A platform and method for moisture management in a cut tobacco process based on an excel server
By using an Excel server-based moisture management platform for tobacco processing, regression prediction models and trend analysis charts were employed to solve the instability problem of moisture control in the tobacco processing process. This enabled low-cost, easily scalable moisture management, improving product stability and energy efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA TOBACCO JIANGXI IND CO LTD
- Filing Date
- 2026-02-03
- Publication Date
- 2026-06-19
AI Technical Summary
In the tobacco processing process, it is difficult to achieve constant temperature and humidity for moisture control. Seasonal and weather changes cause variations in the appropriate moisture levels between processes. Traditional methods rely on experience and lack systematic platform support, making them difficult to promote in small and medium-sized workshops.
A moisture management platform for the silk refining process was built based on an Excel server. By collecting key parameters, a regression prediction model was constructed, and linear regression or random forest algorithms were used to predict moisture values. Combined with trend analysis charts, process parameters were adjusted to achieve dynamic moisture control.
It enables low-cost and easily promoted moisture management, significantly improves product stability, shortens process optimization cycle, saves energy costs, and reduces training costs for new employees.
Smart Images
Figure CN122241045A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of tobacco processing management technology, and in particular to a moisture management platform and method for tobacco processing based on an Excel server. Background Technology
[0002] In the tobacco processing process, moisture control is the core link to ensure the internal quality of cigarettes. The stability of the loosening and rehydration water ratio, the moisture content at the secondary feeding outlet, and the moisture content at the drying inlet directly affects the control effect of the temperature of the thin plate drying cylinder and the temperature of the airflow drying hot air, which in turn determines the combustibility, taste, and physical properties of cigarettes (such as whole tobacco ratio and filling value).
[0003] Currently, most cigarette manufacturing workshops struggle to maintain constant temperature and humidity. Seasonal changes (such as high temperature and humidity in summer and low temperature and humidity in winter), weather fluctuations (sunny / cloudy / rainy), and batch-to-batch variations all lead to changes in optimal moisture levels between processes. Traditional moisture control relies on the experience and judgment of process technicians and simple statistical analysis, which cannot accurately address complex environmental changes. Although some companies have attempted to apply big data technology, it is still in its early stages, lacking systematic platform support, and is costly and complex to develop, making it difficult to promote in small and medium-sized workshops. Summary of the Invention
[0004] In view of the problems existing in the current Excel server-based moisture management platform for silk refining processes, this invention is proposed. Therefore, the problem to be solved by this invention is how to provide an Excel server-based moisture management platform and method for silk refining processes.
[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:
[0006] In a first aspect, the present invention provides a method for moisture management in silk-making process based on an Excel server, which includes: collecting key parameters in the production process from the centralized control system of the silk-making workshop and performing preprocessing, and building a moisture management platform for silk-making process based on an Excel server;
[0007] Using preprocessed historical key parameters as samples, the proportion of loose rehydration water addition, workshop temperature, workshop humidity and season were selected as input variables, and the moisture content at the secondary feeding outlet and the moisture content after drying were selected as output variables. A moisture prediction model was constructed using a regression prediction algorithm, and the moisture prediction model was validated and the model parameters were adjusted to meet the prediction error.
[0008] Based on real-time production parameters input by the user or data retrieved from the centralized control system, the moisture prediction model is invoked to obtain the moisture prediction value. Combined with the generated key parameter trend analysis chart, the silk-making process parameters are adjusted to control moisture.
[0009] As a preferred embodiment of the Excel server-based moisture management method for silk-making process described in this invention, the key parameters include: feeding date, cigarette brand, weather conditions, workshop temperature, workshop humidity, loose rehydration water addition ratio, loose rehydration steam addition ratio, moisture at the secondary feeding outlet, moisture at the SIROX inlet, thin plate cylinder wall temperature or airflow drying process gas temperature, moisture after drying, moisture after flavoring, whole silk rate, broken silk rate, and filling value.
[0010] As a preferred embodiment of the silk-making process moisture management method based on an Excel server according to the present invention, the method for constructing a silk-making process moisture management platform based on an Excel server includes:
[0011] Configure data entry, query, and management permissions for different user roles using the Excel server's permission management module;
[0012] The inter-table formula module is used to establish a data connection between the basic data entry module and the historical moisture data module to synchronize and store the entered data, and to configure the data entry review process and the query result export process.
[0013] The moisture query and analysis module supports setting query conditions by tobacco brand, production month, weather type, and workshop temperature range, and can generate key parameter trend analysis charts based on the query results.
[0014] As a preferred embodiment of the moisture management method for silk refining process based on an Excel server described in this invention, the construction of the moisture prediction model includes:
[0015] A structured dataset was formed using preprocessed historical key parameters as samples. The proportion of loose rehydration water addition, workshop temperature, workshop humidity and season were selected as input variables, and the moisture content at the secondary feeding outlet and the moisture content after drying were selected as output variables.
[0016] The regression prediction algorithm used is either linear regression or random forest. The structured dataset is substituted into the algorithm, and the algorithm learns the mapping relationship between the input variables and the output variables to determine the model parameters.
[0017] When using the linear regression algorithm, the model is expressed as a multiple linear regression equation concerning the moisture content at the secondary feed outlet and the moisture content after drying, as follows:
[0018] ;
[0019] ;
[0020] in: To predict the moisture content at the outlet of the secondary feed, To predict the moisture content after drying, The water ratio is adjusted to allow the loose material to rehydrate. The workshop temperature; Humidity in the workshop; For the season; , These are the regression coefficients for each input variable; and For constant terms;
[0021] When using the random forest algorithm, the samples are trained using multiple decision trees, and the final predicted value is obtained by ensemble decision tree output.
[0022] As a preferred embodiment of the moisture management method for silk refining process based on an Excel server described in this invention, the verification of the moisture prediction model includes:
[0023] 20% of the data in the historical dataset was selected as the validation set, and the relative error between the predicted value and the actual production value of the moisture prediction model was calculated.
[0024] If the relative error is greater than the predetermined value, adjust the weights of the model input variables or change the prediction algorithm, and retrain and validate.
[0025] The conditions for updating the moisture prediction model include: regular quarterly updates, the commissioning of new tobacco brands, continuous exceeding of prediction error limits, and the occurrence of climate events or changes in production processes.
[0026] As a preferred embodiment of the moisture management method for silk refining process based on an Excel server described in this invention, the adjustment of silk refining process parameters includes:
[0027] Based on the trend analysis chart of key parameters, determine the variation law of the ratio of loose rehydration water added during silk refining;
[0028] Based on the predicted values output by the moisture prediction model, the deviation of the current parameters from the standard range, and the changing patterns, the ratio of water added during loose rehydration, the ratio of steam added during loose rehydration, the moisture setting value at the secondary feeding outlet, and the moisture setting value at the drying inlet are dynamically adjusted.
[0029] Record the adjusted production data and quality indicators and feed them back to the platform for subsequent model optimization and adjustment.
[0030] Secondly, the present invention provides a moisture management platform for silk refining process based on an Excel server, comprising:
[0031] Excel server infrastructure is used to provide user permission management, inter-table data association, and workflow configuration functions;
[0032] The basic data entry module is used to input or receive key parameters of silk production from the centralized control system.
[0033] The moisture history data module is used to store historical data of key parameters by cigarette brand and production batch, and the storage format is a structured Excel spreadsheet.
[0034] The moisture query and analysis module is used to receive user query conditions, retrieve data from the historical moisture data module, call the preset moisture prediction model to output the moisture prediction value, and generate a trend analysis chart of key parameters.
[0035] Thirdly, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a moisture management method for silk refining process based on an Excel server.
[0036] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements the steps of a moisture management method for a silk-making process based on an Excel server.
[0037] The beneficial effects of this invention are as follows: Based on no-code development using an Excel server, this invention is characterized by low cost and ease of promotion. Small and medium-sized workshops do not need to bear high IT development costs, and process personnel can get started with minimal training, effectively solving the problems of complex operation and difficulty in implementation of traditional solutions. By quantifying environmental variables and using linear regression modeling, product stability can be significantly improved. Automatic data synchronization and rapid analysis are achieved, shortening the process optimization cycle. Simultaneously, energy costs are saved by dynamically adjusting energy consumption parameters, and the training costs for new employees are significantly reduced. Attached Figure Description
[0038] To more clearly illustrate the technical solutions of 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.
[0039] Figure 1 This is a structural diagram of a moisture management platform for silk refining process based on an Excel server. Detailed Implementation
[0040] To make the above-mentioned objects, features, and advantages of the present invention more readily understood, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. 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 protection scope of the present invention.
[0041] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.
[0042] Secondly, the term "one embodiment" or "example" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the invention. An embodiment appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment that selectively excludes other embodiments.
[0043] Reference Figure 1 This is the first embodiment of the present invention, which provides a method for moisture management in a silk-making process based on an Excel server, including:
[0044] Step S1, Data Acquisition and Preprocessing: Collect key parameters of the silk production process from the centralized control system of the silk production workshop;
[0045] Key parameters include feeding date, cigarette brand, weather conditions during drying, workshop temperature during drying, workshop humidity during drying, water addition ratio for loose rehydration, steam addition ratio for loose rehydration, average moisture content at the secondary feeding outlet, average moisture content at the SIROX inlet, thin plate cylinder wall temperature or airflow drying process gas temperature, average moisture content after drying, average moisture content after flavoring, whole fiber rate, broken fiber rate, and filling value.
[0046] In the data acquisition and preprocessing process, the key parameters collected are preprocessed: outliers are removed, and the data is categorized and summarized by cigarette brand and production batch to form a structured dataset.
[0047] Step S2, Platform Setup: Build a moisture management platform for the silk refining process based on an Excel server;
[0048] Server alternative: The Excel server can be replaced by the WPS server, which also supports table operations, no-code development (such as form design and inter-table formulas), and permission management. It is also compatible with Excel files and can achieve the same data entry, storage and analysis functions. Only the login port (WPS built-in server entry) needs to be adjusted, which does not affect the purpose of the invention (low cost and easy operation).
[0049] Step S21: Configure user roles and operation permissions using the Excel server's permission module;
[0050] Step S22: Use the inter-table formula module to establish the connection between the basic data entry module and the historical moisture data module, so that the data can be automatically synchronized to the historical database after entry.
[0051] Step S23: Configure the data entry and review process and the query result export process using the workflow module;
[0052] Step S24: Deploy the moisture query and analysis module, set query condition options and trend chart generation function.
[0053] Step S3: Establishment and validation of the moisture prediction model;
[0054] Step S31: Using the historical key parameters after preprocessing in Step S1 as samples, select the proportion of loose rehydration and water addition, workshop temperature and humidity, and season as input variables, and the moisture content at the secondary feeding outlet and the moisture content after drying as output variables. Use a linear regression algorithm to construct a moisture prediction model.
[0055] Alternative prediction algorithms: The linear regression algorithm can be replaced by the random forest algorithm, which can handle nonlinear relationships (such as the nonlinear effect of extreme high temperatures on moisture), does not require the assumption of linear correlation between variables, and has a wider range of applicable scenarios; the verification method remains the same (the error is still calculated using 20% of historical data), and the same accuracy of ≤0.5% can be achieved, and the model optimization logic (adjusting feature weights) is consistent;
[0056] The specific steps involved in constructing a moisture prediction model using a linear regression algorithm include:
[0057] Data preparation: Using the historical key parameters after preprocessing in step S1 as samples, the preprocessing includes: removing outliers (such as parameters that are outside the normal range), classifying and summarizing by cigarette brand and production batch to form a structured dataset (the samples need to cover all four seasons and different weather conditions to ensure representativeness).
[0058] Variable determination: Input variables (influencing factors): loose rehydration water ratio, workshop temperature, workshop humidity, season (quantified into numerical values, such as 1=spring, 2=summer, etc.); Output variables (prediction targets): moisture content at the secondary feeding outlet, moisture content after drying.
[0059] Algorithm selection: Core algorithm: Linear regression algorithm (suitable for scenarios with linear relationships between variables, easy to operate and highly interpretable); Alternative algorithm: Random forest algorithm (adapts to nonlinear relationships, such as the nonlinear effect of extreme high temperatures on water, does not require the assumption of linear relationships between variables, and has a wider range of applicable scenarios).
[0060] Model training: Substitute the structured dataset into the selected algorithm, and the algorithm will automatically learn the mapping relationship between the input variables and the output variables to determine the model parameters (such as the coefficients of linear regression and the decision tree structure of random forest).
[0061] Validation and optimization: Select 20% of historical data as the validation set, compare the predicted values output by the model with the actual production values, and calculate the error; if the error is >0.5%, adjust the weights of the model input variables (such as the coefficients of linear regression and the feature importance of random forest), and repeat the validation steps until the error is ≤0.5% (preset acceptable range).
[0062] The model is a multiple linear regression equation (since there are multiple input variables and two output variables, corresponding to two separate equations). The prediction expressions for the moisture content at the secondary feeding outlet and the moisture content after drying are as follows:
[0063] ;
[0064] ;
[0065] in: To predict the moisture content at the outlet of the secondary feed, To predict the moisture content after drying, Water addition ratio (%) for loosening and rehydration; Workshop temperature (°C); Workshop humidity (%); For the season (quantified value, such as 1=spring, 2=summer, 3=autumn, 4=winter); , These are the regression coefficients of each input variable (obtained through training with historical data, reflecting the weight of the variable's influence on the output). and This is the constant term (intercept of the regression equation).
[0066] Random forests do not have a fixed linear expression. Their core logic is as follows: samples are trained based on multiple independent decision trees, and each decision tree outputs a prediction result; the results of all decision trees are integrated through voting (for classification problems) or averaging (for regression problems) to obtain the final prediction value; the model optimization logic is consistent with linear regression, and the accuracy is improved by adjusting the feature weights, and the verification method is the same.
[0067] The prediction will be carried out after the moisture management platform for the silk-making process is built and the model is validated. The entire process will rely on the platform, and the steps are as follows:
[0068] Users can log in to the platform through the client and manually enter the key production parameters for the day (input the real-time data corresponding to the variables, such as the temperature and humidity of the workshop, the season, and the ratio of loose rehydration to water), or directly retrieve the data automatically collected by the centralized control system of the yarn making workshop.
[0069] Users can set query conditions (such as cigarette brand, production batch) through the moisture query and analysis module, which will trigger the platform to automatically call the optimized moisture prediction model (linear regression or random forest).
[0070] Substitute the input real-time key parameters into the model expression (linear regression equation) or random forest algorithm;
[0071] Using a preset coefficient / decision tree structure, the predicted moisture content at the secondary feeding outlet and the predicted moisture content after drying are calculated.
[0072] Generate key parameter trend analysis charts (such as the historical moisture change trend of cigarette brands, and the correlation curve between input and output variables); simultaneously output annual key parameter change tables to assist process technicians in understanding the parameter change patterns (such as the need for a higher water addition ratio in summer when the air is loose and moist compared to winter).
[0073] Based on the predicted values and trend analysis results, the process technicians adjust the daily yarn-making process parameters (such as adjusting the amount of water and steam) to achieve moisture control.
[0074] Step S32: Select a portion of historical data as a validation set, compare the predicted values output by the model with the actual values in production, and calculate the error;
[0075] Step S33: If the error exceeds 0.5%, adjust the weights of the model input variables and repeat step S32 until the error meets the requirements.
[0076] (I) Adjustment of Moisture Prediction Model (Adjustment of Core Technology)
[0077] Regular triggers: once per quarter (incorporating the latest production data); Special triggers: new cigarette brands put into production, forecast error exceeding 0.5%, extreme weather (such as abnormally high / low temperatures), major changes in production processes.
[0078] The adjustment operations specifically include:
[0079] Retrieve new production data (categorized by cigarette brand and batch) from the platform's historical data module, remove outliers, and update the structured dataset;
[0080] 20% of the newly added data was selected as the validation set. The model predictions were compared with the actual production values, and the relative error was calculated.
[0081] If the error is ≤0.5%, only update the dataset; if the error exceeds 0.5%, adjust the weights of the input variables (such as the influence weights of season and workshop temperature and humidity); if there is a non-linear relationship (such as the influence of extreme high temperature on moisture), replace it with the random forest algorithm.
[0082] Repeat the comparison between predicted and actual values until the error is ≤0.5%, then update the platform's built-in model.
[0083] (ii) Dynamic adjustment of yarn-making process parameters;
[0084] Key adjustment items: loose moisture rehydration water ratio (7.9%-10.3%), secondary feeding outlet moisture (20.0%-22.5%), and drying inlet moisture; auxiliary adjustment items: loose moisture rehydration steam ratio, thin plate cylinder wall temperature / airflow drying process gas temperature.
[0085] The adjustment operations specifically include:
[0086] Log in to the platform and enter the key production parameters for the day (cigarette label, weather, workshop temperature and humidity, etc.) or directly retrieve the data automatically collected by the centralized control system;
[0087] Through the moisture query and analysis module, query conditions can be set by cigarette brand + production month + weather + workshop temperature range to obtain historical data, moisture forecast values and key parameter trend charts.
[0088] Based on the trend chart, the core pattern is confirmed: when the temperature of the thin plate cylinder wall / airflow drying temperature is stable, the proportion of loose re-moistening water increases with the season (weather influence is not significant), refer to the annual key parameter change table;
[0089] Process technicians adjust daily parameters based on forecast values and seasonal patterns (e.g., increasing the proportion of water added for loosening and rehydration in summer, and decreasing it in winter).
[0090] Record the adjusted moisture content, whole fiber rate, and broken fiber rate after drying, and synchronize them to the platform's historical data module as a basis for subsequent adjustments.
[0091] Step S4: Platform application and process optimization;
[0092] Step S41: The user logs into the platform through the client;
[0093] Step S42: Enter the key parameters of the day's production through the basic data entry module, or retrieve the data automatically collected by the centralized control system;
[0094] Step S43: Set query conditions through the moisture query and analysis module to obtain historical data and predicted values;
[0095] Step S44: Based on the trend analysis chart, explore the changing patterns of key parameters: When the temperature of the thin plate cylinder wall / the gas temperature of the airflow drying process is stable, the proportion of loose rehydration and water addition increases with the season, and the weather has no significant impact on the parameters. Generate an annual key parameter change table to formulate the standard process parameter range for different seasons.
[0096] The process temperature is stable (the temperature of the thin plate cylinder wall / the temperature of the gas in the airflow drying process is stable). The core judgment criterion is that the temperature fluctuation range is controlled within the quantitative threshold.
[0097] The temperature fluctuation range of the thin plate cylinder wall / airflow drying process gas temperature during continuous production (taking the core process segment of one production batch, such as the entire drying process, usually 2-4 hours) is ≤ ±0.5℃.
[0098] The data is collected in real time from the centralized control system of the yarn processing workshop, with a sampling frequency of once every 5 minutes. All temperature data of the yarn drying process of this batch are recorded, and abnormal values caused by equipment failure (such as data with sudden jumps >2℃) are removed. The difference between the maximum and minimum temperature during this period is calculated. If the difference is ≤±0.5℃ and there are no three consecutive deviations exceeding the threshold, the temperature is judged to be stable.
[0099] Key parameters are stable (proportion of loose rehydration and water addition, moisture content at the outlet of secondary feeding, etc.). The core judgment criterion is that the parameter values are within the standard range and the fluctuation range is ≤ the threshold.
[0100] The ratio of loose, rehydrated material to water should be within the standard range of 7.9%-10.3% (upper limit for summer, lower limit for winter); the fluctuation range of the value for three consecutive production batches should be ≤±0.3%; data should be collected from the basic data entry module / centralized control system, with each batch recorded three times (initial, middle, and final stages of feeding), and the average value should be taken; check whether the average value of a single batch is within the standard range; calculate the maximum difference between the average values of three consecutive batches, and if it is ≤±0.3%, it is considered stable.
[0101] The moisture content at the outlet of the secondary feed should be within the standard range of 20.0%-22.5%; the fluctuation of the measured value within a single batch should be ≤±0.2% (referencing verification data with an error of ≤0.3% for 90% of batches); measured values should be collected once every 15 minutes, with at least 8 sets of data collected per batch; all measured values should be within the standard range; the difference between the maximum and minimum values within a single batch should be ≤±0.2%, which is considered stable.
[0102] After drying, the moisture content is within the standard range of 12.0%-13.0%; the average fluctuation of the average value of 5 consecutive production batches is ≤±0.1%; 3 sets of data (at different locations of the discharge port) are collected after each batch of production is completed, and the average value is taken; the average value of a single batch is within the standard range; the maximum difference of the average value of 5 consecutive batches is ≤±0.1%, which is considered stable.
[0103] II. Objective judgment conditions with insignificant impact (the impact of weather on key parameters)
[0104] Key criteria for judgment: Under different weather conditions in the same month, the parameter fluctuation range is ≤ the threshold and there is no statistical difference.
[0105] The effects of weather on the proportion of loose re-moistening water, the moisture content at the secondary feeding outlet, and the moisture content at the drying inlet were investigated. Production data for three weather types (sunny, cloudy, and rainy) were collected within the same calendar month (at least three production batches for each weather type). The average fluctuation range of the same parameter under different weather conditions was ≤ ±0.2%. Statistical tests were performed using analysis of variance (ANOVA), and the results showed a p-value > 0.05.
[0106] Batch data (taken from historical moisture data module) are stored according to weather type; the average value and fluctuation range of target parameters under different weather conditions in the same month are calculated; if the fluctuation range is ≤ ±0.2% and the P-value of the variance analysis is > 0.05, it is determined that the weather effect is not significant.
[0107] Step S45: The process technician adjusts the daily process parameters according to the pattern to achieve moisture control.
[0108] Specifically, the core objective is to control the error of key indicators such as moisture content at the secondary feeding outlet and moisture content after drying to ≤0.5%, ensuring the stability of physical indicators such as cigarette combustibility, flavor, whole tobacco yield, and filling value; the auxiliary objective is to reduce water consumption (target reduction of 58%) and steam consumption (target reduction of 35%) in the loosening and rehydration process while ensuring quality; the long-term objective is to avoid technical gaps caused by personnel turnover, achieve refined and intelligent management of the tobacco processing process, and comply with industry policy requirements.
[0109] The study focuses on key parameters affecting moisture content in the yarn-making process, specifically including: core control parameters: the ratio of loose rehydration water addition (7.9%-10.3%) and the ratio of loose rehydration steam addition; target control parameters: moisture content at the secondary feeding outlet (20.0%-22.5%), moisture content at the yarn drying inlet, and moisture content after drying (12.0%-13.0%); and auxiliary related parameters: the temperature of the thin-plate cylinder wall / the temperature of the gas in the airflow yarn drying process (which needs to be kept stable as a prerequisite for adjustment).
[0110] II. Core Adjustment Rules: These rules are not based on experience, but rather solidified through platform data closure and predictive models. The core rules include five categories:
[0111] Seasonal dominant rule (the most core rule); Prerequisite: The temperature of the thin plate cylinder wall / the gas temperature of the airflow drying process remains stable (excluding interference from the heating process).
[0112] Specific rules: The proportion of loose rehydration water increases positively with the season, i.e., it is higher in summer and lower in winter; the proportion of loose rehydration water is dynamically adjusted within the range of 7.9%-10.3% (higher value in summer and lower value in winter), and the moisture content at the outlet of the secondary feeding and the moisture content at the inlet of the drying wire are adapted to the season.
[0113] Environmental quantification rules: Quantify the real-time temperature and humidity of the workshop (standard range 20-35℃, 55%-70%) and weather conditions (sunny / cloudy / rainy) as input variables; calculate the influence coefficient of environmental factors on moisture through the prediction model, and output the predicted value, such as appropriately lowering the water addition ratio correction value under high temperature and high humidity environment.
[0114] Error closure rule; accuracy threshold: the relative error between the predicted model output value and the actual production value must be ≤0.5% (90% batch error ≤0.3%).
[0115] Adjustment logic: If the error exceeds the threshold, prioritize optimizing the weights of the model input variables (such as increasing the seasonal weights and fine-tuning the temperature and humidity weights), and after re-verifying that the target has been met, guide the adjustment of process parameters.
[0116] Data traceability rules; Horizontal comparison: Supports parameter comparison across production batches and different cigarette brands, extracting common adjustment patterns; Vertical traceability: Historical adjustment data of a single brand under different seasons and weather conditions is traceable, providing a reference for adding new cigarette brands or extreme environments;
[0117] Dynamic updates: The latest production data is incorporated into the model every quarter, and permission configurations are adjusted annually based on job roles to ensure that rules adapt to new scenarios.
[0118] Energy consumption optimization rules; constraints: parameter adjustments must be based on the premise that cigarette quality meets the standards (the whole tobacco rate, filling value, and other indicators meet the requirements); optimization direction: the adjustment direction should be tilted towards reducing energy consumption, for example, by controlling the water / steam ratio to avoid excessive consumption (target: water consumption reduced by 58%, steam consumption reduced by 35%).
[0119] The adjustment process relies on the Excel Server management platform, and the specific steps are as follows:
[0120] Step 1: Preliminary preparation (one-time deployment, long-term reuse);
[0121] The platform is built on an Excel server (or WPS server, compatible with Excel files), and the login port is configured (direct login for Excel, and adjustment of the built-in server entry for WPS). Roles are assigned through the server permission module (process technician: adjustment suggestion right; administrator: configuration right), and automatic recording of operation logs is enabled (in accordance with tobacco data security specifications).
[0122] Step 2: Data collection and preprocessing (daily / batch pre-processing);
[0123] Key parameters (feeding date, cigarette brand, weather, workshop temperature and humidity, water / steam ratio, moisture value at each stage, yarn yield, etc.) are automatically collected from the workshop's centralized control system or manually entered.
[0124] Outliers are automatically removed, and data is categorized and summarized by cigarette brand and production batch to form a structured Excel dataset. Historical data is backed up to local storage and the cloud every month to ensure traceability.
[0125] Step 3: Predictive model establishment and validation (generating core evidence);
[0126] Using preprocessed historical data as samples, a linear regression algorithm (or random forest algorithm, adapted to nonlinear scenarios) is used. Inputs include the proportion of loose rehydration water, workshop temperature and humidity, and season. Outputs include the moisture content at the secondary feeding outlet and the predicted moisture content after drying.
[0127] Use 20% of historical data as a validation set, compare the predicted values with the actual values, calculate the error, and if the error is >0.5%, adjust the weights of the input variables and repeat the validation until the error is ≤0.5%; incorporate the latest production data (such as newly added cigarette brands) into the model every quarter to optimize it.
[0128] Step 4: Query analysis and pattern discovery (clarify the direction of adjustment);
[0129] Log in via the computer's Excel server port or the built-in port of Excel spreadsheet, filter historical data by cigarette brand + production month (season) + weather + workshop temperature range, generate a key parameter trend analysis chart and an annual parameter change table, and clarify: the impact of season on parameters (e.g., the water ratio needs to be increased in summer); the degree of weather influence (the weather influence in the same month is not significant); and the deviation of the current parameters from the standard range.
[0130] Step 5: Implementation of process parameter adjustment (core execution step);
[0131] Input the real-time parameters for the day (cigarette label, season, workshop temperature and humidity, etc.), and the platform will output the predicted values of moisture content at the secondary feeding outlet and moisture content after drying.
[0132] Process technicians develop adjustment plans based on the following three points: trend chart patterns; model predictions; and current parameter deviations for the day.
[0133] During the production process, process parameters (water / steam ratio, moisture setting value) are modified to ensure that the adjusted parameters fall within the standard range. The actual moisture value and energy consumption data of each step after adjustment are tracked and fed back to the platform.
[0134] Step 6: Validation and Continuous Optimization (Closed-Loop Iteration);
[0135] The error between the adjusted actual value and the predicted value must be ≤0.5%;
[0136] Check whether indicators such as whole fiber rate, filling value, and odor absorption meet the standards;
[0137] Energy consumption verification: Analyze whether the water and steam consumption in the loosening and rehydration process has decreased;
[0138] Dynamic optimization: If the target is met, the data from this adjustment will be archived to the historical database for future reference; if the target is not met, the model will be re-optimized (variable weights will be adjusted), and the adjustment plan will be updated.
[0139] In the event of sudden extreme conditions (such as workshop temperatures exceeding 20-35℃), the adjustment plan will be adjusted based on similar historical cases, and the data will be incorporated into the model optimization for the next quarter.
[0140] Furthermore, this embodiment also provides a moisture management platform for the silk refining process based on an Excel server, including:
[0141] The Excel server infrastructure and three core functional modules deployed on the Excel server infrastructure include: basic data entry module, historical moisture data module, and moisture query and analysis module.
[0142] The Excel server infrastructure provides functions such as access control, inter-table data association, and workflow management. It also supports no-code development with Excel as the user interface and automatically completes the code conversion programming.
[0143] The moisture management platform for silk refining processes, based on an Excel server, offers two client login methods: one is to install the Excel server port on a computer, and the other is to log in directly through the Excel server port in an Excel spreadsheet. After logging in, users can perform operations such as data collection and entry, data querying, and analysis. During data querying and analysis, users can set query conditions according to their actual needs, and the system will return data results that meet the conditions, and support functions such as generating analytical trend charts to help users more intuitively understand the patterns behind the data.
[0144] The basic data entry module is used to input key parameters in the yarn production process.
[0145] The historical moisture data module is used to store historical production data of key parameters by cigarette brand and production batch. The storage format is an Excel structured table, which supports batch retrieval and comparison.
[0146] The moisture query and analysis module receives user-defined query conditions, returns key parameter data that meets the conditions, and outputs moisture prediction values based on a moisture prediction model built on historical key parameter data. It also generates a key parameter trend analysis chart. The moisture prediction model is constructed using statistical analysis algorithms and regression prediction algorithms based on historical key parameters.
[0147] This embodiment also provides a computer device applicable to a moisture management method for a silk-making process based on an Excel server, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement all or part of the steps of the method described in the above embodiments of the present invention.
[0148] This embodiment also provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, it performs the method in any optional implementation of the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.
[0149] The storage medium proposed in this embodiment and the data storage method proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.
[0150] In summary, this invention, developed using no-code Excel server technology, is characterized by low cost and ease of promotion. Small and medium-sized workshops do not need to bear high IT development costs, and process personnel can learn to use it after brief training, effectively solving the problems of complex operation and difficulty in implementation of traditional solutions. By quantifying environmental variables and using linear regression modeling, product stability can be significantly improved. Automatic data synchronization and rapid analysis are achieved, shortening the process optimization cycle. Simultaneously, energy costs are saved by dynamically adjusting energy consumption parameters, and training costs for new employees are significantly reduced.
[0151] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A method for moisture management in silk refining processes based on an Excel server, characterized in that: include, Key parameters from the production process are collected from the centralized control system of the silk-making workshop and preprocessed. A moisture management platform for the silk-making process is built based on an Excel server. Using preprocessed historical key parameters as samples, the proportion of loose rehydration water addition, workshop temperature, workshop humidity and season were selected as input variables, and the moisture content at the secondary feeding outlet and the moisture content after drying were selected as output variables. A moisture prediction model was constructed using a regression prediction algorithm, and the moisture prediction model was validated and the model parameters were adjusted to meet the prediction error. Based on real-time production parameters input by the user or data retrieved from the centralized control system, the moisture prediction model is invoked to obtain the moisture prediction value. Combined with the generated key parameter trend analysis chart, the silk-making process parameters are adjusted to control moisture.
2. The method for moisture management in silk refining process based on an Excel server as described in claim 1, characterized in that: The key parameters include feeding date, cigarette brand, weather conditions, workshop temperature, workshop humidity, loose rehydration water ratio, loose rehydration steam ratio, moisture content at the secondary feeding outlet, moisture content at the SIROX inlet, thin plate cylinder wall temperature or airflow drying process gas temperature, moisture content after drying, moisture content after flavoring, whole fiber rate, broken fiber rate, and filling value.
3. The method for moisture management in silk refining process based on an Excel server as described in claim 1, characterized in that: The moisture management platform for the silk refining process built on an Excel server includes: Configure data entry, query, and management permissions for different user roles using the Excel server's permission management module; The inter-table formula module is used to establish a data connection between the basic data entry module and the historical moisture data module to synchronize and store the entered data, and to configure the data entry review process and the query result export process. The moisture query and analysis module supports setting query conditions by tobacco brand, production month, weather type, and workshop temperature range, and can generate key parameter trend analysis charts based on the query results.
4. The method for moisture management in silk refining process based on an Excel server as described in claim 1, characterized in that: The construction of the moisture prediction model includes: A structured dataset was formed using preprocessed historical key parameters as samples. The proportion of loose rehydration water addition, workshop temperature, workshop humidity and season were selected as input variables, and the moisture content at the secondary feeding outlet and the moisture content after drying were selected as output variables. The regression prediction algorithm used is either linear regression or random forest. The structured dataset is substituted into the algorithm, and the algorithm learns the mapping relationship between the input variables and the output variables to determine the model parameters. When using the linear regression algorithm, the model is expressed as a multiple linear regression equation concerning the moisture content at the secondary feed outlet and the moisture content after drying, as follows: ； ； in: To predict the moisture content at the outlet of the secondary feed, To predict the moisture content after drying, The water ratio is adjusted to allow the loose material to rehydrate. The workshop temperature; Humidity in the workshop; For the season; , These are the regression coefficients for each input variable; and For constant terms; When using the random forest algorithm, the samples are trained using multiple decision trees, and the final predicted value is obtained by ensemble decision tree output.
5. The method for moisture management in silk refining process based on an Excel server as described in claim 4, characterized in that: The validation of the moisture prediction model includes: 20% of the data in the historical dataset was selected as the validation set, and the relative error between the predicted value and the actual production value of the moisture prediction model was calculated. If the relative error is greater than the predetermined value, adjust the weights of the model input variables or change the prediction algorithm, and retrain and validate. The conditions for updating the moisture prediction model include: regular quarterly updates, the commissioning of new tobacco brands, continuous exceeding of prediction error limits, and the occurrence of climate events or changes in production processes.
6. The method for moisture management in silk refining process based on an Excel server as described in claim 1, characterized in that: The adjustment of the fiber-making process parameters includes: Based on the trend analysis chart of key parameters, determine the variation law of the ratio of loose rehydration water added during silk refining; Based on the predicted values output by the moisture prediction model, the deviation of the current parameters from the standard range, and the changing patterns, the ratio of water added during loose rehydration, the ratio of steam added during loose rehydration, the moisture setting value at the secondary feeding outlet, and the moisture setting value at the drying inlet are dynamically adjusted. Record the adjusted production data and quality indicators and feed them back to the platform for subsequent model optimization and adjustment.
7. A moisture management platform for silk refining process based on an Excel server, based on the moisture management method for silk refining process based on an Excel server as described in any one of claims 1 to 6, characterized in that: include, Excel server infrastructure is used to provide user permission management, inter-table data association, and workflow configuration functions; The basic data entry module is used to input or receive key parameters of silk production from the centralized control system. The moisture history data module is used to store historical data of key parameters by cigarette brand and production batch, and the storage format is a structured Excel spreadsheet. The moisture query and analysis module is used to receive user query conditions, retrieve data from the historical moisture data module, call the preset moisture prediction model to output the moisture prediction value, and generate a trend analysis chart of key parameters.
8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the method for moisture management in silk-making process based on an Excel server as described in any one of claims 1 to 6.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the moisture management method for silk-making process based on an Excel server as described in any one of claims 1 to 6.