Method for analyzing root causes of moisture in whole cut tobacco production line based on multi-model fusion

Abstract

The invention discloses a multi-model fusion-based method for analyzing root causes of moisture in a whole silk making line. The method comprises the following steps of: A, collecting and preprocessing environment factors and process parameter data of the whole silk making line; b, constructing a key equipment outlet moisture influence factor model; c, constructing a key equipment inlet moisture influence factor model; d, constructing an important auxiliary connection equipment moisture influence factor model; and E, constructing a moisture change root cause model of different workshop sections. According to the method, the reason of the moisture change of the key equipment is firstly positioned, then the factors and the process parameters influencing the moisture change are found, the target moisture value is stabilized by adjusting the related factors and the process parameters, and then causal inference is quickly carried out to ensure the unbiasedness of the outlet moisture and enhance the interpretability of the influence of the parameters on the outlet moisture; therefore, the reason for fluctuation of the target moisture value can be quickly found, measures can be taken in time to stabilize moisture, and the production efficiency and the product quality of the whole silk making line are improved.

Description

Technical Field

[0001] This invention relates to the field of tobacco technology, specifically to a method for root cause analysis of moisture content in tobacco processing lines based on multi-model fusion. Background Technology

[0002] Although the process data from traditional tobacco production lines is collected and stored in information systems, it is rarely used for further analysis. Usually, only simple statistical and data correlation analyses are performed, without delving into the causal relationships between parameters or fully exploring the value of production process data.

[0003] For example, the Chinese patent "A Method for Mining Fluctuation Relationships in Silk Processing Workshop Data Based on Association Rules," patent number "202111286338.4," provides a method for mining fluctuation relationships in silk processing workshop data based on association rules, belonging to the field of data mining. Addressing the difficulty in describing the fluctuation relationships between silk processing process parameters, this invention leverages association rules to fully mine the fluctuation relationships between different variables. It utilizes data discretization to design a data preprocessing method for mining association rules in silk processing data, transforming the silk processing data into fluctuation data containing original data information; further, it uses the principles of association rules to design fluctuation rule formulas for the data to be mined. This invention fully explores the fluctuation rules between different process parameters, improving the interpretability of the rules and the accuracy of the correlation results. It can accurately determine whether there are fluctuation relationships between various process parameters in the yarn making workshop, which is beneficial for the statistics and management of process parameters in the yarn making workshop. It also makes it easier to optimize the process by adjusting process parameters when there are problems with the quality of yarn products. However, this patent is mainly based on the mining of correlation and fluctuation relationships between different variables based on association rules. Moreover, this patent focuses on analyzing and mining the fluctuation rules between different process parameters. At the same time, this patent uses the association rule analysis method, which cannot construct high-dimensional relationships between parameters. Furthermore, the adjustment of product quality can only be done on a single factor, and cannot be done through the joint adjustment of multiple factors.

[0004] For example, the Chinese patent "A Method, System, Equipment and Storage Medium for Tobacco Processing Operation Data Analysis" (patent number "201910697841.5") provides a method, system, equipment and storage medium for tobacco processing operation data analysis, relating to the field of tobacco industry technology. This method includes: acquiring production information from various tobacco plants; obtaining tobacco production technology standards corresponding to the tobacco plants from the production information according to a preset mapping relationship; using the production information as an index to retrieve the tobacco processing operation data of the tobacco plants from a tobacco database; and determining whether the tobacco processing operation data conforms to the tobacco production technology standards, obtaining a judgment result. This enables data analysis of tobacco production processes from different tobacco plants, offering flexible, configurable, and maintainable technical advantages. However, this patent focuses on collecting tobacco processing operation data from different production lines and analyzing whether the data conforms to the standards. Furthermore, this patent only performs simple statistics and judgments on whether the data meets the standards, without designing causal relationship modeling, and it cannot adjust process parameters to make the results tend towards the standard values. Summary of the Invention

[0005] The purpose of this invention is to provide a multi-model fusion-based method for analyzing the root causes of moisture in the yarn-making process, in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for analyzing the root causes of moisture content in yarn processing lines based on multi-model fusion, the method comprising the following steps: A. Collection and preprocessing of environmental factors and process parameters for the entire yarn production line; B. Construct a model of the influencing factors of moisture content at the outlet of key equipment; C. Construct a model of influencing factors of inlet moisture in key equipment; D. Construct a model of the influencing factors of moisture content in important auxiliary equipment; E. Construct root cause models of moisture changes in different work sections; F. Water stability control based on root cause model.

[0007] Preferably, in step A, the environmental factors include ambient temperature and humidity, atmospheric pressure, atmospheric moisture pressure, moisture meter data, seasonal meteorological information, and morning / evening shift information, while the process parameters include tobacco brand, tobacco batch, equipment control parameters, inlet and outlet moisture, hot air temperature, water addition, drum speed, water addition ratio between front and rear chambers, cumulative material amount, and dehumidification damper opening.

[0008] Preferably, in step D, the auxiliary equipment includes an HT heating and humidifying machine, a leaf storage cabinet, a SIROX heating and humidifying machine, a temporary storage cabinet, and a yarn mixing cabinet.

[0009] Preferably, in step E, based on the moisture loss model and the equipment parameter data within the section, the silk-making line is divided into three sections: pretreatment, leaf silk production, and blending and flavoring. The abnormal moisture content detection, root cause localization, and multi-dimensional element combination root cause are used as three key models, and the abnormal moisture content detection model, root cause localization model, and multi-dimensional element combination root cause model within the section are constructed.

[0010] Preferably, the root cause localization model needs to consider the cyclical nature of moisture within the process section, while the multi-dimensional element combination root cause model finds the causes of moisture fluctuations within the process section by analyzing the moisture influencing factors at the outlet of key equipment, the moisture influencing factors at the inlet of key equipment, and the moisture influencing factors at key auxiliary equipment. Through the establishment of these models, the root cause models of moisture changes in the pretreatment section, leaf fiber production section, and blending and flavoring section of the silk production line are obtained. Furthermore, the abnormal moisture content detection model within the process section is based on the detection of moisture stability data within the current batch and the detection of minute-by-minute moisture stability data within the current batch.

[0011] Preferably, in step F, the root cause model is based on key features, and the specific process includes the following steps: a) detecting moisture in the yarn-making line: measuring the moisture at the inlet and outlet of the yarn-making line in real time and comparing it with the target moisture to obtain the moisture difference; b) determining root cause factors: based on the root cause model of moisture in the yarn-making line, analyzing the root causes of the moisture difference and determining the key factors affecting moisture; c) adjusting process parameters: adjusting process parameters according to the analysis results of the root cause model; d) monitoring moisture in the yarn-making line: tracking the changes in moisture at the inlet and outlet of the yarn-making line in real time according to the adjusted process parameters; e) continuous optimization: continuously optimizing the control strategy of key factors based on the root cause model of moisture in the yarn-making line.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention first identifies the causes of moisture variations in key equipment, then finds the factors and process parameters affecting these variations, and subsequently stabilizes the target moisture value by adjusting these factors and parameters. It then rapidly performs causal inference to ensure the unbiasedness of the outlet moisture content and enhances the interpretability of the parameters' influence on outlet moisture. This allows for the rapid identification of the causes of fluctuations in the target moisture value, enabling timely measures to stabilize moisture, improve the overall production efficiency and product quality of the silk-making line, and identify the factors and process parameters related to moisture variations. This helps in better understanding the influencing factors in the production process, facilitating more precise control of the production flow. Finally, causal inference allows for a better understanding of the relationship between parameters and outlet moisture content, improving the model's reliability and interpretability, and ultimately helping to optimize the overall production process of the silk-making line. Detailed Implementation

[0013] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] A method for analyzing the root causes of moisture content in a yarn-making line based on multi-model fusion, comprising the following steps: A. Collection and preprocessing of environmental factors and process parameters for the entire yarn production line; B. Construct a model of key equipment outlet moisture influencing factors. Through real-time analysis of environmental temperature and humidity, raw tobacco, equipment parameters and process inspection factors, conduct parameter analysis on the outlet moisture of loosening rehumidifier, feeding machine, drying machine and flavoring machine, study the key parameters affecting moisture, and conduct quantitative analysis to achieve rapid causal inference. Through this quantitative analysis, ensure the unbiasedness of outlet moisture and enhance the interpretability of the parameters' influence on outlet moisture. C. Construct a model of inlet moisture in key equipment. By analyzing the inlet moisture inlet factors of the feeder, dryer and flavoring machine in the silk making line, study the correlation between the inlet moisture of each equipment and the parameters of the preceding process. Further study the key parameters that affect inlet moisture and how these key parameters affect inlet moisture. The key is to achieve quantitative analysis of the impact of parameters on inlet moisture. D. Construct a model of the influencing factors of moisture content in important auxiliary equipment; E. Construct root cause models of moisture changes in different work sections; F. Water stability control based on root cause model.

[0015] In step A, environmental factors include ambient temperature and humidity, atmospheric pressure, atmospheric moisture pressure, moisture meter data, seasonal weather information, and morning / evening shift information. Process parameters include tobacco brand, tobacco batch, equipment control parameters, inlet and outlet moisture, hot air temperature, water addition, drum speed, water addition ratio between front and rear chambers, cumulative material amount, and exhaust damper opening.

[0016] In step D, the auxiliary equipment includes HT heating and humidifying machine, leaf storage cabinet, SIROX heating and humidifying machine, temporary storage cabinet and blending cabinet. The study investigates the impact of each piece of equipment on the moisture content of tobacco, explores the influence of important parameters within the equipment on moisture, and establishes a moisture influencing factor model related to these parameters. These models can help people quantitatively analyze the impact of auxiliary equipment on the moisture content of tobacco, thereby better controlling the moisture content at the outlet.

[0017] In step E, based on the moisture loss model and the equipment parameter data within the section, the entire silk-making line is divided into three sections: pretreatment, leaf silk production, and blending and flavoring. The abnormal moisture content detection, root cause localization, and multi-dimensional element combination root cause are used as three key models, and the abnormal moisture content detection model, root cause localization model, and multi-dimensional element combination root cause model within the section are constructed.

[0018] Root cause localization models need to consider the cyclical nature of moisture within each process section, while multi-dimensional element combination root cause models analyze the moisture influencing factors at the outlet, inlet, and auxiliary equipment of key equipment to identify the causes of moisture fluctuations within each process section. By establishing these models, root cause models of moisture changes in the pretreatment, leaf fiber, and blending and flavoring sections of the silk-making line can be obtained, thereby helping to predict and control moisture changes throughout the entire silk-making process. Furthermore, the abnormal moisture content detection model within each process section is based on the detection of moisture stability data within the current batch and the minute-by-minute moisture stability data within the current batch, used to detect the moisture situation within the process section.

[0019] In step F, the root cause model is based on key features. When applied to process parameter control, it adjusts the target moisture value by controlling process parameters / single factors. The specific process includes the following steps: a) Detecting moisture in the yarn-making line: Real-time measurement of moisture at the inlet and outlet of the yarn-making line is performed and compared with the target moisture value to obtain the moisture difference; b) Determining root cause factors: Based on the root cause model of moisture in the yarn-making line, the root causes of the moisture difference are analyzed to determine the key factors affecting moisture; c) Adjusting process parameters: According to… The analysis results of the root cause model are used to adjust process parameters to control key factors affecting moisture; d) Monitor the moisture content of the entire spinning line: Based on the adjusted process parameters, the moisture content changes at the inlet and outlet of the spinning line are tracked in real time to ensure that it gradually approaches the target moisture content; e) Continuous optimization: Based on the root cause model of moisture content in the spinning line, the control strategy of key factors is continuously optimized to improve the overall spinning quality and efficiency. The root cause model can be effectively applied to control the moisture content of the spinning line, improve the accuracy and stability of moisture control, avoid the adverse effects of moisture instability, and greatly reduce waste and cost expenditures in the spinning process.

[0020] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for analyzing the root causes of moisture content in yarn processing lines based on multi-model fusion, characterized in that: The method includes the following steps: A. Collection and preprocessing of environmental factors and process parameters for the entire yarn production line; B. Construct a model of the influencing factors of moisture content at the outlet of key equipment; C. Construct a model of influencing factors of inlet moisture in key equipment; D. Construct a model of the influencing factors of moisture content in important auxiliary equipment; E. Construct root cause models of moisture changes in different work sections; F. Water stability control based on root cause model.

2. The method for analyzing the root causes of moisture content in yarn-making lines based on multi-model fusion as described in claim 1, characterized in that: In step A, environmental factors include ambient temperature and humidity, atmospheric pressure, atmospheric moisture pressure, moisture meter data, seasonal weather information, and morning / evening shift information. Process parameters include tobacco brand, tobacco batch, equipment control parameters, inlet and outlet moisture, hot air temperature, water addition, drum speed, water addition ratio between front and rear chambers, cumulative material amount, and dehumidification damper opening.

3. The method for analyzing the root causes of moisture content in yarn-making lines based on multi-model fusion as described in claim 1, characterized in that: In step D, the auxiliary equipment includes an HT heating and humidifying machine, a leaf storage cabinet, a SIROX heating and humidifying machine, a temporary storage cabinet, and a yarn mixing cabinet.

4. The method for analyzing the root causes of moisture content in a yarn-making line based on multi-model fusion as described in claim 1, characterized in that: In step E, based on the moisture loss model and the equipment parameter data within the section, the silk-making line is divided into three sections: pretreatment, leaf silk production, and blending and flavoring. The abnormal moisture content detection, root cause localization, and multi-dimensional element combination root cause are used as three key models, and the abnormal moisture content detection model, root cause localization model, and multi-dimensional element combination root cause model within the section are constructed.

5. The method for analyzing the root causes of moisture content in yarn-making lines based on multi-model fusion as described in claim 4, characterized in that: The root cause localization model needs to consider the cyclical nature of moisture within the process section, while the multi-dimensional element combination root cause model finds the causes of moisture fluctuations within the process section by analyzing the moisture influencing factors at the outlet of key equipment, the inlet of key equipment, and the moisture influencing factors of key auxiliary equipment. Through the establishment of these models, the root cause models of moisture changes in the pretreatment section, leaf fiber production section, and blending and flavoring section of the silk production line are obtained. Furthermore, the abnormal moisture content detection model within the process section is based on the detection of moisture stability data within the current batch and the detection of minute-by-minute moisture stability data within the current batch.

6. The method for analyzing the root causes of moisture content in yarn-making lines based on multi-model fusion according to claim 1, characterized in that: In step F, the root cause model is based on key features, and the specific process includes the following steps: a) Detecting moisture in the yarn processing line: Real-time measurement of moisture at the inlet and outlet of the yarn processing line is performed and compared with the target moisture level to obtain the moisture difference; b) Determining root cause factors: Based on the root cause model of moisture in the yarn processing line, the root causes of the moisture difference are analyzed, and key factors affecting moisture are determined; c) Adjusting process parameters: Based on the analysis results of the root cause model, process parameters are adjusted; d) Monitoring moisture in the yarn processing line: Based on the adjusted process parameters, the changes in moisture at the inlet and outlet of the yarn processing line are tracked in real time; e) Continuous optimization: Based on the root cause model of moisture in the yarn processing line, the control strategy for key factors is continuously optimized.