. Among the many processes in tobacco processing, the drying process holds a crucial position. The core objective of drying is to ensure that the moisture content and temperature of the exported tobacco strictly meet the process specifications for finished tobacco products by precisely controlling the moisture content and temperature of the tobacco.

[0003] However, during the tobacco drying process, numerous factors influence the stability of the tobacco outlet temperature, often exhibiting significant fluctuations. Because these factors are numerous, intertwined, and closely related, accurately tracing the root cause of these temperature fluctuations is extremely difficult, posing a serious challenge to the stable control of the drying process and the improvement of the finished tobacco quality.

[0004] Therefore, how to accurately identify the key driving factors that cause fluctuations in the outlet temperature of the wire drying process from a large number of parameters, and lay the foundation for more accurate control and prediction of the stability of the outlet temperature of the wire drying process, is an urgent problem to be solved. Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, and medium for determining the parameters affecting the wire drying outlet temperature. It can solve the problem that existing technologies cannot quantitatively identify the specific driving factors affecting the fluctuation of the wire drying outlet temperature, resulting in difficulties in tracing the source of fluctuations and a lack of targets for process optimization.

[0006] According to one aspect of the present invention, a method for determining the influence parameters of the wire drying outlet temperature is provided, comprising: Obtain the set of steady-state drying data corresponding to the target production batch in the target yarn processing workshop; A parameter correlation analysis was performed on the steady-state drying data set, and the basic influencing parameters corresponding to the target yarn-making workshop were determined based on the parameter correlation analysis results corresponding to each target production batch. The standard deviation of the outlet temperature of the steady-state drying data set is calculated based on a preset sliding window, and the label is divided based on the standard deviation calculation result to obtain the drying outlet temperature fluctuation dataset corresponding to the target production batch. The degree of influence of the basic influence parameters on the global yarn drying outlet temperature fluctuation dataset is calculated based on the preset lag step size and target lag time, and the candidate influence parameters corresponding to the target yarn making workshop are determined based on the degree of influence of the variables. The steady-state yarn drying data set is divided based on the candidate influence parameters to obtain a first data subset containing the candidate influence parameters and a second data subset not containing the candidate influence parameters. The exit temperature of the first data subset and the second data subset is predicted based on the target machine learning model. The target influence parameters corresponding to the target yarn making workshop are determined based on the prediction accuracy of the prediction results.

[0007] According to another aspect of the present invention, an apparatus for determining the influence parameter of the wire drying outlet temperature is provided, comprising: The data acquisition module is used to acquire the set of steady-state drying data corresponding to the target production batch in the target yarn processing workshop; The correlation analysis module is used to perform parameter correlation analysis on the steady-state drying data set, and determine the basic influence parameters corresponding to the target yarn-making workshop based on the parameter correlation analysis results corresponding to each target production batch. The labeling module is used to calculate the standard deviation of the outlet temperature of the steady-state drying data set based on a preset sliding window, and to label the steady-state drying data set based on the standard deviation calculation result, so as to obtain the drying outlet temperature fluctuation dataset corresponding to the target production batch. The first determining module is used to calculate the degree of influence of the basic influence parameters on the global drying outlet temperature fluctuation dataset based on the preset lag step size and target lag time, and to determine the candidate influence parameters corresponding to the target yarn making workshop based on the degree of influence of the variables. The second determining module is used to divide the steady-state drying data set based on the candidate influence parameters to obtain a first data subset containing the candidate influence parameters and a second data subset not containing the candidate influence parameters, and to predict the outlet temperature of the first data subset and the second data subset based on the target machine learning model, and to determine the target influence parameters corresponding to the target yarn making workshop based on the prediction accuracy of the prediction results.

[0008] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the method for determining the influence parameters of the wire drying outlet temperature according to any embodiment of the present invention.

[0009] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the method for determining the influence parameters of the wire drying outlet temperature as described in any embodiment of the present invention.

[0010] According to another aspect of the present invention, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method for determining the influence parameters of the wire drying outlet temperature as described in any embodiment of the present invention.

[0011] The technical solution of this invention involves acquiring a steady-state drying data set corresponding to a target production batch in a target yarn-making workshop. Then, parameter correlation analysis is performed on the steady-state drying data set, and the basic influencing parameters corresponding to the target yarn-making workshop are determined based on the parameter correlation analysis results for each target production batch. Further, the standard deviation of the outlet temperature of the steady-state drying data set is calculated based on a preset sliding window, and the steady-state drying data set is labeled based on the standard deviation calculation results to obtain a drying outlet temperature fluctuation dataset corresponding to the target production batch. Then, the degree of influence of the basic influencing parameters on the global drying outlet temperature fluctuation dataset is calculated based on a preset lag step size and a target lag time, and candidate influencing parameters corresponding to the target yarn-making workshop are determined based on the degree of variable influence. Finally, the steady-state drying data set is partitioned based on the candidate influencing parameters to obtain a first data subset containing the candidate influencing parameters and a second data subset not containing the candidate influencing parameters. The outlet temperature of the first and second data subsets is predicted based on a target machine learning model, and the target influencing parameters corresponding to the target yarn-making workshop are determined based on the prediction accuracy of the prediction results. By establishing a causal relationship analysis method to identify the key driving factors affecting the fluctuation of the tobacco drying outlet temperature, this method solves the problem that existing technologies cannot quantitatively identify the specific driving factors affecting the fluctuation of the tobacco drying outlet temperature, leading to difficulties in tracing the source of fluctuations and a lack of targets for process optimization. It can accurately identify the key driving factors causing the fluctuation of the tobacco drying outlet temperature from a large number of parameters in the drying process, providing more accurate and reliable results for manually identifying the factors that play a key role in temperature fluctuations. At the same time, it also lays the foundation for controlling relevant production process parameters to maintain the stability of the tobacco drying outlet temperature in the later stages, thereby ensuring the overall quality of the tobacco.

[0012] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a flowchart of a method for determining the influence parameters of the wire drying outlet temperature according to Embodiment 1 of the present invention; Figure 2 This is a flowchart of a method for determining the influence parameters of the wire drying outlet temperature according to Embodiment 2 of the present invention; Figure 3 This is a flowchart of an optional method for determining the influence parameters of the wire drying outlet temperature according to Embodiment 2 of the present invention; Figure 4 This is a schematic diagram of a candidate influence parameter determination process provided in Embodiment 2 of the present invention; Figure 5 This is a schematic diagram of a device for determining the influence parameters of the wire drying outlet temperature according to Embodiment 3 of the present invention; Figure 6 This is a schematic diagram of the structure of an electronic device for implementing the method for determining the influence parameters of the wire drying outlet temperature according to an embodiment of the present invention. Detailed Implementation

[0015] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0016] It should be noted that the terms "first," "second," "target," "original," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0017] Example 1 Figure 1 This is a flowchart of a method for determining the influencing parameters of the wire drying outlet temperature according to Embodiment 1 of the present invention. This embodiment is applicable to situations where key driving factors causing fluctuations in the wire drying outlet temperature are accurately identified from a large number of parameters in the wire drying process. This method can be executed by a device for determining the influencing parameters of the wire drying outlet temperature, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes: S110. Obtain the set of steady-state drying data corresponding to the target production batch in the target yarn-making workshop.

[0018] In this context, "tobacco processing workshop" refers to the production workshop responsible for processing raw tobacco leaves into shredded tobacco during cigarette manufacturing. For example, a tobacco processing workshop may include a series of production lines such as drying machines, heating and humidifying equipment, and conveying equipment. Typically, a tobacco processing workshop may contain multiple production lines. "Target tobacco processing workshop" refers to the specific tobacco processing workshop where the influence factors of the drying outlet temperature need to be analyzed. Typically, the target tobacco processing workshop can be determined based on actual application requirements. "Production batch" refers to a complete production unit from the start to the end of the drying process according to the production process flow. Typically, a production batch corresponds to a continuous time series data, including process parameters and environmental parameters at all sampling times within that batch. "Target production batch" refers to one or more specific production batches selected for analysis. For example, the target production batch can be a qualified batch selected from historical production data that meets the requirements. "Steady-state drying data" refers to data collected when the drying process is in a relatively stable operating state. For example, steady-state drying data can be production data after removing the initial and final stages of production and abnormal disturbances. A steady-state drying data set can refer to a dataset composed of various steady-state drying data corresponding to the same target production batch.

[0019] S120. Perform parameter correlation analysis on the steady-state drying data set, and determine the basic influencing parameters corresponding to the target yarn-making workshop based on the parameter correlation analysis results corresponding to each target production batch.

[0020] Parameter correlation analysis refers to the process of using statistical methods to calculate the degree and significance of the linear correlation between each parameter in the steady-state wire drying data set and the wire drying outlet temperature. The parameter correlation analysis results can refer to a list of parameters significantly correlated with the outlet temperature of each target production batch, obtained after parameter correlation analysis. Influencing parameters can refer to process or environmental parameters that may affect the wire drying outlet temperature. Basic influencing parameters can refer to the global parameter set selected through cross-batch statistical synthesis based on the parameter correlation analysis results of each target production batch.

[0021] S130. Calculate the standard deviation of the outlet temperature of the steady-state drying data set based on a preset sliding window, and label the steady-state drying data set based on the standard deviation calculation result to obtain the drying outlet temperature fluctuation dataset corresponding to the target production batch.

[0022] The preset sliding window can refer to a pre-defined time window length. For example, the preset sliding window size can be 300. Standard deviation calculation refers to the process of calculating the sample standard deviation for the drying wire outlet temperature sequence within each preset sliding window. The standard deviation calculation result can refer to the sequence composed of the standard deviation values ​​calculated for each preset sliding window. Label partitioning refers to the process of determining labels for each window based on the distribution of the standard deviation values ​​in the standard deviation calculation result. For example, in this embodiment of the invention, label partitioning can include a stable period label and a high-fluctuation period label. A stable period can refer to a period where the fluctuation of the drying wire outlet temperature is relatively small. A stable period label can refer to a label used to mark a stable period. For example, a stable period label can be represented as 0. A high-fluctuation period can refer to a period where the fluctuation of the drying wire outlet temperature is relatively large. A high-fluctuation period label can refer to a label used to mark a high-fluctuation period. For example, a high-fluctuation period label can be represented as 1. The drying wire outlet temperature fluctuation dataset can refer to a dataset formed by adding a fluctuation label column to the steady-state drying wire data set.

[0023] In one optional implementation, the wire drying outlet temperature fluctuation dataset includes: a stable period data subset and a high fluctuation period data subset. The stable period data subset may refer to the data subset consisting of all sample points labeled as stable periods in the wire drying outlet temperature fluctuation dataset. The high fluctuation period data subset may refer to the data subset consisting of all sample points labeled as high fluctuation periods in the wire drying outlet temperature fluctuation dataset.

[0024] S140. Calculate the degree of influence of the basic influence parameters on the global yarn drying outlet temperature fluctuation dataset based on the preset lag step size and target lag time, and determine the candidate influence parameters corresponding to the target yarn making workshop based on the degree of influence of the variables.

[0025] The preset lag step can refer to a pre-set time offset. For example, the preset lag step can be 20, meaning the parameter values ​​from the 20 sampling points prior to the current moment are used. The lag time can refer to the lag step multiplied by the time length corresponding to the current sampling period's sequence number. For example, if the current sampling period is the first, the lag time can be 20; if the current sampling period is the second, the lag time can be 40. The target lag time can refer to the maximum lag time determined in practical applications. In this embodiment, for situations where the data lengths of high-fluctuation periods and stable periods are inconsistent, the target lag time can adopt the dynamic actual maximum lag time. The global drying wire outlet temperature fluctuation dataset can refer to the overall dataset obtained by stitching together the drying wire outlet temperature fluctuation datasets of all target production batches. The degree of variable influence can refer to the difference in causal strength between the high-fluctuation period and the stable period for each basic influence parameter. The candidate influence parameter can refer to the influence parameters selected from the basic influence parameters based on the degree of variable influence.

[0026] S150. Based on the candidate influence parameters, the steady-state yarn drying data set is divided into a first data subset containing the candidate influence parameters and a second data subset not containing the candidate influence parameters. The exit temperature of the first data subset and the second data subset is predicted based on the target machine learning model. The target influence parameters corresponding to the target yarn making workshop are determined based on the prediction accuracy of the prediction results.

[0027] The first data subset can refer to the data subset of the steady-state wire drying dataset that retains all basic variables and includes candidate influencing parameters. For example, the first data subset can be equivalent to the steady-state wire drying dataset. The second data subset can refer to the data subset of the steady-state wire drying dataset after removing candidate influencing parameters and retaining only the remaining basic variables. The target machine learning model can refer to a pre-trained machine learning algorithm used to predict the wire drying outlet temperature. The prediction result can refer to the predicted value of the wire drying outlet temperature output by the target machine learning model. The prediction accuracy can refer to the degree of agreement between the model's prediction result and the actual value. For example, the prediction accuracy can include the mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R²). 2 The target influencing parameter can refer to the parameter that, after verification through machine learning, has a significant impact on the fluctuation of the wire drying outlet temperature.

[0028] In an optional implementation, after predicting the outlet temperature of the first and second data subsets based on the target machine learning model and determining the target influencing parameters corresponding to the target tobacco processing workshop based on the prediction accuracy of the prediction results, the method may further include: during the production process, monitoring the parameter fluctuations corresponding to the target influencing parameters in real time; when fluctuations occur in the tobacco drying outlet temperature, prioritizing the inspection of the target influencing parameters; and adjusting the corresponding control parameters or control logic according to the fluctuation characteristics of the target influencing parameters to stabilize the tobacco drying outlet temperature, thereby improving the stability of the tobacco moisture content.

[0029] Fluctuation characteristics refer to the patterns and regularities of change of target influencing parameters over time. For example, fluctuation characteristics may include the amplitude, frequency, duration, and trend of fluctuation. Control parameters refer to setpoints or coefficients that can be directly adjusted in a production control system. For example, control parameters may include upper and lower limits for valve opening, limits for damper opening speed, and limits for the rate of change of temperature settings. Control logic refers to the pre-set judgment rules, decision-making processes, or algorithms in a control system.

[0030] Specifically, after identifying the target influencing parameters for the yarn-making workshop and entering the production application stage: First, in the real-time monitoring system of the production line, the target influencing parameters are set as key monitoring variables. The system collects the actual values ​​of these parameters in real time at a frequency of seconds or sub-seconds and calculates their rolling standard deviation, rate of change, and other fluctuation characteristic indicators. Second, when the monitoring system detects abnormal fluctuations in the yarn drying outlet temperature, it automatically triggers an early warning and prioritizes checking the real-time trend chart of the target influencing parameters, such as whether there have been drastic jumps in hot air temperature or abnormal oscillations in the opening of the dehumidification damper in the past two minutes. If it is found that the fluctuation of a certain target influencing parameter is highly consistent with the fluctuation of the outlet temperature in time, the parameter is quickly identified as the source of the fluctuation. Then, based on the fluctuation characteristics of the driving factor, targeted adjustment measures are taken. Finally, after the adjustment is completed, the system continues to monitor the changing trends of the outlet temperature and the target influencing parameters in real time. If the outlet temperature returns to stability within a few minutes and the fluctuation of the target influencing parameters is suppressed, the adjustment is confirmed to be effective; if the fluctuation persists, the above investigation and adjustment process is repeated until the outlet temperature stabilizes within the allowable range of the process. Therefore, through the above closed-loop implementation process, the driving factors identified offline can be transformed into the core basis for online real-time monitoring and active control, thereby significantly improving the stability of the drying outlet temperature and the consistency of the finished tobacco quality.

[0031] The technical solution of this invention involves acquiring a steady-state drying data set corresponding to a target production batch in a target yarn-making workshop. Then, parameter correlation analysis is performed on the steady-state drying data set, and the basic influencing parameters corresponding to the target yarn-making workshop are determined based on the parameter correlation analysis results for each target production batch. Further, the standard deviation of the outlet temperature of the steady-state drying data set is calculated based on a preset sliding window, and the steady-state drying data set is labeled based on the standard deviation calculation results to obtain a drying outlet temperature fluctuation dataset corresponding to the target production batch. Then, the degree of influence of the basic influencing parameters on the global drying outlet temperature fluctuation dataset is calculated based on a preset lag step size and a target lag time, and candidate influencing parameters corresponding to the target yarn-making workshop are determined based on the degree of variable influence. Finally, the steady-state drying data set is partitioned based on the candidate influencing parameters to obtain a first data subset containing the candidate influencing parameters and a second data subset not containing the candidate influencing parameters. The outlet temperature of the first and second data subsets is predicted based on a target machine learning model, and the target influencing parameters corresponding to the target yarn-making workshop are determined based on the prediction accuracy of the prediction results. By establishing a causal relationship analysis method to identify the key driving factors affecting the fluctuation of the tobacco drying outlet temperature, this method solves the problem that existing technologies cannot quantitatively identify the specific driving factors affecting the fluctuation of the tobacco drying outlet temperature, leading to difficulties in tracing the source of fluctuations and a lack of targets for process optimization. It can accurately identify the key driving factors causing the fluctuation of the tobacco drying outlet temperature from a large number of parameters in the drying process, providing more accurate and reliable results for manually identifying the factors that play a key role in temperature fluctuations. At the same time, it also lays the foundation for controlling relevant production process parameters to maintain the stability of the tobacco drying outlet temperature in the later stages, thereby ensuring the overall quality of the tobacco.

[0032] Example 2 Figure 2 This is a flowchart of a method for determining the influence parameter of the wire drying outlet temperature according to Embodiment 2 of the present invention. This embodiment is a refinement based on the above embodiment. Specifically, this embodiment refines the "parameter correlation analysis of the steady-state wire drying data set", which may include: calculating the correlation coefficient between the target wire drying parameter and the corresponding wire drying outlet temperature in the steady-state wire drying data set; determining the significance coefficient corresponding to the target wire drying parameter based on the correlation coefficient and a preset significance check; performing a threshold judgment on the correlation coefficient based on a preset correlation threshold to obtain a first threshold judgment result, and performing a threshold judgment on the significance coefficient based on a preset significance threshold to obtain a second threshold judgment result; marking the target wire drying parameter as the original influence parameter based on the first threshold judgment result and the second threshold judgment result, and using the original influence parameter as the parameter correlation analysis result. Figure 2 As shown, the method includes: S210. Obtain the set of original drying data corresponding to the original production batch in the target yarn processing workshop.

[0033] The original production batch can refer to all production batches directly collected from the yarn-making workshop without any screening. For example, the original production batch can include normal batches and batches that may contain anomalies. The original yarn-drying data can refer to raw data collected directly during the yarn-drying process according to a set sampling period, without any processing. For example, the original yarn-drying data can include production process data and environmental data. Production process data can include steam pressure, hot air temperature, exhaust damper opening, and hot air velocity, etc. Environmental data can include workshop ambient temperature and humidity, etc. In this embodiment of the invention, the set sampling period can be 1 second. The original yarn-drying data set can refer to a dataset composed of all original yarn-drying data corresponding to the same original production batch. Typically, the size of the original yarn-drying data set is related to the number of sampling points.

[0034] S220. Based on the original yarn drying data set, batch data selection is performed on the original production batch to determine the target production batch corresponding to the target yarn making workshop.

[0035] Batch data selection refers to the process of selecting qualified production batches from the original production batches according to preset quality standards as target production batches. For example, the preset quality standards can be to exclude batches with equipment failures, start-up / shutdown transition periods, and incomplete records of manual intervention.

[0036] S230. Based on the cumulative amount of the electronic scale, perform data cleaning on the original drying data set corresponding to the target production batch, remove the initial data and later data in the original drying data set, and obtain the steady-state drying data set corresponding to the target production batch in the target yarn making workshop.

[0037] The cumulative amount measured by the electronic scale refers to the total amount of tobacco shreds accumulated at the entrance of the tobacco drying process. Typically, the cumulative amount reflects the progress of material entering the drying machine. Data cleaning refers to the process of cleaning the original tobacco drying data corresponding to the target production batch, removing invalid, erroneous, or non-steady-state data segments, and retaining high-quality, representative steady-state data. Initial data refers to the tobacco drying data corresponding to the beginning stage of the production batch. Typically, when the cumulative amount measured by the electronic scale is <= 500 kg, the tobacco drying data corresponding to that time period is considered initial data. Later data refers to the tobacco drying data corresponding to the end of the production batch. Typically, when the cumulative amount measured by the electronic scale reaches 0.9 times the maximum cumulative amount measured by the electronic scale, the tobacco drying data corresponding to that time period is considered later data.

[0038] Specifically, firstly, from the production execution system or centralized control system of the target yarn processing workshop, export the original yarn drying data sets corresponding to all original production batches. Each original yarn drying data set contains second-level sampling data of the entire process from feeding to discharging for the corresponding production batch. Then, based on the batch quality records, eliminate abnormal batches with interruptions, shutdowns, equipment failures, or severely missing data, and select qualified batches with continuous production and complete data as target production batches. Further, for each target production batch, obtain the time-series record of the cumulative amount of the electronic scale within that batch. Set a lower threshold for the cumulative amount of the electronic scale, such as 500 kg, and an upper threshold, such as 90% of the maximum cumulative amount of the batch. Mark the data segments with the cumulative amount of the electronic scale below the lower threshold as early production data, the data segments above the upper threshold as late production data, and the remaining intermediate data as steady-state production data. Finally, remove the early and late data from the original yarn drying data sets, retaining only the steady-state production data to form a steady-state yarn drying data set for subsequent correlation analysis and fluctuation feature extraction.

[0039] S240. Calculate the correlation coefficient between the target drying parameters and the corresponding drying outlet temperature in the steady-state drying data set.

[0040] The target drying parameter can refer to the drying parameter selected for correlation analysis from the steady-state drying data set. For example, the target drying parameter can be any candidate process parameter or environmental parameter in the steady-state drying data set except for the drying outlet temperature. The correlation coefficient can refer to a statistic used to measure the degree of linear correlation between the target drying parameter and the corresponding drying outlet temperature. Typically, the Pearson correlation coefficient is used. For example, using the Pearson correlation coefficient, the steady-state drying data set is shown in Table 1 below, which contains n time-series sample points. Each time-series sample point contains k target drying parameters such as hot air velocity, steam pressure, and outlet temperature. Each row in Table 1 corresponds to the same batch and the same sampling time. Therefore, for a certain target drying parameter... and the corresponding wire drying outlet temperature You can follow the formula: Calculate the correlation coefficient. Wherein, , This represents a pair of values ​​corresponding to the j-th sampling time point. Represents the target wire drying parameters at time j. The corresponding single numeric value; This represents the wire outlet temperature at time j. Indicating steady-state wire drying data set The overall mean of all samples for the variable. Y represents the total average value of all samples of the wire drying outlet temperature in the steady-state wire drying data set.

[0041] Table 1. Set of steady-state wire drying data S250. Based on the correlation coefficient and the preset significance check, determine the significance coefficient corresponding to the target wire drying parameter.

[0042] The pre-defined significance check can refer to a pre-set statistical hypothesis testing method used to determine whether the calculated correlation coefficient is statistically significant. For example, the pre-defined significance check can be a t-test. The significance coefficient can refer to the quantitative indicator obtained after the significance check. Typically, the significance coefficient can be the p-value. For example, continuing with the above example, taking the t-test as the pre-defined significance check, the formula can be used: Calculate the statistical values ​​corresponding to the target drying parameters, and then look up the t-distribution table according to df=n-2 to obtain the p-value.

[0043] S260. The correlation coefficient is judged based on a preset correlation threshold to obtain a first threshold judgment result, and the significance coefficient is judged based on a preset significance threshold to obtain a second threshold judgment result.

[0044] The preset relevance threshold can refer to a pre-set value used to filter parameters with strong relevance. For example, the preset relevance threshold can be the minimum acceptable value of the absolute value of the correlation coefficient, such as 0.5. The first threshold judgment result can refer to the result obtained by comparing the absolute value of the correlation coefficient of the target drying parameter with the preset relevance threshold. For example, the first threshold judgment result can be either a correlation coefficient less than the preset significance threshold or a correlation coefficient exceeding the preset significance threshold.

[0045] The preset significance threshold can refer to a pre-set value used to determine whether the significance coefficient is sufficiently small. For example, the preset significance threshold could be 0.05. The second threshold judgment result can refer to the result obtained by comparing the significance coefficient of the target drying wire parameter with the preset significance threshold. For example, the second threshold judgment result could be a significance coefficient less than the preset significance threshold, or a significance coefficient exceeding the preset significance threshold.

[0046] S270. Based on the first threshold judgment result and the second threshold judgment result, the target wire drying parameter is marked as the original influence parameter, and the original influence parameter is used as the parameter correlation analysis result.

[0047] Here, the original influence parameter can refer to the influence parameter obtained from preliminary judgment. For example, the original influence parameter can be a target drying wire parameter that satisfies both strong correlation and statistical significance. Continuing with the above example, the original influence parameter can be a target drying wire parameter whose correlation coefficient exceeds a preset significance threshold and whose significance coefficient is less than a preset significance threshold.

[0048] Specifically, for the steady-state wire drying data set corresponding to the target production batch, for each target wire drying parameter within that batch, its time-series data is paired with the time-series data of the wire drying outlet temperature, and the Pearson correlation coefficient r is calculated. Simultaneously, a t-test is used to calculate the p-value corresponding to the correlation coefficient. Correlation thresholds and significance thresholds are set. For the current target wire drying parameter, firstly, it is determined whether |r| is greater than or equal to the preset correlation threshold: if yes, the first threshold judgment result is satisfied; otherwise, it is not satisfied. Then, it is determined whether the p-value is less than the preset significance threshold: if yes, the second threshold judgment result is significant; otherwise, it is not significant. Only when both judgment results are yes, the target wire drying parameter is marked as the original influencing parameter of that batch. After traversing all target wire drying parameters in that batch, the original influencing parameter list for that batch is obtained. The above process is repeated for all target production batches to obtain the original influencing parameter list for each batch, which serves as the result of parameter correlation analysis and is used for subsequent cross-batch frequency statistics and basic influencing parameter screening.

[0049] S280. Summarize and process the frequency of occurrence of the original influencing parameters in each steady-state wire drying data set.

[0050] The frequency of occurrence can be defined as the ratio of the number of times a certain original influencing parameter appears in the original influencing parameter list of all target production batches to the total number of target production batches. For example, if the original influencing parameter is hot air velocity, and there are 100 target production batches, and hot air velocity appears in the original influencing parameter list of 85 batches, then the frequency of occurrence of the original influencing parameter hot air velocity can be: 85 / 100 = 0.85.

[0051] S290. Based on a preset frequency threshold, a threshold judgment is made on the occurrence frequency to obtain a third threshold judgment result.

[0052] The preset frequency threshold can refer to a pre-set minimum frequency value used to determine whether a certain original influence parameter is sufficiently prevalent in multiple batches. For example, the preset frequency threshold can be set to 0.5. The third threshold judgment result can refer to the result obtained by comparing the occurrence frequency of a certain original influence parameter with the preset frequency threshold. For example, the third threshold judgment result can be an occurrence frequency greater than or equal to the preset frequency threshold, or an occurrence frequency less than the preset frequency threshold.

[0053] S2100. Based on the third threshold judgment result, the original influence parameter is determined as the basic influence parameter corresponding to the target yarn-making workshop.

[0054] Specifically, after completing the parameter correlation analysis for all target production batches, a list of original influencing parameters was obtained for each batch. For example, Batch 1: [Hot air velocity, exhaust damper opening, steam pressure]; Batch 2: [Hot air velocity, workshop humidity, cylinder wall temperature]; Batch 3: [Hot air velocity, exhaust damper opening, hot air temperature], etc. The frequency of each original influencing parameter, such as hot air velocity, exhaust damper opening, steam pressure, or workshop humidity, was calculated across all batches and divided by the total number of batches. A preset frequency threshold was set, for example, 0.5. For each original influencing parameter, its frequency was checked to see if it was greater than or equal to 0.5: if it was, the third threshold judgment result was passed; otherwise, it was failed. All original influencing parameters that passed the third threshold judgment were determined as the basic influencing parameters corresponding to the target yarn-making workshop. For example, if hot air velocity occurs in 85% of batches, exhaust damper opening in 72% of batches, and steam pressure in 48% of batches, then when the threshold is 0.5, hot air velocity and exhaust damper opening are identified as the fundamental influencing parameters, while steam pressure is excluded due to insufficient frequency. Thus, by using frequency statistics to exclude parameters that are only occasionally relevant in a few specific batches but irrelevant in most batches, the stability and generalization ability of the fundamental influencing parameters under different production conditions can be ensured.

[0055] S2110. Calculate the standard deviation of the outlet temperature of the steady-state drying data set based on a preset sliding window, and label the steady-state drying data set based on the standard deviation calculation result to obtain the drying outlet temperature fluctuation dataset corresponding to the target production batch.

[0056] Specifically, after determining the steady-state drying data set corresponding to each target production batch, the standard deviation of the drying outlet temperature within each preset sliding window can be calculated sequentially according to the production time order to obtain the rolling standard deviation sequence of the target production batch, i.e., the standard deviation calculation result. Next, the rolling standard deviation sequence corresponding to the target production batch is sorted in ascending order, and a quantile threshold is set, such as the 90th quantile. Time points corresponding to windows exceeding this quantile threshold are marked as high-fluctuation periods (labeled 1), and other time points are marked as stable periods (labeled 0). Finally, the generated fluctuation labels are added as a new column to the steady-state drying data set, forming a drying outlet temperature fluctuation dataset with fluctuation labels. This dataset contains a stable period subset labeled 0 and a high-fluctuation period subset labeled 1, for subsequent causal analysis.

[0057] S2120. Calculate the first causal strength of the basic influence parameter in the global stationary period data subset based on the preset lag step size and target lag time.

[0058] Here, causal strength refers to a quantitative indicator used to measure the magnitude of the causal effect of a certain basic influence parameter on the wire drying outlet temperature. For example, with the basic influence parameter X, the wire drying outlet temperature Y, a preset lag step of 20 seconds, and a target lag time of... This allows us to determine the set of lag points. For example, we can first use the Gaussian kernel density approximation to calculate the probability density distribution of each basic influence parameter, and then use the formula: Calculate the differential entropy of the data. The sample point can represent the current Y and lagged Y, the univariate lagged Y, the lagged X and lagged Y, and the lagged X and the current Y and lagged Y. This is the probability density of the sample points. Furthermore, according to the formula: Calculate the causal strength. In the formula... As the lag point, t represents the current sampling point. N represents the total number of valid sample points after merging under this condition. Taking a sampling period of 1 second, a total sample size of 8000 after merging all batches during the high-fluctuation period, a preset lag step of 20 seconds, and a target lag time of 200 seconds as an example, the final total number of valid samples N' = 8000 - 200 = 7800; =20: The number of valid sample points is 7800-20=7780; for =40: The number of valid sample points is 7800-40=7760; for =200: The number of valid sample points is 7800-200=7600.

[0059] It is worth noting that, in this embodiment of the invention, the average causal strength of each lag step can be calculated as the final causal strength, and the formula is as follows: .

[0060] The global stationary period data subset can refer to a large dataset formed by merging sample points marked as stationary periods from all target production batches. The first causal strength can refer to the causal strength value of a certain basic influence parameter on the wire drying outlet temperature, calculated on the global stationary period data subset.

[0061] S2130. Calculate the second causal strength of the basic influence parameter in the global high-fluctuation period data subset based on the preset lag step size and target lag time.

[0062] The global high-fluctuation period data subset can refer to a large dataset formed by merging sample points marked as high-fluctuation periods from all target production batches. The second causal strength can refer to the causal strength value of the same basic influence parameter on the wire drying outlet temperature, calculated on the global high-fluctuation period data subset.

[0063] S2140. Based on the calculation result of the difference between the first causal intensity and the second causal intensity, determine the degree of influence of the basic influence parameter on the variable in the global wire drying outlet temperature fluctuation dataset.

[0064] The difference calculation result can refer to the ratio obtained by subtracting the first causal strength from the second causal strength. For example, taking the first causal strength as... The second causal strength is For example, the degree of influence of a variable can be expressed as: Generally, the greater the influence of a variable, the stronger the causal effect of that fundamental parameter on the outlet temperature during periods of high fluctuation, and the more likely it is to be a driving factor causing temperature fluctuations.

[0065] Specifically, after obtaining the dataset of temperature fluctuations at the drying outlet corresponding to the target production batch, the sample points of the stable period of all target production batches can be concatenated in chronological order to form a global stable period data subset, and the sample points of the high-fluctuation period of all target production batches can be merged to form a global high-fluctuation period data subset. A preset lag step and target lag time are set. For each basic influence parameter, on the global stable period data subset, the conditional mutual information is calculated for each lag step as the causal strength at that lag step. Then, the average of the causal strengths for all lag steps is taken to obtain the first causal strength. Similarly, the second causal strength is calculated on the global high-fluctuation period data subset. Then, the difference ratio is calculated as the degree of influence of the basic influence parameter under the global drying outlet temperature fluctuation dataset. The above calculation is repeated for all basic influence parameters to obtain the degree of influence of each basic influence parameter. Thus, by calculating the causal strengths of the stable period and the high-fluctuation period separately, the change in the role of the same parameter under different fluctuation states can be quantitatively assessed, thereby identifying parameters that only have a significant effect during the fluctuation period, rather than general correlation parameters that are significant in all periods.

[0066] S2150. Based on the degree of influence of the variables, determine the candidate influence parameters corresponding to the target silk-making workshop.

[0067] Specifically, after determining the influence of the basic influence parameter on the global drying wire outlet temperature fluctuation dataset, a preset influence threshold, such as 75%, can be used to judge the influence of the variable. If the influence of the variable exceeds the preset influence threshold, the basic influence parameter is used as a candidate influence parameter.

[0068] S2160. Based on the candidate influence parameters, the steady-state wire drying data set is divided into a first data subset containing the candidate influence parameters and a second data subset not containing the candidate influence parameters.

[0069] Specifically, after determining the list of candidate influencing parameters, two identical datasets are copied from the steady-state wire drying dataset, serving as the original templates for the first and second data subsets, respectively. For the first data subset, all original feature columns are retained without any removal; for the second data subset, the columns corresponding to the candidate influencing parameters are removed, retaining only the remaining basic variables. Essentially, the two data subsets are identical in sample size, sample order, and other non-candidate variables, the only difference being that the second data subset does not contain information about the candidate influencing parameters. After partitioning, the first and second data subsets are used for training and prediction of subsequent machine learning models, respectively, to compare the impact of including and omitting candidate influencing parameters on the accuracy of outlet temperature prediction.

[0070] S2170. Based on the target machine learning model, the outlet temperature of the first data subset is predicted to obtain a first prediction result, and based on the target machine learning model, the outlet temperature of the second data subset is predicted to obtain a second prediction result.

[0071] The first prediction result can refer to the predicted temperature of the wire drying outlet output by the target machine learning model after inputting the first subset of data. The second prediction result can refer to the predicted temperature of the wire drying outlet output by the same target machine learning model after inputting the second subset of data.

[0072] S2180. Calculate the prediction accuracy of the first prediction result based on the true values ​​in the first data subset to obtain a first prediction accuracy, and calculate the prediction accuracy of the second prediction result based on the true values ​​in the second data subset to obtain a second prediction accuracy.

[0073] The true value refers to the actual measured value of the wire drying outlet temperature, which is recorded and verified in the dataset. Typically, the true value is objectively existing data collected by a high-precision sensor. The first prediction accuracy refers to the model prediction performance index calculated based on the first prediction result and the corresponding true value. Typically, the first prediction accuracy can be used to quantify the prediction accuracy of the model when candidate influencing parameters are included. The second prediction accuracy refers to the model prediction performance index calculated based on the second prediction result and the corresponding true value. Typically, the second prediction accuracy can be used to quantify the prediction accuracy of the model when candidate influencing parameters are not included.

[0074] S2190. Compare the first prediction accuracy and the second prediction accuracy numerically, and use the candidate influence parameter as the target influence parameter corresponding to the target yarn-making workshop based on the numerical comparison result.

[0075] The numerical comparison result refers to the conclusion drawn after comparing the first prediction accuracy with the second prediction accuracy. For example, the numerical comparison result could show a large difference between the first and second prediction accuracies, or it could show a small difference. Generally, if the numerical comparison result shows a large difference between the first and second prediction accuracies, it indicates that the candidate influencing parameter is indeed important. Conversely, if the numerical comparison result shows a small difference between the first and second prediction accuracies, it indicates that the candidate influencing parameter is invalid.

[0076] Specifically, after obtaining a first data subset containing candidate influence parameters and a second data subset not containing candidate influence parameters, the first and second data subsets can be divided into training and test sets according to time sequence or batch, respectively. Then, using the same target machine learning model, the model is trained on the training set of the first data subset and predicted on the test set of the first data subset to obtain a first prediction result; similarly, the model is trained on the training set of the second data subset and predicted on its test set to obtain a second prediction result. Further, the actual measured values ​​of the yarn drying outlet temperature in the test set are extracted as the true values. The prediction accuracy of the first prediction result and the true value, and the prediction accuracy of the second prediction result and the true value are calculated respectively. The first and second prediction accuracies are compared: if the first prediction accuracy is significantly different from the second prediction accuracy, the numerical comparison result indicates that the candidate influence parameter is valid and is determined as the target influence parameter corresponding to the target yarn processing workshop; otherwise, the candidate influence parameter is considered to have no significant contribution to the prediction and is not adopted. Thus, through this verification, the final output is a target influence parameter that has been doubly confirmed, which can be used for subsequent real-time monitoring and process optimization.

[0077] The technical solution of this invention involves obtaining the original drying data set corresponding to the original production batch in the target yarn-making workshop. Based on the original drying data set, batch data selection is performed on the original production batch to determine the target production batch corresponding to the target yarn-making workshop. Data cleaning is performed on the original drying data set corresponding to the target production batch based on the accumulated amount of the electronic scale, removing initial and later data to obtain the steady-state drying data set corresponding to the target production batch in the target yarn-making workshop. Further, the correlation coefficient between the target drying parameter and the corresponding drying outlet temperature in the steady-state drying data set is calculated. Based on the correlation coefficient and a preset significance check, the significance coefficient corresponding to the target drying parameter is determined. A threshold judgment is performed on the correlation coefficient based on a preset correlation threshold to obtain a first threshold judgment result, and a threshold judgment is performed on the significance coefficient based on a preset significance threshold to obtain a second threshold judgment result. Based on the first and second threshold judgment results, the target drying parameter is marked as the original influencing parameter, and the original influencing parameter is used as the parameter correlation analysis result. Further, the frequency of occurrence of the original influencing parameter in each steady-state drying data set is summarized and processed. A threshold judgment is performed on the frequency of occurrence based on a preset frequency threshold to obtain a third threshold judgment result. Based on the third threshold judgment result, the original influencing parameters are determined as the basic influencing parameters corresponding to the target yarn-making workshop. Further, the standard deviation of the outlet temperature is calculated for the steady-state yarn-drying data set based on a preset sliding window, and the steady-state yarn-drying data set is labeled based on the standard deviation calculation result to obtain the yarn-drying outlet temperature fluctuation dataset corresponding to the target production batch. The first causal strength of the basic influencing parameters in the global stable period data subset is calculated based on a preset lag step size and a target lag time. The second causal strength of the basic influencing parameters in the global high-fluctuation period data subset is calculated based on a preset lag step size and a target lag time. Based on the difference between the first and second causal strengths, the variable influence degree of the basic influencing parameters in the global yarn-drying outlet temperature fluctuation dataset is determined. Further, candidate influencing parameters corresponding to the target yarn-making workshop are determined based on the variable influence degree. The steady-state yarn-drying data set is divided based on the candidate influencing parameters to obtain a first data subset containing the candidate influencing parameters and a second data subset not containing the candidate influencing parameters. Finally, the outlet temperature is predicted for the first data subset based on the target machine learning model, yielding a first prediction result. Similarly, the outlet temperature is predicted for the second data subset based on the same target machine learning model, yielding a second prediction result. The prediction accuracy is calculated for the first prediction result based on the actual values ​​in the first data subset, yielding a first prediction accuracy. The prediction accuracy is then calculated for the second prediction result based on the actual values ​​in the second data subset, yielding a second prediction accuracy. The first and second prediction accuracies are numerically compared, and based on the numerical comparison results, candidate influencing parameters are selected as the target influencing parameters corresponding to the target spinning workshop.By establishing a causal relationship analysis method to identify the key driving factors affecting the fluctuation of the tobacco drying outlet temperature, this method solves the problem that existing technologies cannot quantitatively identify the specific driving factors affecting the fluctuation of the tobacco drying outlet temperature, leading to difficulties in tracing the source of fluctuations and a lack of targets for process optimization. It can accurately identify the key driving factors causing the fluctuation of the tobacco drying outlet temperature from a large number of parameters in the drying process, providing more accurate and reliable results for manually identifying the factors that play a key role in temperature fluctuations. At the same time, it also lays the foundation for controlling relevant production process parameters to maintain the stability of the tobacco drying outlet temperature in the later stages, thereby ensuring the overall quality of the tobacco.

[0078] Figure 3 The diagram shows a flowchart of an optional method for determining the influence parameters of the drying outlet temperature provided by an embodiment of the present invention. Specifically, firstly, the original drying datasets corresponding to the original production batches in the target yarn-making workshop are collected and merged. Batch data selection and data cleaning are then performed on the original drying datasets corresponding to the original production batches to obtain a steady-state drying dataset corresponding to the target production batch in the target yarn-making workshop. Next, the correlation coefficient, significance coefficient, and frequency of occurrence between the target drying parameters and the corresponding drying outlet temperature in each steady-state drying dataset are calculated. Preset correlation thresholds, preset significance thresholds, and preset frequency thresholds are used to filter the correlation coefficients, significance coefficients, and frequencies to obtain the basic influence parameters. Further, the standard deviation of the outlet temperature is calculated for the steady-state drying datasets based on a preset sliding window. Based on the standard deviation calculation results, the steady-state drying datasets are labeled to obtain a drying outlet temperature fluctuation dataset corresponding to the target production batch, containing a stable period data subset and a high-fluctuation period data subset. Further, the causal strength of the basic influence parameters in the stable period and the high-fluctuation period is calculated. Based on this causal strength, the variable influence degree of the basic influence parameters under the global drying outlet temperature fluctuation dataset is calculated. If the influence of a variable exceeds a preset threshold, the basic influence parameter is used as a candidate influence parameter. Further, the steady-state yarn drying dataset is partitioned using the candidate influence parameters to obtain a first subset containing the candidate influence parameters and a second subset not containing them, followed by data normalization. A target machine learning model is then used to predict the outlet temperature for both the first and second subsets, and the influence of the candidate influence parameters is verified by comparing the prediction accuracy, thus determining the target influence parameter corresponding to the target yarn processing workshop.

[0079] Figure 4The diagram illustrates a process for determining candidate influencing parameters according to an embodiment of the present invention. Specifically, firstly, the standard deviation of the outlet temperature is calculated for the steady-state drying data set based on a preset sliding window. Then, the steady-state drying data set is labeled based on the standard deviation calculation results to obtain the drying outlet temperature fluctuation dataset corresponding to the target production batch. Next, a preset lag step and a target lag time are set. The differential entropy of the sample points is calculated using kernel density estimation probability. The causal strength of each lag step is calculated using the differential entropy, and the average causal strength of each lag step is calculated as the final causal strength. Thus, the first causal strength under the steady-state period and the second causal strength under the high-fluctuation period are obtained. Further, based on the difference between the first and second causal strengths, the degree of influence of the basic influencing parameters on the global drying outlet temperature fluctuation dataset is determined. Finally, a preset influence degree threshold is set to determine the candidate influencing parameters corresponding to the target yarn-making workshop.

[0080] Example 3 Figure 5 This is a schematic diagram of a device for determining the influence parameters of the wire drying outlet temperature according to Embodiment 3 of the present invention. Figure 5 As shown, the device includes: a data acquisition module 310, a correlation analysis module 320, a label segmentation module 330, a first determination module 340, and a second determination module 350; Among them, the data acquisition module 310 is used to acquire the set of steady-state drying data corresponding to the target production batch in the target yarn making workshop; The correlation analysis module 320 is used to perform parameter correlation analysis on the steady-state drying data set, and determine the basic influence parameters corresponding to the target yarn making workshop based on the parameter correlation analysis results corresponding to each target production batch. The labeling module 330 is used to calculate the standard deviation of the outlet temperature of the steady-state drying data set based on a preset sliding window, and to label the steady-state drying data set based on the standard deviation calculation result, so as to obtain the drying outlet temperature fluctuation dataset corresponding to the target production batch. The first determining module 340 is used to calculate the degree of influence of the basic influence parameters on the global drying outlet temperature fluctuation dataset based on the preset lag step size and target lag time, and to determine the candidate influence parameters corresponding to the target yarn making workshop based on the degree of influence of the variables. The second determining module 350 is used to divide the steady-state drying data set based on the candidate influence parameters to obtain a first data subset containing the candidate influence parameters and a second data subset not containing the candidate influence parameters, and to predict the outlet temperature of the first data subset and the second data subset based on the target machine learning model, and to determine the target influence parameters corresponding to the target yarn making workshop based on the prediction accuracy of the prediction results.

[0081] The technical solution of this invention involves acquiring a steady-state drying data set corresponding to a target production batch in a target yarn-making workshop. Then, parameter correlation analysis is performed on the steady-state drying data set, and the basic influencing parameters corresponding to the target yarn-making workshop are determined based on the parameter correlation analysis results for each target production batch. Further, the standard deviation of the outlet temperature of the steady-state drying data set is calculated based on a preset sliding window, and the steady-state drying data set is labeled based on the standard deviation calculation results to obtain a drying outlet temperature fluctuation dataset corresponding to the target production batch. Then, the degree of influence of the basic influencing parameters on the global drying outlet temperature fluctuation dataset is calculated based on a preset lag step size and a target lag time, and candidate influencing parameters corresponding to the target yarn-making workshop are determined based on the degree of variable influence. Finally, the steady-state drying data set is partitioned based on the candidate influencing parameters to obtain a first data subset containing the candidate influencing parameters and a second data subset not containing the candidate influencing parameters. The outlet temperature of the first and second data subsets is predicted based on a target machine learning model, and the target influencing parameters corresponding to the target yarn-making workshop are determined based on the prediction accuracy of the prediction results. By establishing a causal relationship analysis method to identify the key driving factors affecting the fluctuation of the tobacco drying outlet temperature, this method solves the problem that existing technologies cannot quantitatively identify the specific driving factors affecting the fluctuation of the tobacco drying outlet temperature, leading to difficulties in tracing the source of fluctuations and a lack of targets for process optimization. It can accurately identify the key driving factors causing the fluctuation of the tobacco drying outlet temperature from a large number of parameters in the drying process, providing more accurate and reliable results for manually identifying the factors that play a key role in temperature fluctuations. At the same time, it also lays the foundation for controlling relevant production process parameters to maintain the stability of the tobacco drying outlet temperature in the later stages, thereby ensuring the overall quality of the tobacco.

[0082] Optionally, the correlation analysis module 320 can be specifically used to: calculate the correlation coefficient between the target wire drying parameter and the corresponding wire drying outlet temperature in the steady-state wire drying data set; determine the significance coefficient corresponding to the target wire drying parameter based on the correlation coefficient and a preset significance check; perform threshold judgment on the correlation coefficient based on a preset correlation threshold to obtain a first threshold judgment result, and perform threshold judgment on the significance coefficient based on a preset significance threshold to obtain a second threshold judgment result; and mark the target wire drying parameter as the original influencing parameter based on the first threshold judgment result and the second threshold judgment result, and use the original influencing parameter as the parameter correlation analysis result.

[0083] Optionally, the correlation analysis module 320 can be used to: summarize and process the frequency of occurrence of the original influence parameter in each steady-state yarn drying data set; perform threshold judgment on the frequency of occurrence based on a preset frequency threshold to obtain a third threshold judgment result; and determine the original influence parameter as the basic influence parameter corresponding to the target yarn processing workshop according to the third threshold judgment result.

[0084] Optionally, the data set of temperature fluctuation at the wire drying outlet includes: a subset of data during stable periods and a subset of data during periods of high fluctuation; The first determining module 340 can be specifically used to: calculate the first causal intensity of the basic influence parameter in the global stable period data subset based on a preset lag step size and a target lag time; calculate the second causal intensity of the basic influence parameter in the global high fluctuation period data subset based on a preset lag step size and a target lag time; and determine the degree of variable influence of the basic influence parameter in the global drying wire outlet temperature fluctuation dataset based on the difference between the first causal intensity and the second causal intensity.

[0085] Optionally, the data acquisition module 310 can be used to: acquire the original drying data set corresponding to the original production batch in the target yarn processing workshop; select batch data for the original production batch based on the original drying data set to determine the target production batch corresponding to the target yarn processing workshop; and clean the original drying data set corresponding to the target production batch based on the accumulated amount of the electronic scale, removing the initial and later data in the original drying data set to obtain the steady-state drying data set corresponding to the target production batch in the target yarn processing workshop.

[0086] Optionally, the second determining module 350 may be specifically used for: predicting the outlet temperature of the first data subset based on the target machine learning model to obtain a first prediction result, and predicting the outlet temperature of the second data subset based on the target machine learning model to obtain a second prediction result; calculating the prediction accuracy of the first prediction result based on the true values ​​in the first data subset to obtain a first prediction accuracy, and calculating the prediction accuracy of the second prediction result based on the true values ​​in the second data subset to obtain a second prediction accuracy; numerically comparing the first prediction accuracy and the second prediction accuracy, and using the candidate influence parameter as the target influence parameter corresponding to the target spinning workshop based on the numerical comparison result.

[0087] Optionally, the device for determining the influencing parameters of the tobacco drying outlet temperature may further include: a post-processing module, used to monitor the parameter fluctuations corresponding to the target influencing parameters in real time during the production process after the outlet temperature of the first data subset and the second data subset is predicted based on the target machine learning model and the target influencing parameters are determined based on the prediction accuracy of the prediction results; when the tobacco drying outlet temperature fluctuates, the target influencing parameters are checked first; and the corresponding control parameters or control logic are adjusted according to the fluctuation characteristics of the target influencing parameters to stabilize the tobacco drying outlet temperature and thereby improve the stability of the tobacco moisture content.

[0088] The device for determining the influence parameters of the wire drying outlet temperature provided in this embodiment of the invention can execute the method for determining the influence parameters of the wire drying outlet temperature provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0089] Example 4 Figure 6 A schematic diagram of an electronic device 410 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0090] like Figure 6 As shown, the electronic device 410 includes at least one processor 420 and a memory, such as a read-only memory (ROM) 430 or a random access memory (RAM) 440, communicatively connected to the at least one processor 420. The memory stores computer programs executable by the at least one processor. The processor 420 can perform various appropriate actions and processes based on the computer program stored in the ROM 430 or loaded into the RAM 440 from storage unit 490. The RAM 440 may also store various programs and data required for the operation of the electronic device 410. The processor 420, ROM 430, and RAM 440 are interconnected via a bus 450. An input / output (I / O) interface 460 is also connected to the bus 450.

[0091] Multiple components in electronic device 410 are connected to I / O interface 460, including: input unit 470, such as keyboard, mouse, etc.; output unit 480, such as various types of monitors, speakers, etc.; storage unit 490, such as disk, optical disk, etc.; and communication unit 4100, such as network card, modem, wireless transceiver, etc. Communication unit 4100 allows electronic device 410 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0092] Processor 420 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 420 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 420 performs the various methods and processes described above, such as the method for determining the influence parameters of the wire drying outlet temperature.

[0093] The method includes: Obtain the set of steady-state drying data corresponding to the target production batch in the target yarn processing workshop; A parameter correlation analysis was performed on the steady-state drying data set, and the basic influencing parameters corresponding to the target yarn-making workshop were determined based on the parameter correlation analysis results corresponding to each target production batch. The standard deviation of the outlet temperature of the steady-state drying data set is calculated based on a preset sliding window, and the label is divided based on the standard deviation calculation result to obtain the drying outlet temperature fluctuation dataset corresponding to the target production batch. The degree of influence of the basic influence parameters on the global yarn drying outlet temperature fluctuation dataset is calculated based on the preset lag step size and target lag time, and the candidate influence parameters corresponding to the target yarn making workshop are determined based on the degree of influence of the variables. The steady-state yarn drying data set is divided based on the candidate influence parameters to obtain a first data subset containing the candidate influence parameters and a second data subset not containing the candidate influence parameters. The exit temperature of the first data subset and the second data subset is predicted based on the target machine learning model. The target influence parameters corresponding to the target yarn making workshop are determined based on the prediction accuracy of the prediction results.

[0094] In some embodiments, the method for determining the influence parameter of the wire drying outlet temperature can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 490. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 410 via ROM 430 and / or communication unit 4100. When the computer program is loaded into RAM 440 and executed by processor 420, one or more steps of the method for determining the influence parameter of the wire drying outlet temperature described above can be performed. Alternatively, in other embodiments, processor 420 can be configured to perform the method for determining the influence parameter of the wire drying outlet temperature by any other suitable means (e.g., by means of firmware).

[0095] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0096] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0097] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0098] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0099] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0100] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0101] This application also discloses a computer program product, which includes a computer program that, when executed by a processor, implements the method for determining the influence parameter of the wire drying outlet temperature provided in any embodiment of this application. This program product and the method for determining the influence parameter of the wire drying outlet temperature disclosed in the embodiments of this application belong to the same inventive concept, and therefore will not be described in detail here.

[0102] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0103] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for determining the parameters affecting the outlet temperature of the drying wire, characterized in that, include: Obtain the set of steady-state drying data corresponding to the target production batch in the target yarn processing workshop; A parameter correlation analysis was performed on the steady-state drying data set, and the basic influencing parameters corresponding to the target yarn-making workshop were determined based on the parameter correlation analysis results corresponding to each target production batch. The standard deviation of the outlet temperature of the steady-state drying data set is calculated based on a preset sliding window, and the label is divided based on the standard deviation calculation result to obtain the drying outlet temperature fluctuation dataset corresponding to the target production batch. The degree of influence of the basic influence parameters on the global yarn drying outlet temperature fluctuation dataset is calculated based on the preset lag step size and target lag time, and the candidate influence parameters corresponding to the target yarn making workshop are determined based on the degree of influence of the variables. The steady-state yarn drying data set is divided based on the candidate influence parameters to obtain a first data subset containing the candidate influence parameters and a second data subset not containing the candidate influence parameters. The exit temperature of the first data subset and the second data subset is predicted based on the target machine learning model. The target influence parameters corresponding to the target yarn making workshop are determined based on the prediction accuracy of the prediction results.

2. The method according to claim 1, characterized in that, The parameter correlation analysis of the steady-state wire drying data set includes: Calculate the correlation coefficient between the target wire drying parameters and the corresponding wire drying outlet temperature in the steady-state wire drying data set; Based on the correlation coefficient and the preset significance check, the significance coefficient corresponding to the target wire drying parameter is determined; The correlation coefficient is judged based on a preset correlation threshold to obtain a first threshold judgment result, and the significance coefficient is judged based on a preset significance threshold to obtain a second threshold judgment result. Based on the first threshold judgment result and the second threshold judgment result, the target wire drying parameter is marked as the original influencing parameter, and the original influencing parameter is used as the parameter correlation analysis result.

3. The method according to claim 2, characterized in that, The determination of the basic influencing parameters corresponding to the target yarn-making workshop based on the parameter correlation analysis results for each target production batch includes: The frequency of occurrence of the original influencing parameters in each steady-state wire drying data set is summarized and processed; A third threshold judgment result is obtained by performing a threshold judgment on the occurrence frequency based on a preset frequency threshold. Based on the third threshold judgment result, the original influence parameter is determined as the basic influence parameter corresponding to the target yarn-making workshop.

4. The method according to claim 1, characterized in that, The data set of temperature fluctuations at the wire drying outlet includes: a subset of data during stable periods and a subset of data during periods of high fluctuations; The calculation of the influence of the basic influence parameters on the global drying wire outlet temperature fluctuation dataset based on the preset lag step size and target lag time includes: The first causal strength of the basic influence parameter in the global stationary period data subset is calculated based on the preset lag step size and target lag time. The second causal strength of the basic influence parameter in the global high-fluctuation period data subset is calculated based on the preset lag step size and target lag time. Based on the calculation result of the difference between the first causal intensity and the second causal intensity, the degree of influence of the basic influence parameter on the variable in the global wire drying outlet temperature fluctuation dataset is determined.

5. The method according to claim 1, characterized in that, The acquisition of the steady-state drying data set corresponding to the target production batch in the target yarn-making workshop includes: Obtain the set of original yarn drying data corresponding to the original production batch in the target yarn processing workshop; Based on the original yarn drying data set, batch data of the original production batch is selected to determine the target production batch corresponding to the target yarn making workshop; Based on the cumulative amount measured by the electronic scale, the original drying data set corresponding to the target production batch is cleaned by removing the initial and later data from the original drying data set, thereby obtaining the steady-state drying data set corresponding to the target production batch in the target yarn-making workshop.

6. The method according to claim 1, characterized in that, The step of predicting the outlet temperature of the first and second data subsets based on the target machine learning model, and determining the target influence parameters corresponding to the target spinning workshop based on the prediction accuracy of the prediction results, includes: Based on the target machine learning model, the outlet temperature of the first data subset is predicted to obtain a first prediction result, and based on the target machine learning model, the outlet temperature of the second data subset is predicted to obtain a second prediction result. The prediction accuracy of the first prediction result is calculated based on the true values ​​in the first data subset to obtain the first prediction accuracy, and the prediction accuracy of the second prediction result is calculated based on the true values ​​in the second data subset to obtain the second prediction accuracy. The first prediction accuracy and the second prediction accuracy are numerically compared, and the candidate influence parameters are used as the target influence parameters corresponding to the target yarn-making workshop based on the numerical comparison results.

7. The method according to claim 1, characterized in that, After predicting the outlet temperature of the first and second data subsets based on the target machine learning model, and determining the target influence parameters corresponding to the target spinning workshop based on the prediction accuracy of the prediction results, the method further includes: During the production process, the fluctuations of the parameters corresponding to the target influencing parameters are monitored in real time. When the temperature at the wire drying outlet fluctuates, the target influencing parameters are checked first. Based on the fluctuation characteristics of the target influencing parameters, the corresponding control parameters or control logic are adjusted to stabilize the drying outlet temperature and thus improve the stability of the tobacco moisture content.

8. A device for determining the parameters affecting the outlet temperature of a wire drying machine, characterized in that, include: The data acquisition module is used to acquire the set of steady-state drying data corresponding to the target production batch in the target yarn processing workshop; The correlation analysis module is used to perform parameter correlation analysis on the steady-state drying data set, and determine the basic influence parameters corresponding to the target yarn-making workshop based on the parameter correlation analysis results corresponding to each target production batch. The labeling module is used to calculate the standard deviation of the outlet temperature of the steady-state drying data set based on a preset sliding window, and to label the steady-state drying data set based on the standard deviation calculation result, so as to obtain the drying outlet temperature fluctuation dataset corresponding to the target production batch. The first determining module is used to calculate the degree of influence of the basic influence parameters on the global drying outlet temperature fluctuation dataset based on the preset lag step size and target lag time, and to determine the candidate influence parameters corresponding to the target yarn making workshop based on the degree of influence of the variables. The second determining module is used to divide the steady-state drying data set based on the candidate influence parameters to obtain a first data subset containing the candidate influence parameters and a second data subset not containing the candidate influence parameters, and to predict the outlet temperature of the first data subset and the second data subset based on the target machine learning model, and to determine the target influence parameters corresponding to the target yarn making workshop based on the prediction accuracy of the prediction results.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method for determining the influence parameters of the drying wire outlet temperature as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method for determining the influence parameters of the wire drying outlet temperature as described in any one of claims 1-7.

Images