Model prediction-based automatic zero setting method for online moisture meter in cigarette cut tobacco production

Abstract

The invention provides an automatic zero setting method for an online moisture meter in cigarette cut tobacco production based on model prediction, and belongs to the technical field of cigarette production. The automatic zeroing method comprises the following steps: constructing a prediction model based on historical data of tobacco shred moisture in a cigarette shred making process; variable data in the cigarette cut tobacco making process are collected online in real time and input into the prediction model, and a predicted moisture value is obtained; the actual detection value of the tobacco shred moisture in the cigarette shred making process is collected online in real time; calculating a deviation value between the predicted moisture value and the actual detection value, and setting a deviation threshold value; comparing the deviation value with the deviation threshold value to obtain a comparison result, and performing zero setting on the online moisture meter according to the comparison result. Through the automatic zero setting mechanism, manual intervention can be reduced, the precision and stability of moisture measurement are improved, the production efficiency is enhanced, the maintenance cost is reduced, and production management and quality control are supported and optimized through real-time data.

Description

Technical Field

[0001] The invention relates to the technical field of cigarette production, and in particular to an automatic zeroing method for an online moisture meter in cigarette shreds based on model prediction. Background Art

[0002] In the process of making cigarettes, accurate measurement of tobacco moisture is one of the key factors to ensure product quality and an important parameter to reflect working conditions. At present, near-infrared spectroscopy is usually used to measure tobacco moisture in industrial sites. However, near-infrared spectroscopy still has some limitations in practical applications, especially in complex industrial environments with large fluctuations in temperature and humidity and dust interference, and the accuracy and stability of measurement are often difficult to guarantee. At the tobacco making site, near-infrared detectors often experience performance degradation and failure, and require manual regular calibration, generally sampling and testing every day and every brand, and adjusting the zero point if there is a large deviation. This manual calibration method is labor-intensive and inefficient, which not only increases the labor intensity and maintenance costs of operators, but also affects the continuity and efficiency of production. In addition, the existing near-infrared spectroscopy technology still has deficiencies in automation integration and long-term stability, and cannot fully meet the needs of modern industrial production. Although some patents and technical literature have proposed improvement measures, these are mostly focused on local optimization and lack a systematic and comprehensive solution.

[0003] In view of this, the inventor, based on many years of experience in production design in this field and related fields, has designed a method for automatic zeroing of an online moisture meter in cigarette making based on model prediction after repeated experiments, in order to solve the problems existing in the prior art. Summary of the invention

[0004] The object of the present invention is to provide an automatic zeroing method for an online moisture meter in cigarette shreds making based on model prediction, which can automatically zero the online moisture meter.

[0005] To achieve the above-mentioned purpose, the present invention proposes an automatic zeroing method for an online moisture meter in cigarette shreds based on model prediction, wherein the automatic zeroing method for an online moisture meter in cigarette shreds comprises:

[0006] Construct a prediction model based on historical data of moisture content in cut tobacco during cigarette making;

[0007] Collect variable data in the cigarette making process in real time online and input them into the prediction model to obtain the predicted moisture value;

[0008] Collect the actual detection value of tobacco moisture in the process of cigarette making online and in real time;

[0009] Calculate the deviation between the predicted moisture value and the actual detected value, and set a deviation threshold;

[0010] Compare the deviation value and the deviation threshold and obtain the comparison result, and adjust the online moisture meter to zero according to the comparison result.

[0011] The present invention also provides a computer-readable storage medium on which a computer program is stored, wherein the computer program implements the steps according to the above method when executed by a processor.

[0012] The present invention further provides a computer device, comprising a memory and a processor, wherein a computer program that can be run on the processor is stored in the memory, wherein the steps of the above method are implemented when the processor executes the computer program.

[0013] Compared with the prior art, the present invention has the following characteristics and advantages:

[0014] The invention proposes an automatic zeroing method for an online moisture meter in cigarette shreds based on model prediction, which significantly reduces the labor intensity of operators by introducing the automatic zeroing method for an online moisture meter based on model prediction. Compared with the traditional manual calibration method, automatic zeroing can reduce frequent manual intervention, reduce the workload of operators, improve the efficiency of the zeroing process, reduce production delays caused by waiting for manual calibration, enhance production continuity, and improve overall production efficiency.

[0015] The present invention proposes an automatic zeroing method for an online moisture meter in cigarette shred processing based on model prediction. By constructing a soft measurement prediction model and realizing real-time automatic calibration of the moisture meter, not only the accuracy and stability of moisture measurement are improved, but also the automation integration capability of the system is enhanced, so that the moisture measurement equipment can better adapt to the needs of modern industrial production, and a long-term stable moisture measurement solution is provided.

[0016] The present invention proposes an automatic zeroing method for an online moisture meter in cigarette shred processing based on model prediction. By comprehensively considering multiple aspects such as data preprocessing, feature engineering, model construction, model training and verification, a complete automatic zeroing system is constructed. The system can respond to changes in the shred processing process in real time, automatically adjust the zero point of the moisture meter, and ensure the accuracy of the measurement results. It overcomes the limitations of the existing technology and provides a new automated solution for moisture measurement in the cigarette shred processing process.

[0017] The present invention proposes an automatic zeroing method for an online moisture meter in tobacco shreds based on model prediction, which constructs a soft measurement prediction model between multiple process variables and tobacco shred moisture by applying an artificial intelligence algorithm. The soft measurement prediction model between multiple process variables and tobacco shred moisture constructed by the present invention can predict the moisture value in real time and compare it with the actual detection value. When the deviation between the predicted value and the detection value is within the set threshold, zeroing is automatically performed, thereby improving the accuracy and stability of moisture measurement, reducing measurement errors caused by improper manual calibration or environmental changes, and ensuring the consistency of product quality.

[0018] The present invention proposes an automatic zeroing method for an online moisture meter in cigarette shreds based on model prediction. The moisture zeroing information management system developed using the automatic zeroing method for an online moisture meter in cigarette shreds based on model prediction of the present invention can record and manage all zeroing-related data in real time, provide data analysis and statistical functions, help managers understand the working status of the moisture meter, and optimize the zeroing strategy. This data-driven method makes production management more scientific and accurate, and improves the level of product quality control. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings described herein are only for explanation purposes and are not intended to limit the scope of the present invention in any way. In addition, the shapes and proportional dimensions of the various components in the figures are only schematic, used to help understand the present invention, and are not specifically limited to the shapes and proportional dimensions of the various components of the present invention. Those skilled in the art can select various possible shapes and proportional dimensions to implement the present invention according to the teachings of the present invention.

[0020] Figure 1 It is a schematic flow chart of the automatic zeroing method of the online moisture meter in cigarette shreds making of the present invention;

[0021] Figure 2 A schematic diagram of the process of constructing a prediction model based on historical data of cut tobacco moisture in the present invention. DETAILED DESCRIPTION

[0022] The details of the present invention can be more clearly understood with reference to the accompanying drawings and the description of the specific embodiments of the present invention. However, the specific embodiments of the present invention described herein are only used for the purpose of explaining the present invention and cannot be understood as limiting the present invention in any way. Under the guidance of the present invention, technicians can conceive of any possible variations based on the present invention, which should all be considered to belong to the scope of the present invention.

[0023] The present invention proposes an automatic zeroing method for an online moisture meter in cigarette shreds based on model prediction, and an automatic zeroing method for an online moisture meter in cigarette shreds, such as Figure 1As shown, it includes: building a prediction model based on historical data of tobacco moisture in the cigarette making process; collecting variable data in the cigarette making process online in real time and inputting it into the prediction model to obtain a predicted moisture value; collecting the actual detection value of tobacco moisture in the cigarette making process online in real time; calculating the deviation between the predicted moisture value and the actual detection value, and setting a deviation threshold; comparing the deviation value and the deviation threshold to obtain a comparison result, and zeroing the online moisture meter according to the comparison result.

[0024] The present invention also provides a computer-readable storage medium on which a computer program is stored, wherein the computer program implements the steps according to the above method when executed by a processor.

[0025] The present invention also proposes a computer-readable storage medium on which a computer program is stored, which is based on a model-predicted automatic zeroing method for an online moisture meter in cigarette shred processing, and designs and develops a moisture zeroing information management system for collecting, storing, analyzing and displaying zeroing process data.

[0026] Data management: Develop an information management system to record and manage all zeroing related data. This system should include data collection, analysis, storage and display functions.

[0027] System functions:

[0028] Data recording: Record detailed information of each zero adjustment, including time, deviation value, zero adjustment result, etc.

[0029] Data analysis: Provides data analysis and statistical functions to help managers understand the working status of the moisture meter and optimize the zero adjustment strategy.

[0030] Alarm function: When the detector has an abnormality, the system can alarm in time to remind relevant personnel to deal with it.

[0031] Report generation: The system can automatically generate zero adjustment reports for easy traceability and review.

[0032] The present invention further provides a computer device, comprising a memory and a processor, wherein a computer program that can be run on the processor is stored in the memory, wherein the steps of the above method are implemented when the processor executes the computer program.

[0033] The invention proposes an automatic zeroing method for an online moisture meter in cigarette cutting based on model prediction, which reduces manual intervention through an automatic zeroing mechanism, improves the accuracy and stability of moisture measurement, enhances production efficiency, reduces maintenance costs, and optimizes production management and quality control through real-time data support.

[0034] In an optional embodiment of the present invention, the comparison result includes that the deviation value is within the deviation threshold or the deviation value is greater than the deviation threshold.

[0035] Specifically, the automatic zeroing method of the online moisture meter in the cigarette shreds based on model prediction in the present invention realizes efficient automatic zeroing of the moisture content of the tobacco shreds through accurate model prediction and real-time data collection. First, a prediction model is constructed based on historical data, and the model can predict the moisture content of the tobacco shreds according to the variable data in the shreds making process. The variable data collected online in real time are input into the prediction model to obtain the predicted moisture value. At the same time, the actual detection value of the moisture content of the tobacco shreds in the shreds making process is collected online in real time. By calculating the deviation value between the predicted moisture value and the actual detection value and comparing it with the preset deviation threshold, a comparison result can be obtained. This comparison result clearly indicates whether the deviation value is within the deviation threshold or greater than the deviation threshold, thereby triggering the corresponding zeroing action. When the deviation value is within the threshold, the system believes that the moisture meter is working normally and does not need to be zeroed; if the deviation value is greater than the threshold, the zeroing procedure is automatically triggered to ensure the measurement accuracy and stability of the moisture meter.

[0036] The automatic zeroing method of the online moisture meter of the present invention can significantly improve the measurement accuracy and stability of the online moisture meter. By comparing the deviation between the predicted value and the actual value in real time and making an intelligent decision on whether to perform zeroing based on the relationship between the deviation and the threshold, the need for manual calibration is reduced, the labor intensity of the operator is reduced, and the potential errors introduced by manual operation are reduced.

[0037] In an optional example of this implementation, zeroing is performed according to the comparison result, including: when the deviation value is within the deviation threshold, automatically zeroing the online moisture meter; when the deviation value is greater than the deviation threshold, generating a verification work order and pushing it to relevant personnel, prompting manual inspection and calibration.

[0038] In the automatic zeroing method of the online moisture meter of the present invention, the zeroing operation is performed according to the comparison result of the deviation value between the predicted moisture value and the actual detection value and the preset deviation threshold. Specifically, when the calculated deviation value is within the preset deviation threshold range, the system will automatically trigger the zeroing program of the online moisture meter without manual intervention, thereby realizing automatic zeroing. This automatic zeroing mechanism can quickly respond to measurement deviations, ensure accurate readings of the moisture meter, and reduce product quality problems caused by measurement errors.

[0039] In an optional example, automatic zeroing of an online moisture meter includes: calculating a zeroing step and a zeroing direction according to the size and direction of a deviation value; determining a zeroing time interval and a zeroing amplitude according to a dynamic response characteristic of the online moisture meter; and zeroing the online moisture meter based on the zeroing step, the zeroing direction, the zeroing time interval and the zeroing amplitude.

[0040] In the online moisture automatic zeroing method of the present invention, after detecting the deviation value, the system will intelligently calculate the required zeroing step and zeroing direction according to the size and direction of the deviation value. Specifically, the system will comprehensively consider the historical zeroing data of the moisture meter and the current working conditions, and dynamically optimize the zeroing parameters to achieve accurate zeroing. Based on the calculated zeroing step, zeroing direction, zeroing time interval and zeroing amplitude, the system will automatically zero the online moisture meter to ensure that the deviation between its measured value and the actual value remains within an acceptable range, thereby improving the accuracy and reliability of the measurement.

[0041] Furthermore, the automatic zeroing of the online moisture meter also includes: setting a zeroing dead zone, and not triggering the automatic zeroing when the deviation value is within a predetermined range.

[0042] Specifically, the automatic zeroing function requires the design of reasonable control algorithms and strategies. According to the size and direction of the deviation value, calculate the zeroing parameters, such as zeroing step length, zeroing direction, etc. Consider the dynamic response characteristics of the moisture meter, design a suitable zeroing time interval and zeroing amplitude to avoid over-adjustment or under-adjustment. Set a zeroing dead zone, and do not trigger zeroing when the deviation is within a certain range, reducing unnecessary zeroing operations. Add an adaptive mechanism to dynamically optimize the zeroing parameters based on historical zeroing data to improve zeroing efficiency and stability. By introducing the concept of zeroing dead zone, intelligent zeroing management of online moisture meters is realized, significantly improving the stability and efficiency of the system. Setting a zeroing dead zone can avoid zeroing operations that are frequently triggered due to small deviations, reduce unnecessary intervention in equipment, reduce the risk of equipment wear and errors, and reduce production interruptions that may be caused by frequent zeroing.

[0043] In an optional example of this implementation, the deviation threshold ranges from 0.2 to 0.3.

[0044] Specifically, when the deviation between the predicted value and the detected value is within the threshold, the moisture detector is considered to be working normally, and the automatic zeroing program is triggered. The software system calculates the zeroing parameters according to the deviation value and issues instructions to adjust the zero point of the moisture meter. When the deviation is greater than the threshold, a calibration work order is generated and pushed to relevant personnel, prompting manual inspection and calibration. The work order contains key information such as deviation value and timestamp, which is convenient for problem tracking and analysis. The deviation between the predicted value and the detected value must be 0.3. The empirically optimal value is 0.3, which can be 0.2-0.3, especially for materials with low moisture content, which have higher requirements. By accurately setting the deviation threshold within the range of 0.2 to 0.3, the accuracy and response speed of moisture measurement are improved, and unnecessary zeroing operations are reduced. By setting the threshold precisely and limiting the scope of zeroing operations, moisture measurement is made more accurate, which can reduce equipment wear and errors caused by frequent zeroing, improve production continuity, and reduce production interruptions caused by waiting for zeroing, thereby improving production efficiency.

[0045] In an optional embodiment of the present invention, the method for automatically zeroing an online moisture meter in cigarette shred processing further includes preprocessing historical data.

[0046] Specifically, historical data of the cigarette making process is collected, including process variables and moisture values ​​of tobacco at corresponding moments. Appropriate data preprocessing methods are selected to preprocess the historical data, including: missing value processing, outlier processing and noise processing.

[0047] The present invention proposes an automatic zeroing method for an online moisture meter in cigarette shred processing based on model prediction, which can achieve higher-precision moisture measurement and more reliable equipment zeroing by introducing a preprocessing step of historical data. The preprocessing step eliminates noise and outliers in the data, fills in missing values, and ensures the integrity and consistency of the data. It not only improves the accuracy of the prediction model, but also enhances the model's adaptability to changes in the shred processing process. Therefore, the preprocessing step is crucial to improving the accuracy and reliability of the automatic zeroing of the online moisture meter. It helps to reduce the zeroing error caused by data quality problems, improve the continuity and efficiency of production, and ultimately improve the quality and consistency of cigarette products. Through this method, we can ensure that the online moisture meter can provide accurate measurement results under various working conditions, providing a solid data foundation for the automation and intelligence of the cigarette shred processing process.

[0048] In an optional example of this implementation, preprocessing the historical data includes: filling missing values ​​in the historical data using linear interpolation; detecting and eliminating outliers in the historical data using the 3-σ principle; and smoothing the historical data using a moving average filter.

[0049] Missing value processing: For missing values ​​in the data, linear interpolation is used to fill them. For the missing value $x_i$ at the i-th moment, linear interpolation estimation is performed through the data $x_{i-1}$ and $x_{i+1}$ at the two moments before and after it:

[0050] $x_i=(1-t)*x_{i-1}+t*x_{i+1}$, where $t=(i-i_1) / (i_2-i_1)$, $i_1$ and $i_2$ are the time indices before and after the missing value respectively.

[0051] Outlier processing: The 3-σ principle is used to detect and eliminate outliers in the data. For each variable, its mean $μ$ and standard deviation $σ$ are calculated, and data outside the range of $μ±3σ$ are considered outliers and eliminated.

[0052] Noise processing: Use a moving average filter to smooth the data to remove high-frequency noise. For the data $x_i$ at the i-th moment, its smoothing value is the average value of the data at the k moments before and after it:

[0053] $x_i=(x_{ik / 2}+...+x_i+...+x_{i+k / 2}) / k$, where k is the size of the sliding window.

[0054] By using linear interpolation to fill missing values, 3-σ principle to remove outliers, and moving average filter to smooth historical data, the quality and efficiency of data preprocessing can be significantly improved, which not only ensures the integrity and consistency of data, but also improves the reliability of data and reduces the negative impact of outliers on model training. Through the preprocessing steps, the automatic zeroing method based on model prediction can be made more accurate and stable, the zeroing accuracy of the online moisture meter is improved, and the zeroing error caused by data quality problems is reduced, thereby improving the automation level of the cigarette wire making process and product quality.

[0055] In an optional embodiment of the present invention, a prediction model is constructed based on historical data of moisture content of cut tobacco in the cigarette making process, including: extracting time series features from the historical data and obtaining time series features; based on the time series features, using a support vector regression algorithm and a random forest regression algorithm to construct a prediction model; and using k-fold cross validation and leave-one-out method to evaluate the performance of the prediction model.

[0056] Specifically, Figure 2 As shown, in the present invention, a prediction model is constructed based on historical data of tobacco moisture in the cigarette shred process, including:

[0057] Model selection: Considering the high dimensionality and nonlinear characteristics of cigarette making process data, two machine learning algorithms, Support Vector Regression (SVR) and Random Forest Regression (RFR), were chosen to construct the prediction model.

[0058] Model training:

[0059] Support Vector Regression: The radial basis function (RBF) is used as the kernel function of SVR. The hyperparameters of SVR are optimized by grid search, including the penalty coefficient C, kernel function parameter γ, and insensitive loss ε. The SVR model is trained using the training set data, and the model performance is evaluated and hyperparameters are tuned using the validation set data.

[0060] Random Forest Regression: The grid search method is used to optimize the hyperparameters of RFR, including the number of decision trees, the maximum depth of each tree, the minimum number of samples per node, etc. The RFR model is trained using the training set data, and the validation set data is used to evaluate the model performance and tune the hyperparameters.

[0061] Model integration: The trained SVR model and RFR model are integrated by weighted average to improve the robustness and accuracy of the prediction. The integration weight can be determined by cross-validation or optimization method based on the performance of the validation set.

[0062] Model validation: K-fold cross validation and leave-one-out method are used to evaluate model performance. Root mean squared error (RMSE) and mean absolute error (MAE) are used to measure the prediction accuracy of the model. At the same time, independent test set data is used to verify the model application to ensure the reliability of the model in the actual production environment.

[0063] Model application: The trained prediction model is deployed to the online system, which receives process variable data in real time and outputs the prediction results of tobacco moisture value. The prediction results can be used to guide the process optimization and quality control of cigarette making process.

[0064] By adopting the support vector regression (SVR) algorithm and the random forest regression (RFR) algorithm, the generalization ability and prediction accuracy of the model can be comprehensively evaluated to ensure the reliability and effectiveness of the model in practical applications.

[0065] The present invention builds a prediction model based on the extracted time series features, and feature engineering includes: feature scaling, feature selection and time series feature extraction.

[0066] Feature scaling: Use the minimum-maximum standardization method to scale the data and scale the value range of each variable to the interval [0,1]. For the jth feature $x_{ij}$ of the i-th sample, its scaled value is:

[0067] $x_{ij}^{scaled}=(x_{ij}-min(x_j)) / (max(x_j)-min(x_j))$, where $min(x_j)$ and $max(x_j)$ are the minimum and maximum values ​​of the j-th feature respectively.

[0068] Feature selection: Recursive Feature Elimination (RFE) is used to screen features. RFE recursively trains the model and evaluates the importance of features, gradually eliminating features with low importance until the preset number of features or performance indicators are reached.

[0069] In an optional example of this implementation, time series feature extraction is performed on historical data, including: scaling the historical data using a minimum-maximum normalization method to scale the value range of each variable to the interval [0,1]; calculating the lag term or sliding window statistics of each variable and using the lag term or sliding window statistics as time series features.

[0070] Specifically, time series feature extraction: Based on the process knowledge of the cigarette making process, construct time series features that reflect dynamic characteristics. For example, calculate the lag terms of each process variable (such as the previous moment, the previous two moments, etc.) and sliding window statistics (such as sliding mean, sliding standard deviation, etc.) as new features. Time series features can capture the trend and pattern of variables changing over time in the making process, providing rich dynamic information for the prediction model, so that the model can more accurately predict the moisture value of tobacco.

[0071] The method for automatically zeroing an online moisture meter in cigarette shreds based on model prediction proposed by the present invention also includes a prediction comparison step. Specifically, process variable data is collected online in real time and input into the constructed prediction model to obtain the predicted moisture value of the detection site. At the same time, the actual detection value of the moisture detector is obtained. The deviation between the predicted value and the detection value is calculated, and a reasonable deviation threshold (such as 0.3) is set.

[0072] Online near-infrared moisture meters are installed at the entrance and exit of key processes. The installation position is required by the Cigarette Process Specifications. It should be installed within about 3.0m of the entrance or exit of the detected process equipment; for processes such as rehumidification and airflow drying where the outlet is a high-temperature material, it should be installed after the equipment discharge conveyor device is overlapped with another conveyor device. The detection probe should be perpendicular to the surface of the tobacco material to be detected, and the height from the probe to the material surface is about (250±100)mm. The material flow is continuous in the detection direction and the material thickness should be greater than 5.0cm.

[0073] In another optional embodiment of the present invention, a computer-readable storage medium storing a computer program is used to design and develop a moisture zeroing information management system for collecting, storing, analyzing and displaying zeroing process data based on a model-predicted automatic zeroing method for an online moisture meter in cigarette shred processing.

[0074] The system mainly consists of the following modules:

[0075] Data acquisition module

[0076] Function: Real-time collection of moisture meter detection values, process parameters and key data in the automatic zeroing process, such as zeroing time, deviation before and after zeroing, zeroing status, etc.

[0077] Connect with the online moisture detection system and automatic zeroing system to exchange data through communication protocols (such as OPC, Modbus, etc.).

[0078] Perform preliminary processing on the collected data, such as data format conversion, time synchronization, outlier filtering, etc.

[0079] The processed data is transferred to the data storage module.

[0080] Data storage module

[0081] Function: Store the collected data persistently to ensure data integrity and security.

[0082] Select a suitable database system (such as relational database, time series database, etc.) and design an optimized data table structure based on data characteristics.

[0083] Provides basic operation interfaces such as data writing, querying, updating and deleting for other modules to call.

[0084] Back up and archive data regularly to prevent data loss.

[0085] Data Analysis Module

[0086] Function: Analyze and mine the stored historical data to extract valuable information and rules.

[0087] Implement common statistical analysis functions, such as mean, variance, correlation, etc., and explore the performance of the moisture meter and the effect of automatic zeroing.

[0088] Develop trend prediction and anomaly detection algorithms to detect abnormal conditions of moisture meters early.

[0089] Optimize the automatic zeroing algorithm and dynamically adjust the zeroing parameters, such as zeroing threshold, zeroing step, etc., based on historical data.

[0090] Report generation module

[0091] Function: Generate various zero adjustment reports according to user needs and display key information intuitively.

[0092] Design flexible report templates to support customized report content and format.

[0093] Regularly and automatically generate common reports, such as daily, weekly, monthly reports, etc., to facilitate management personnel to understand the situation in a timely manner.

[0094] Provides report export function and supports multiple file formats, such as PDF, Excel, etc.

[0095] Visual display module

[0096] Function: Intuitively display moisture zeroing related data and analysis results through a graphical interface.

[0097] Design an interactive dashboard to display key indicators such as current moisture value, deviation value, zeroing status, etc. in real time.

[0098] The historical data is displayed in the form of trend charts, scatter plots, etc., which is convenient for users to analyze the performance of the moisture meter and the zero adjustment effect.

[0099] Providing data drilling and filtering functions, users can flexibly select the data range and dimensions of interest.

[0100] System Management Module

[0101] Function: Responsible for the system's basic functions such as user management, permission control, and configuration management.

[0102] Provide account management functions such as user registration, login, password modification, etc. to ensure system access security.

[0103] Set corresponding operation permissions for different user roles (such as operators, engineers, managers, etc.).

[0104] Manage various configuration parameters of the management system, such as data collection cycle, data storage strategy, alarm threshold, etc.

[0105] Connection relationship between modules:

[0106] The data collection module transmits the collected data to the data storage module for persistent storage.

[0107] The data analysis module reads historical data from the data storage module for analysis and writes the analysis results back to the database or passes them to other modules.

[0108] The report generation module generates various reports based on the data in the data storage module for users to review.

[0109] The visualization module reads data from the data storage module and calls the algorithm of the data analysis module for graphical display.

[0110] The system management module interacts with other modules to manage user permissions, system configuration, etc.

[0111] The logical relationship between the various parts is as follows:

[0112] The online moisture detection system and automatic zeroing system serve as data sources, generating moisture detection values, process parameters and zeroing data in real time. These data are transmitted to the data acquisition module through the communication protocol.

[0113] The data acquisition module is responsible for receiving data from the online moisture detection system and the automatic zeroing system, and performing preliminary processing, such as data format conversion, time synchronization, etc. The processed data is transmitted to the data storage module.

[0114] The data storage module stores the received data persistently in the database. It provides data reading and writing interfaces for other modules and is the data center of the entire system.

[0115] The data analysis module reads historical data from the data storage module, performs statistical analysis, trend prediction, anomaly detection, etc., and mines valuable information from the data. The analysis results can be written back to the database or directly passed to the visualization display module.

[0116] The report generation module generates various zero adjustment reports according to the data in the data storage module and the predefined report template. The report can be generated automatically or manually triggered by the user, and the generated report can be viewed and exported by the user.

[0117] The visualization module is responsible for presenting data and analysis results to users in a graphical way. It reads data from the data storage module, calls the algorithm of the data analysis module for processing, and finally displays it in the form of dashboards, trend charts, etc.

[0118] The system management module plays a role in management and configuration of the entire system. It manages user accounts and permissions and controls the access rights of each module. At the same time, it is also responsible for managing various configuration parameters of the system, such as data collection cycle, data storage strategy, etc.

[0119] There is also interaction between the data analysis module and the visualization module. The analysis results of the data analysis module can be directly passed to the visualization module for display; at the same time, the visualization module can also call the algorithm of the data analysis module to process and display the data in real time.

[0120] The automatic zeroing method of the online moisture meter in the cigarette shreds processing based on model prediction proposed by the present invention also includes system integration and testing. The various modules are integrated into a complete automatic zeroing system, including data flow, control logic and user interaction. Functional testing, performance testing and stability testing are carried out in a test environment to verify the reliability of the system. Necessary optimization and improvement are carried out according to the test results, such as adjusting control parameters and optimizing database performance.

[0121] The automatic zeroing method of online moisture meter in cigarette shreds based on model prediction proposed in this invention is to deploy the system to the actual cigarette shreds production line for on-site trial operation and verification. Collect operation data over a period of time to evaluate the zeroing effect and stability of the system. Compare the frequency and workload of manual calibration to quantify the benefits of the automatic zeroing system. Summarize the lessons learned and continuously optimize and improve the system functions.

[0122] The detailed explanation of the above-mentioned embodiments is only intended to explain the present invention so as to facilitate a better understanding of the present invention. However, these descriptions cannot be interpreted as limitations on the present invention for any reason. In particular, the various features described in different embodiments may also be arbitrarily combined with each other to form other embodiments. Unless there is a clear description to the contrary, these features should be understood to be applicable to any embodiment and are not limited to the described embodiments.

Claims

1. A method for automatically adjusting the zero value of an online moisture meter in cigarette shreds based on model prediction, characterized in that: The method for automatically adjusting the zero of an online moisture meter in cigarette shreds comprises: Construct a prediction model based on historical data of moisture content in cut tobacco during cigarette making; Collect variable data in the cigarette shred process online and in real time and input them into the prediction model to obtain the predicted moisture value; Online real-time collection of actual detection values ​​of tobacco moisture in the cigarette making process; Calculating the deviation between the predicted moisture value and the actual detected value, and setting a deviation threshold; The deviation value is compared with the deviation threshold value to obtain a comparison result, and the online moisture meter is zeroed according to the comparison result.

2. The method for automatically adjusting the zero value of an online moisture meter in cigarette shreds based on model prediction according to claim 1, characterized in that: The comparison result includes that the deviation value is within the deviation threshold or the deviation value is greater than the deviation threshold.

3. The method for automatically adjusting the zero value of an online moisture meter in cigarette shreds based on model prediction according to claim 2, characterized in that: Performing zero adjustment according to the comparison result includes: When the deviation value is within the deviation threshold, automatically zeroing the online moisture meter; When the deviation value is greater than the deviation threshold, a verification work order is generated and pushed to relevant personnel, prompting them to perform manual inspection and calibration.

4. The method for automatically adjusting the moisture content of cigarette shreds online based on model prediction according to claim 3, characterized in that: Automatically zeroing the online moisture meter includes: Calculate the zeroing step length and the zeroing direction according to the magnitude and direction of the deviation value; Determining a zero adjustment time interval and a zero adjustment amplitude according to the dynamic response characteristics of the online moisture meter; The online moisture meter is zeroed based on the zeroing step, the zeroing direction, the zeroing time interval and the zeroing amplitude.

5. The method for automatically adjusting the moisture content of cigarette shreds online based on model prediction according to claim 4, characterized in that: Automatically zeroing the online moisture meter further comprises: The zero adjustment dead zone is set, and the automatic zero adjustment is not triggered when the deviation value is within the preset range.

6. The method for automatically adjusting the zero value of an online moisture meter in cigarette shreds based on model prediction according to claim 2, characterized in that: The deviation threshold ranges from 0.2 to 0.

3.

7. The method for automatically adjusting the zero value of an online moisture meter in cigarette shreds based on model prediction according to claim 1, characterized in that: The method for automatically zeroing an online moisture meter in cigarette shred processing also includes preprocessing the historical data.

8. The method for automatically adjusting the zero value of an online moisture meter in cigarette shreds based on model prediction according to claim 7, characterized in that: Preprocessing the historical data includes: Filling missing values ​​in the historical data using linear interpolation; Using the 3-σ principle to detect and eliminate outliers in the historical data; The historical data is smoothed using a moving average filter.

9. The method for automatically adjusting the zero value of an online moisture meter in cigarette shreds based on model prediction according to claim 1, characterized in that: The prediction model is built based on the historical data of moisture content in tobacco during the cigarette making process, including: Extracting time series features from the historical data and obtaining time series features; Based on the time series characteristics, a prediction model is constructed using a support vector regression algorithm and a random forest regression algorithm; The performance of the prediction model was evaluated using k-fold cross validation and leave-one-out method.

10. The method for automatically adjusting the zero value of an online moisture meter in cigarette shreds based on model prediction according to claim 9, characterized in that: Extracting time series features from the historical data includes: The historical data is scaled using a minimum-maximum standardization method to scale the value range of each variable to within the interval [0,1]; The lag term or sliding window statistic of each of the variables is calculated and the lag term or the sliding window statistic is used as the time series feature.

11. A computer-readable storage medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 10 are implemented.

12. A computer device comprising a memory and a processor, wherein a computer program capable of running on the processor is stored in the memory, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 10 are implemented.

Images