A method, system, medium and product for intelligent monitoring and early warning of tobacco cutting moisture

Abstract

The application discloses a kind of intelligent monitoring and early warning method, system, medium and product of tobacco manufacturing moisture, and relates to data processing technical field.In the method, first, the sample data and process parameters of adjacent moisture nodes are obtained, and the process parameter set and moisture deviation transmission coefficient are determined;Second, the initial moisture loss mapping relationship is constructed, and the target mapping relationship is obtained by combining historical temporary storage time and environmental parameter correction;Then, real-time production data is obtained and substituted into the target mapping relationship, and the normalized moisture loss deviation rate of the current process is calculated;Subsequently, based on the deviation rate and the transmission coefficient, the inlet moisture benchmark of the downstream unfinished process is corrected step by step, and the cascade corrected predicted moisture value of the finished product node is obtained;Finally, the cumulative moisture deviation risk index is calculated, the finished product prediction deviation is determined by combining the above predicted moisture value, and then the moisture early warning is executed.The technical scheme provided by the application can improve the accuracy of moisture intelligent monitoring and early warning in the process of tobacco manufacturing.

Description

Technical Field

[0001] This application relates to the field of data processing technology, specifically to a method, system, medium, and product for intelligent monitoring and early warning of moisture content in silk production. Background Technology

[0002] In tobacco processing, moisture content is one of the key technological indicators affecting the quality and processing performance of tobacco shreds. The tobacco processing process typically involves multiple continuous steps, including formula addition, shredding, drying, flavoring, and stem processing. The materials undergo multiple processes of water addition, heating, and drying. Appropriate moisture content not only ensures the physical properties of tobacco shreds, such as flexibility, filling capacity, and hygroscopicity, but also directly affects the smoothness of subsequent cigarette processing and the sensory quality of the finished cigarette. Stable control of these characteristics is crucial for achieving consistent tobacco product quality and improving production efficiency.

[0003] Currently, moisture monitoring in silk refining production lines mainly relies on online moisture detection systems. These systems utilize infrared moisture meters, microwave moisture meters, and other detection devices deployed at key process nodes to continuously collect moisture content data during material flow, obtaining real-time moisture status information for each process. This online detection-based monitoring method can acquire basic moisture characteristics at each node and has been widely used in silk refining process management.

[0004] However, with the continuous improvement of tobacco product quality requirements and the increasing complexity of production processes, the moisture variation patterns during the tobacco processing have become increasingly complex. In actual production, online detection-based monitoring methods struggle to accurately predict the final moisture state of the current batch of materials after all processes are completed, and their ability to provide early warnings of abnormal moisture levels in the finished product is difficult to guarantee. Especially when there are varying degrees of moisture deviation in each process, and these deviations propagate between processes and show a cumulative diffusion trend, the lack of effective process correlation analysis and deviation propagation prediction mechanisms can easily lead to the monitoring system failing to identify the risk of moisture accumulation in a timely and accurate manner, resulting in delayed early warnings. This reduces the accuracy of intelligent moisture monitoring and early warning during the tobacco processing process. Summary of the Invention

[0005] This application provides a method, system, medium, and product for intelligent monitoring and early warning of moisture content in silk production, which can improve the accuracy of intelligent monitoring and early warning of moisture content during the silk production process.

[0006] The first aspect of this application provides a method for intelligent monitoring and early warning of moisture content in silk production, including: Acquire sample moisture detection data from multiple adjacent moisture detection nodes on the silk production line and sample process parameters for each process. Based on the sample moisture detection data and the sample process parameters, determine the process parameter set for each process and the moisture deviation transfer coefficient between adjacent processes; Based on the set of process parameters, an initial moisture loss mapping relationship for each process is constructed, and the initial moisture loss mapping relationship is corrected by combining the historical storage time and historical environmental parameters between processes to obtain the target moisture loss mapping relationship for each process. The real-time moisture detection data, real-time process parameters, real-time storage time and real-time environmental parameters of the current batch of materials are obtained, and the real-time process parameters, real-time storage time and real-time environmental parameters are substituted into the target moisture loss mapping relationship, and the normalized moisture loss deviation rate of the current process is calculated in combination with the real-time moisture detection data. Based on the normalized moisture loss deviation rate and the moisture deviation transmission coefficient, the inlet moisture benchmark of the downstream unfinished process is corrected and transmitted step by step to obtain the cascaded corrected predicted moisture value of the finished product node. The cumulative moisture deviation risk index is calculated based on the normalized moisture loss deviation rate of each completed process. The finished product prediction deviation is determined by combining the cascaded correction prediction moisture value. Moisture warning is then executed based on the cumulative moisture deviation risk index and the finished product prediction deviation.

[0007] By adopting the above technical solution, the system first acquires sample moisture detection data from multiple adjacent moisture detection nodes on the yarn production line and sample process parameters for each process. Then, based on the sample moisture detection data and sample process parameters, it determines the process parameter set for each process and the moisture deviation transmission coefficient between adjacent processes. On this basis, an initial moisture loss mapping relationship for each process is constructed based on the process parameter set. This initial moisture loss mapping relationship is then corrected by incorporating historical storage time and historical environmental parameters between processes to obtain the target moisture loss mapping relationship. When real-time data for the current batch of material is acquired, the real-time process parameters, real-time storage time, and real-time environmental parameters are substituted into the target moisture loss mapping relationship, and the normalized moisture loss deviation rate for the current process is calculated based on the real-time moisture detection data. Furthermore, based on the normalized moisture loss deviation rate and the moisture deviation transmission coefficient, the inlet moisture benchmark for downstream unfinished processes is progressively corrected and transmitted to obtain the cascaded corrected predicted moisture value for the finished product node. This establishes a deviation transmission prediction mechanism from intermediate processes to the final finished product, enabling the system to predict the final moisture state of the material after all processes are completed during production. Finally, the cumulative moisture deviation risk index is calculated based on the normalized moisture loss deviation rate of each completed process. The predicted deviation of the finished product is determined by combining the cascaded correction prediction moisture value. Moisture warning is then executed based on the cumulative moisture deviation risk index and the predicted deviation of the finished product. This achieves effective identification and prediction of the cumulative diffusion trend of moisture deviation between processes, solving the problem of delayed warning caused by the lack of process correlation analysis and deviation transmission prediction mechanism in the existing technology. This improves the accuracy of intelligent moisture monitoring and warning in the yarn making process.

[0008] Optionally, based on the sample moisture detection data, the moisture loss between adjacent moisture detection nodes is calculated, and the correlation coefficient between the moisture loss and each sample process parameter is calculated; the sample process parameters with correlation coefficient values ​​greater than a preset correlation threshold are extracted and combined to form the sample process parameter set; the measured moisture values ​​of the same production batch at adjacent moisture detection nodes are extracted, and the moisture deviation between the measured moisture values ​​and the historical overall mean of the corresponding nodes is calculated; the ratio of moisture deviation values ​​between adjacent processes is calculated, and the ratio is used as the moisture deviation transmission coefficient between adjacent processes.

[0009] Optionally, the initial water loss mapping relationship is obtained by performing regression fitting with the parameters in the process parameter set as independent variables and the sample water loss as dependent variable; a temporary correction term is obtained by fitting according to the historical temporary storage time; an environmental correction term is obtained by fitting according to the historical environmental parameters; and the temporary correction term and the environmental correction term are superimposed on the initial water loss mapping relationship to obtain the target water loss mapping relationship.

[0010] Optionally, the real-time process parameters, the real-time storage time, and the real-time environmental parameters are substituted into the target moisture loss mapping relationship to calculate the downstream node moisture prediction benchmark value of the current process; the instantaneous flow rate of the material is extracted from the real-time process parameters, and the measurement delay of the moisture detection probe is compensated based on the instantaneous flow rate of the material to correct the real-time moisture detection data to the target moisture measured value; the temperature and humidity fluctuation characteristics in the real-time environmental parameters are extracted, the environmental penalty coefficient is calculated based on the temperature and humidity fluctuation characteristics, and the initial moisture tolerance range of the current process is dynamically scaled using the environmental penalty coefficient to obtain the dynamic moisture tolerance range; the deviation between the downstream node moisture prediction benchmark value and the target moisture measured value is calculated, and the deviation is divided by the dynamic moisture tolerance range to obtain the normalized moisture loss deviation rate.

[0011] Optionally, the real-time moisture detection data of the downstream node of the current process is used as the initial benchmark for the inlet moisture of the next process; the product of the normalized moisture loss deviation rate, the moisture deviation transmission coefficient, and the moisture tolerance range of the current process is used as the deviation transmission correction amount; the deviation transmission correction amount is superimposed on the initial benchmark for inlet moisture to obtain the corrected benchmark value for inlet moisture of the next process; based on the corrected benchmark value for inlet moisture of the next process and the target moisture loss mapping relationship of the next process, the cascaded correction prediction value of the next node is calculated and transmitted step by step to the final process to obtain the cascaded correction prediction moisture value of the finished product node.

[0012] Optionally, the process weighting coefficient is determined based on the historical average moisture loss ratio of each process; the normalized moisture loss deviation rate of each completed process is weighted and summed based on the process weighting coefficient to obtain the cumulative moisture deviation risk index; the difference between the cascaded correction predicted moisture value and the median of the finished product process moisture standard is calculated, and the difference is divided by the half-width of the finished product moisture acceptable range to obtain the finished product predicted deviation; it is determined whether the cumulative moisture deviation risk index or the finished product predicted deviation exceeds the corresponding critical warning threshold; if at least one of the cumulative moisture deviation risk index or the finished product predicted deviation exceeds the corresponding critical warning threshold, a moisture warning signal is triggered.

[0013] Optionally, the measured moisture content of the finished product at the finished product node of the current batch is obtained, and the terminal residual between the measured moisture content of the finished product and the cascaded correction predicted moisture content is calculated; the terminal residual is allocated according to the proportion of the normalized moisture loss deviation rate of each completed process to obtain the residual component of each process; the residual component is used to compensate the target moisture loss mapping relationship of each process to generate an updated moisture loss mapping relationship; based on the cumulative moisture deviation risk index when the moisture warning is triggered and the updated moisture loss mapping relationship, the moisture deviation transmission coefficient between adjacent processes is dynamically adjusted.

[0014] Secondly, embodiments of this application provide a silk-making moisture intelligent monitoring and early warning system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, which includes computer instructions, and the one or more processors call the computer instructions to cause the silk-making moisture intelligent monitoring and early warning system to perform the method described in the first aspect and any possible implementation thereof.

[0015] Thirdly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a silk-making moisture intelligent monitoring and early warning system, cause the silk-making moisture intelligent monitoring and early warning system to perform the method described in the first aspect and any possible implementation thereof.

[0016] Fourthly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a silk-making moisture intelligent monitoring and early warning system, causes the silk-making moisture intelligent monitoring and early warning system to execute the method described in the first aspect and any possible implementation thereof.

[0017] In summary, one or more technical solutions provided in this application have at least the following technical effects or advantages: By adopting the above technical solution, the system first acquires sample moisture detection data from multiple adjacent moisture detection nodes on the yarn production line and sample process parameters for each process. Then, based on the sample moisture detection data and sample process parameters, it determines the process parameter set for each process and the moisture deviation transmission coefficient between adjacent processes. On this basis, an initial moisture loss mapping relationship for each process is constructed based on the process parameter set. This initial moisture loss mapping relationship is then corrected by incorporating historical storage time and historical environmental parameters between processes to obtain the target moisture loss mapping relationship. When real-time data for the current batch of material is acquired, the real-time process parameters, real-time storage time, and real-time environmental parameters are substituted into the target moisture loss mapping relationship, and the normalized moisture loss deviation rate for the current process is calculated based on the real-time moisture detection data. Furthermore, based on the normalized moisture loss deviation rate and the moisture deviation transmission coefficient, the inlet moisture benchmark for downstream unfinished processes is progressively corrected and transmitted to obtain the cascaded corrected predicted moisture value for the finished product node. This establishes a deviation transmission prediction mechanism from intermediate processes to the final finished product, enabling the system to predict the final moisture state of the material after all processes are completed during production. Finally, the cumulative moisture deviation risk index is calculated based on the normalized moisture loss deviation rate of each completed process. The predicted deviation of the finished product is determined by combining the cascaded correction prediction moisture value. Moisture warning is then executed based on the cumulative moisture deviation risk index and the predicted deviation of the finished product. This achieves effective identification and prediction of the cumulative diffusion trend of moisture deviation between processes, solving the problem of delayed warning caused by the lack of process correlation analysis and deviation transmission prediction mechanism in the existing technology. This improves the accuracy of intelligent moisture monitoring and warning in the yarn making process. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the intelligent monitoring and early warning method for silk moisture content disclosed in the embodiments of this application; Figure 2 This is another schematic diagram of the intelligent monitoring and early warning method for silk production moisture disclosed in the embodiments of this application; Figure 3 This is a schematic diagram of the structure of a system provided in an embodiment of this application.

[0019] Explanation of reference numerals in the attached drawings: 301, Central Processing Unit; 302, Read-Only Memory; 303, Random Access Memory; 304, Bus; 305, Input / Output Interface; 306, Input Section; 307, Output Section; 308, Storage Section; 309, Communication Section; 310, Driver; 311, Removable Media. Detailed Implementation

[0020] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0021] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0022] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0023] This application provides a method for intelligent monitoring and early warning of moisture content in silk processing, referring to... Figure 1 , Figure 1 This is a flowchart illustrating a method for intelligent monitoring and early warning of moisture content in silk processing, provided in an embodiment of this application. The method is applied to a system, which refers to a hardware and software integrated platform capable of executing an intelligent monitoring and early warning program for moisture content in silk processing. The system can execute an intelligent monitoring and early warning program for moisture content in silk processing. The method includes steps 101 to 106, as follows: Step 101: Obtain sample moisture detection data from multiple adjacent moisture detection nodes on the silk production line and sample process parameters for each process.

[0024] In this embodiment, moisture detection nodes refer to measurement points installed on the yarn production line for detecting the moisture content of materials. These typically include detection devices such as infrared moisture meters and microwave moisture meters, which can collect moisture content data in real time during material flow. Sample moisture detection data refers to the moisture content measurements collected from various moisture detection nodes during historical production processes, forming a historical dataset used to build the model. Sample process parameters represent the operational parameters of various processes related to moisture control, including but not limited to key process settings such as yarn drying temperature, steam pressure, water addition, and yarn drying speed. These data originate from historical records of the production process.

[0025] Specifically, the process of acquiring sample moisture detection data and sample process parameters includes extracting historical production batch moisture detection records and process parameter records from the data acquisition system of the yarn production line. First, a valid historical data time window is determined, typically selecting production data from the most recent 3-6 months to ensure the timeliness and representativeness of the sample. Then, a complete dataset of records from all adjacent moisture detection nodes within this time window is queried and exported from the process management system database, including the measurement timestamp, batch identifier, material identifier, and moisture content value for each node. Simultaneously, process parameter records for each process within the corresponding time period are extracted, and these parameters are correlated and integrated according to batch and time to form a structured sample dataset. Outliers and missing values ​​that may exist in the data are identified and processed using statistical methods to ensure the data quality for subsequent modeling and analysis.

[0026] Step 102: Based on the sample moisture detection data and sample process parameters, determine the process parameter set for each process and the moisture deviation transfer coefficient between adjacent processes.

[0027] In this embodiment, the moisture deviation transmission coefficient represents a coefficient value that quantifies the degree of influence of moisture deviation transmission between adjacent processes. It reflects the degree of influence of the moisture state of the upstream process on the downstream process and is used to predict the transmission effect of moisture deviation on the production line.

[0028] Specifically, determining the process parameter set and moisture deviation transmission coefficient involves two main steps. First, for each process step, the correlation coefficient between each process parameter and the moisture changes at the inlet and outlet of the process is calculated. Parameters with an absolute correlation coefficient greater than a preset threshold (usually 0.3) are selected as key parameters, and these key parameters are combined into the process parameter set for that process. For parameters exhibiting multicollinearity, variance inflation factor analysis is used to screen and retain parameters with strong independence. Second, for the moisture deviation relationship between adjacent processes, a linear regression model is constructed, using the upstream process outlet moisture deviation as the independent variable and the downstream process inlet moisture deviation as the dependent variable. The regression coefficient is estimated using the least squares method, and this regression coefficient is the corresponding moisture deviation transmission coefficient. Simultaneously, the correlation index of moisture deviation for each pair of adjacent processes is calculated as a basis for evaluating the reliability of the transmission coefficient. For sample data from multiple batches, a weighted average method is used to synthesize the transmission coefficients of each batch, with the weights proportional to the sample size and time proximity of the batches.

[0029] In one possible implementation, based on sample moisture detection data and sample process parameters, the process parameter set for each process and the moisture deviation transfer coefficient between adjacent processes are determined, specifically including steps 1021-1023, as follows: Step 1021: Based on the sample moisture detection data, calculate the water loss between adjacent moisture detection nodes, and calculate the correlation coefficient between the water loss and the process parameters of each sample.

[0030] In this embodiment, the correlation coefficient value represents the statistical correlation strength between water loss and various process parameters, serving as an indicator to quantify the influence of different process parameters on moisture changes. The correlation coefficient value typically ranges from -1 to 1; the closer the absolute value is to 1, the stronger the correlation. A positive value indicates a positive correlation (i.e., as the parameter value increases, water loss increases), while a negative value indicates a negative correlation (i.e., as the parameter value increases, water loss decreases). For example, the correlation coefficient between drying temperature and water loss is 0.85, indicating that water loss usually increases significantly with increasing temperature, and there is a strong positive correlation between the two.

[0031] Specifically, calculating the correlation coefficient first requires determining the pairing relationship between adjacent moisture detection nodes on the yarn production line, establishing a correspondence between nodes according to the material flow sequence. Then, the moisture detection data for each pair of adjacent nodes are matched according to production batch and timestamp to ensure that the calculation represents the moisture data of the same batch of material at different detection points. The moisture loss between each pair of adjacent nodes is calculated, i.e., the moisture value of the preceding node minus the moisture value of the following node, yielding the moisture reduction in that process segment. Simultaneously, the process parameter data for each batch of material is compiled, including control parameters such as temperature, pressure, and speed. For each process parameter, a data pair with the moisture loss is constructed, forming two data sequences. Using the Pearson correlation coefficient calculation method, the covariance of these two sets of data is divided by the product of their respective standard deviations to obtain the correlation coefficient value between the process parameter and the moisture loss. This calculation process is repeated for all sample process parameters to finally obtain a complete correlation coefficient matrix, which represents the strength of the correlation between all process parameters and the moisture loss.

[0032] Step 1022: Extract and combine the sample process parameters whose correlation coefficient values ​​are greater than the preset correlation threshold to form a sample process parameter set.

[0033] In this embodiment, the sample process parameter set represents the combination of key process parameters that have a significant impact on moisture loss after correlation screening. It is the core set of input variables for subsequently constructing the moisture mapping relationship. This parameter set excludes minor factors with weak influence on moisture changes and retains the main process parameters with statistically significant correlations, improving the model's specificity and effectiveness. For example, the sample process parameter set for a certain yarn drying process may include several key parameters such as drying temperature, drying speed, and hot air flow rate, while excluding other auxiliary parameters that have no significant impact on moisture.

[0034] Specifically, constructing a sample process parameter set first requires setting a suitable preset correlation threshold. This threshold is usually determined based on the characteristics of the silk-making process and data analysis experience, and is generally set between 0.3 and 0.5. The correlation coefficients between each process parameter calculated in step 1021 and water loss are compared with this preset threshold, and process parameters with absolute correlation coefficient values ​​greater than the threshold are selected. For the selected parameters, their data integrity is checked, and parameters with too many missing values ​​are removed. Then, the multicollinearity between the retained parameters is evaluated. If the absolute correlation coefficient between two parameters is too high (usually greater than 0.8), the parameter with a stronger correlation to water loss is retained to avoid information redundancy in the model. Finally, all the key process parameters that have passed the selection are combined into a set to form a sample process parameter set, which serves as the variable set for subsequent modeling and analysis. This parameter set is stored in structured data format, containing information such as parameter names, units, and correlation coefficient values, facilitating its use in constructing moisture mapping relationships.

[0035] Step 1023: Extract the measured moisture values ​​of the same production batch at adjacent moisture detection nodes, and calculate the moisture deviation value between the measured moisture value and the historical overall mean of the corresponding node; calculate the ratio of the moisture deviation values ​​between adjacent processes, and use the ratio as the moisture deviation transmission coefficient between adjacent processes.

[0036] In this embodiment, the moisture deviation transmission coefficient represents the proportional relationship of moisture deviation transmission between adjacent processes, and is used to quantify the degree of influence of moisture state changes in upstream processes on the moisture state of downstream processes. This coefficient reflects the series effect characteristics in the moisture control system and is a key indicator for predicting the propagation path and intensity of moisture changes along the production line. For example, if the moisture deviation transmission coefficient between the shredding process and the drying process is 0.75, it means that a 1% moisture deviation in the shredding process will result in a corresponding 0.75% moisture deviation at the inlet of the drying process.

[0037] Specifically, calculating the moisture deviation transfer coefficient first requires extracting actual moisture measurement records for the same production batch at adjacent moisture detection nodes from the production database, ensuring that the extracted data has batch identification consistency and time series continuity. For each moisture detection node, sufficient historical detection data is collected, and its historical overall mean is calculated as the moisture baseline value for that node. Then, the difference between each actual measured moisture value and the corresponding node's historical mean is calculated to obtain the moisture deviation value. These deviation values ​​can be positive (indicating moisture is higher than the historical mean) or negative (indicating moisture is lower than the historical mean). For each pair of adjacent processes, their moisture deviation values ​​are obtained separately, forming a deviation value pair. The moisture deviation value of the subsequent process is divided by the moisture deviation value of the previous process to obtain the ratio, which is the moisture deviation transfer coefficient between adjacent processes. During the calculation process, special handling is needed for cases where the deviation value of the previous process is close to zero. This can be avoided by setting a minimum threshold or using data smoothing techniques to prevent calculation instability caused by a divisor close to zero. For multiple batches of sample data, the transfer coefficient for each batch is calculated, and then their weighted average is taken as the final moisture deviation transfer coefficient, where the weights can be determined based on the temporal proximity of the data and the representativeness of the sample.

[0038] Step 103: Construct the initial moisture loss mapping relationship for each process based on the process parameter set, and modify the initial moisture loss mapping relationship by combining the historical storage time and historical environmental parameters between processes to obtain the target moisture loss mapping relationship for each process.

[0039] In the embodiments of this application, the target moisture loss mapping relationship refers to a more accurate moisture change prediction model obtained by further considering the inter-process storage time and environmental factors based on the initial mapping relationship, which can more comprehensively reflect the moisture changes in actual production. For example, the initial moisture loss mapping relationship of the shredding process may mainly consider process parameters such as shredding speed and blade sharpness, while the target mapping relationship will further consider the storage time of the shredded wire on the conveyor belt and the temperature and humidity conditions of the storage area.

[0040] Specifically, constructing the moisture loss mapping relationship for each process involves two stages. The first stage constructs the initial mapping relationship, using a defined set of process parameters as input features and the moisture difference between the process inlet and outlet as the output target. A mapping function is established using multiple regression analysis. A suitable regression model type, such as linear regression, multinomial regression, or locally weighted regression, is selected based on the data distribution characteristics, and the model parameters are optimized using the gradient descent algorithm with training data. Model performance is evaluated through cross-validation, and the model with the best fit is selected as the initial mapping relationship. The second stage corrects the mapping relationship by introducing historical storage time and historical environmental parameter data (such as temperature and humidity) between processes, analyzing the correlation between these factors and the prediction deviation of the initial mapping relationship. A deviation correction function is constructed, with storage time and environmental parameters as independent variables and prediction deviation as the dependent variable. The predicted values ​​of the initial mapping relationship are combined with the output value of the deviation correction function to form a target moisture loss mapping relationship that comprehensively considers process parameters, storage time, and environmental factors. For data under different seasons or climatic conditions, corresponding mapping relationships are established or seasonal factors are introduced for adjustment to adapt to the periodic changes in environmental conditions.

[0041] In one possible implementation, an initial moisture loss mapping relationship for each process is constructed based on a set of process parameters. This initial moisture loss mapping relationship is then corrected by incorporating historical storage times and historical environmental parameters between processes to obtain the target moisture loss mapping relationship for each process. Specifically, this includes steps 1031-1033, as follows: Step 1031: Using the parameters in the process parameter set as independent variables and the sample water loss as the dependent variable, perform regression fitting to obtain the initial water loss mapping relationship.

[0042] In this embodiment, the initial moisture loss mapping relationship represents the mathematical function relationship between process parameters and moisture loss under standard conditions. It serves as a fundamental model describing how key process parameters affect material moisture content during the yarn-making process. This mapping relationship is extracted from historical production data using mathematical fitting methods, quantifying the degree and manner of influence of different process parameters on moisture loss, and providing a theoretical basis for moisture prediction and process optimization. For example, the initial moisture loss mapping relationship for a yarn drying process might indicate that for every 10°C increase in drying temperature, the average moisture loss increases by 0.5% while other parameters remain constant.

[0043] Specifically, constructing the initial water loss mapping relationship first requires organizing the selected set of process parameters and corresponding water loss data to ensure data integrity and consistency. For each sample point, each parameter in the process parameter set is used as the independent variable, and the corresponding water loss is used as the dependent variable to construct a data matrix. Necessary preprocessing is performed on the independent and dependent variables, including data standardization or normalization, to make parameters of different dimensions comparable. An appropriate regression model type is selected based on the data characteristics; this could be multiple linear regression, multinomial regression, or a more complex nonlinear model. Numerical optimization methods such as least squares and gradient descent are used to solve for the model parameters, minimizing the sum of squared errors between the predicted and actual water loss values. Cross-validation is performed on the fitting results to evaluate the model's prediction accuracy and generalization ability, adjusting the model structure or parameters as necessary. The final determined regression equation is the initial water loss mapping relationship, which can calculate the expected water loss based on the input values ​​of the process parameters.

[0044] Step 1032: Obtain the temporary correction term by fitting the historical temporary storage time; obtain the environmental correction term by fitting the historical environmental parameters.

[0045] In this embodiment, the temporary storage correction term represents a correction value for the impact of discontinuous flow stages (such as waiting, temporary storage, and stagnation) of materials on moisture changes during the production process. This correction term compensates for temporary storage effects that cannot be covered by standard continuous production models. It is typically a function of the temporary storage time, varying with the duration of storage. It can capture the natural evaporation or absorption of moisture caused by the material's contact with air in a static environment, improving the accuracy of moisture prediction under non-standard production rhythms. For example, if tobacco shreds are temporarily stored for 2 hours after drying, a 0.2% moisture loss correction may be needed to compensate for the additional moisture evaporation during the temporary storage period.

[0046] Specifically, calculating the temporary storage correction term first requires extracting material temporary storage time data from historical production records, recording moisture changes under different temporary storage conditions. Samples with the same process conditions but different temporary storage times are grouped and compared to isolate moisture changes purely caused by differences in temporary storage time. A data pair is established with temporary storage time as the independent variable and additional moisture change as the dependent variable, typically using linear, exponential, or logarithmic functions to fit the data and obtain the functional relationship between temporary storage time and moisture change. Simultaneously, historical parameter data of the production environment are collected, mainly including factors that significantly affect moisture evaporation, such as ambient temperature and humidity. Moisture change data under different environmental conditions are compared and analyzed to isolate the independent influence of environmental factors. A regression model is established with environmental parameters as independent variables and environmentally induced moisture change as the dependent variable, typically using multivariate linear regression or nonlinear regression methods. The fitting process considers possible interactions between environmental parameters, such as the combined effect of temperature and humidity. After fitting, statistical tests are used to evaluate the significance and applicability of the correction term model, ensuring that the correction term has good predictive ability within the expected environmental conditions and temporary storage time range.

[0047] Step 1033: Superimpose the temporary correction term and the environmental correction term onto the initial water dispersion mapping relationship to obtain the target water dispersion mapping relationship.

[0048] In this embodiment, the "target moisture loss mapping relationship" represents a modified comprehensive moisture loss prediction model, which is a complete mathematical model that integrates temporary storage effects and environmental influences based on the initial mapping relationship. This relationship can more accurately describe the combined impact of process parameters, temporary storage time, and environmental factors on moisture changes under actual production conditions, providing a moisture prediction tool that includes multiple influencing factors. For example, the final target moisture loss mapping relationship may indicate that, under conditions of 25°C, 60% humidity, 120°C drying temperature, 2m / min speed, and 30-minute temporary storage time, the expected moisture loss is 3.2%.

[0049] Specifically, constructing the target water loss mapping relationship first requires defining three components: the initial water loss mapping relationship, the mathematical expressions for temporary correction terms and environmental correction terms. A suitable combination method is determined, typically using an additive model, where each correction term is directly superimposed on the initial mapping relationship. For some special cases, multiplicative correction or more complex combination functions may also be used. The temporary correction term function and the environmental correction term function are substituted with the actual storage time and environmental parameter values ​​to calculate the specific correction amounts. These correction amounts are then combined with the initial water loss mapping relationship according to the determined combination method to form the complete target water loss mapping relationship equation. The prediction accuracy of the combined mapping relationship under different operating conditions is verified. If necessary, the combination weights or methods are fine-tuned to ensure the model performs well under various combinations of conditions. The final determined target water loss mapping relationship is stored in an appropriate mathematical form for easy retrieval in the production control system. This mapping relationship can calculate a more realistic water loss prediction value based on current process parameters, environmental conditions, and expected storage time, providing a reliable basis for moisture control decisions.

[0050] Step 104: Obtain the real-time moisture detection data, real-time process parameters, real-time storage time, and real-time environmental parameters of the current batch of materials, and substitute the real-time process parameters, real-time storage time, and real-time environmental parameters into the target moisture loss mapping relationship, and calculate the normalized moisture loss deviation rate of the current process in combination with the real-time moisture detection data.

[0051] In this embodiment, the normalized moisture loss deviation rate represents the relative deviation between the actual moisture loss in the current process and the theoretically expected moisture loss. Normalization ensures comparability of moisture deviations across different processes and batches, serving as a quantitative indicator reflecting the degree of moisture abnormality in the current batch of material at that process. For example, if the theoretical moisture loss of a process is 2%, while the actual moisture loss is 2.4%, then the moisture loss deviation rate for that process is 20%, indicating that the actual moisture loss exceeds the expected value by 20%.

[0052] Specifically, the process of calculating the normalized moisture loss deviation rate first requires obtaining the latest moisture detection data of the current batch of materials from the production execution system in real time, including the moisture content values ​​at the inlet and outlet of each process. Simultaneously, it requires obtaining real-time process parameter data for each process, such as control parameters like temperature, pressure, and speed, as well as records of the actual temporary storage time of materials between processes and environmental monitoring data (temperature and humidity) of the temporary storage area. These real-time process parameters, storage time, and environmental parameters are then substituted into the established target moisture loss mapping relationship to calculate the theoretical expected moisture loss value under the current process conditions. Next, the actual moisture loss value is calculated using the real-time moisture detection data, i.e., the difference between the moisture content at the process inlet and outlet. The difference between the actual moisture loss value and the theoretical expected moisture loss value is divided by the theoretical expected moisture loss value to obtain the moisture loss deviation rate. To eliminate the influence of differences in moisture baselines between different processes, the calculated deviation rate is normalized and converted into a standard score form, i.e., the original deviation rate is subtracted from the historical average deviation rate and then divided by the standard deviation of the historical deviation rate, finally yielding a comparable normalized moisture loss deviation rate.

[0053] In one possible implementation, real-time process parameters, real-time storage time, and real-time environmental parameters are substituted into the target moisture loss mapping relationship, and the normalized moisture loss deviation rate of the current process is calculated in combination with real-time moisture detection data. Specifically, this includes steps 1041-1044, as follows: Step 1041: Substitute the real-time process parameters, real-time temporary storage time, and real-time environmental parameters into the target moisture loss mapping relationship to calculate the downstream node moisture prediction benchmark value for the current process.

[0054] In this embodiment, the downstream node moisture prediction benchmark value represents the expected moisture content of the material at the next process detection node, calculated through the target moisture loss mapping relationship. It serves as an important reference standard for assessing moisture change trends and controlling the production process. This benchmark value comprehensively considers the influence of current process parameters, storage time, and environmental parameters, and can predict the moisture state of the material after flowing through downstream processes. For example, if the moisture content of the tobacco shreds in the shredding process is 22%, after passing through the drying process under known process parameters and environmental conditions, the downstream node moisture prediction benchmark value may be 13.5%, representing the expected moisture content of the tobacco shreds after the drying treatment.

[0055] Specifically, calculating the downstream node moisture prediction baseline first requires acquiring real-time process parameters of the current production process, including key control parameters such as equipment temperature, pressure, and speed. These parameters are typically collected in real-time through a sensor network on the production line. Simultaneously, real-time material storage time information is collected, recording the duration of material residence at each buffer stage. This data may come from the production scheduling system or dedicated time recording devices. Real-time parameters of the production environment are also collected, primarily including environmental factors affecting moisture changes such as workshop temperature and humidity. This data is obtained through an environmental monitoring system. The collected real-time process parameters, real-time storage time, and real-time environmental parameters are preprocessed according to the data format defined in the target moisture dispersion mapping relationship, including unit conversion and outlier handling. The processed parameter values ​​are then substituted into the previously established target moisture dispersion mapping relationship, which is a mathematical function that maps multiple input variables to moisture changes. Through numerical calculation, the expected moisture content of the material after flowing through the downstream process under the current conditions is obtained, i.e., the downstream node moisture prediction baseline. This baseline reflects the theoretical change in material moisture under given process conditions and environmental factors.

[0056] Step 1042: Extract the instantaneous flow rate of the material from the real-time process parameters, and compensate for the measurement delay of the moisture detection probe based on the instantaneous flow rate of the material to correct the real-time moisture detection data to the target measured moisture value.

[0057] In this embodiment, the target moisture measured value represents the actual moisture content data after measurement delay compensation. It is a more accurate representation of the moisture state obtained by time synchronization and flow correction of the original moisture detection data. This measured value eliminates the time difference caused by the inconsistency between the detection location and the actual control point during material flow, and provides reference data that can be directly compared with the predicted baseline value in the time dimension. For example, when the original moisture detection reading is 15.2%, but the material flow rate is large, resulting in a measurement delay of 30 seconds, the compensated target moisture measured value may be 15.5%, more accurately reflecting the actual moisture state of the material at a specific process point.

[0058] Specifically, obtaining the target moisture content first requires extracting the instantaneous material flow rate information from the real-time process parameter data stream. This is typically calculated indirectly from the flow meter or the rotational speed of the material conveying equipment. A model is then established to model the relationship between material flow rate and detection delay time; generally, the higher the flow rate, the shorter the time delay from the material passing the actual control point to the detection probe. Based on the current instantaneous material flow rate, the corresponding measurement delay time value is queried or calculated. This delay time is dynamically adjusted as the flow rate changes. Raw real-time moisture detection data, including timestamps, is collected from moisture detection probes installed on the production line. Based on the calculated measurement delay time, the timestamps of the raw moisture detection data are adjusted, shifting the data points forward by the corresponding delay time to achieve alignment on the time axis. If necessary, the aligned data may require interpolation to obtain the moisture value at a specific moment. Common interpolation methods include linear interpolation or spline interpolation. After time adjustment and necessary data interpolation, the resulting moisture data is the target moisture content measurement value. These values ​​are aligned with the control decision point in the time dimension and can be directly compared with the predicted values.

[0059] Step 1043: Extract the temperature and humidity fluctuation characteristics from the real-time environmental parameters, calculate the environmental penalty coefficient based on the temperature and humidity fluctuation characteristics, and use the environmental penalty coefficient to dynamically scale the initial moisture tolerance range of the current process to obtain the dynamic moisture tolerance range.

[0060] In this embodiment, the dynamic moisture tolerance range represents an adaptively adjusted allowable range for moisture control based on the degree of fluctuation in environmental conditions, representing a dynamic optimization and upgrade of the traditional fixed tolerance range. This range scales the initial moisture tolerance range using an environmental penalty coefficient, providing stricter control standards when the environment is stable and appropriately relaxing the control tolerance when environmental fluctuations are severe, thus making the moisture control requirements more adaptable to the actual environment. For example, when the temperature and humidity of the workshop environment are stable, the dynamic tolerance range for moisture at the yarn drying outlet may be ±0.3%, while under conditions of severe temperature and humidity fluctuations, the tolerance range may be relaxed to ±0.5% to adapt to the influence of uncontrollable environmental factors.

[0061] Specifically, determining the dynamic moisture tolerance range first requires extracting time-series data of temperature and humidity from real-time environmental parameter data, typically including continuous measurements within a recent time window (e.g., the past 1-4 hours). Statistical characteristics of temperature and humidity within this time window are calculated, including standard deviation, coefficient of variation, range, or spectral analysis results. These indicators reflect the degree of fluctuation in environmental parameters. Based on the extracted temperature and humidity fluctuation characteristics, a calculation model for the environmental penalty coefficient is established. This model maps the degree of environmental fluctuation to a penalty coefficient value; generally, the greater the fluctuation, the lower the penalty coefficient (i.e., a positive number less than 1). The initial moisture tolerance range for the current process is obtained. This is a preset allowable range for moisture control under standard environmental conditions, usually expressed as a percentage or absolute value range of the target value. The calculated environmental penalty coefficient is then mathematically calculated with the initial moisture tolerance range. A common method is to divide the initial range value by the penalty coefficient (when the penalty coefficient is less than 1, the result is an expanded tolerance range). The final range obtained is the dynamic moisture tolerance range. This range automatically adjusts with changes in environmental conditions, providing a reasonable control standard under different environmental stability conditions.

[0062] Step 1044: Calculate the deviation between the predicted baseline value of moisture at downstream nodes and the measured value of target moisture, divide the deviation by the dynamic moisture tolerance range, and obtain the normalized moisture loss deviation rate.

[0063] In this embodiment, the normalized moisture loss deviation rate represents the proportional relationship between the predicted and measured moisture values ​​relative to the allowable variation range, and is a dimensionless deviation assessment index. This index, by standardizing the actual deviation by dividing it by the allowable deviation range, eliminates the influence of inconsistent tolerance standards across different processes or conditions, providing a unified evaluation scale that can be compared across processes and conditions. For example, a normalized moisture loss deviation rate of 0.8 indicates that the current moisture deviation is 80% of the allowable deviation range, close to but not exceeding the control limit; while a value of 1.2 indicates that the deviation has exceeded the allowable range by 20%, requiring process adjustments.

[0064] Specifically, calculating the normalized moisture loss deviation rate first requires obtaining the downstream node moisture prediction baseline value, which is the theoretical expected value calculated based on the target moisture loss mapping relationship. Simultaneously, the measured target moisture value after measurement delay compensation is obtained, which is the measurement result from the actual production process. The difference between these two values ​​is calculated, i.e., the downstream node moisture prediction baseline value minus the measured target moisture value, to obtain the original moisture deviation value. This value can be positive (measured value lower than predicted value) or negative (measured value higher than predicted value). A defined dynamic moisture tolerance range is then obtained, which is a reasonable tolerance interval considering environmental fluctuations. The calculated original moisture deviation value is divided by the dynamic moisture tolerance range to obtain the normalized moisture loss deviation rate. In actual calculations, half of the tolerance range (i.e., half the width) is usually used as the divisor, so that the absolute value of the deviation rate of 1 corresponds exactly to the tolerance boundary. The final normalized moisture loss deviation rate is a dimensionless value, representing the degree of actual deviation relative to the allowable deviation, facilitating comparison of moisture control accuracy across processes and batches.

[0065] Step 105: Based on the normalized moisture loss deviation rate and moisture deviation transmission coefficient, the inlet moisture benchmark of the downstream unfinished process is corrected and transmitted step by step to obtain the cascaded corrected predicted moisture value of the finished product node.

[0066] In this embodiment, the "cascaded correction predicted moisture value" refers to the moisture content value of the finished product node predicted by the moisture deviation propagation model. It considers the actual moisture deviation of the completed processes and the cumulative effect of these deviations in subsequent processes through the moisture deviation propagation coefficient, and is a prediction result of the final finished product moisture state. For example, by analyzing the moisture deviation state of the currently completed processes and considering the impact of these deviations on subsequent processes, the moisture content of the finished product node can be predicted to be 13.2%, which is 0.3 percentage points lower than the standard value of 13.5%.

[0067] Specifically, the process of predicting cascaded correction of the predicted moisture value first identifies the completed and incomplete processes of the current material batch. For incomplete processes, the baseline inlet moisture value under normal operating conditions is obtained, derived from historical data statistics or process standard specifications. Then, based on the normalized moisture loss deviation rate of the completed processes, the impact of upstream process moisture deviation on downstream process inlet moisture is calculated using a determined moisture deviation transmission coefficient. Starting from the current process, the moisture deviation impact is transmitted downstream level by level, correcting the expected inlet moisture value of each incomplete process. The specific transmission calculation method is to multiply the outlet moisture deviation of the upstream process by the corresponding moisture deviation transmission coefficient to obtain the impact value on the inlet moisture of the downstream process, which is then superimposed with the baseline inlet moisture of the downstream process to form the corrected expected inlet moisture value. This transmission is carried downstream sequentially until the finished product node, ultimately obtaining the cascaded corrected predicted moisture value for the finished product node. During the transmission process, if the corrected inlet moisture of a certain process exceeds the normal fluctuation range, a nonlinear correction factor is introduced into subsequent calculations to reflect the moisture adjustment mechanism under extreme conditions.

[0068] In one possible implementation, based on the normalized moisture loss deviation rate and moisture deviation propagation coefficient, the inlet moisture benchmark of the downstream unfinished process is progressively corrected and propagated to obtain the cascaded corrected predicted moisture value of the finished product node. Specifically, this includes steps 1051-1053, as follows: Step 1051: Use the real-time moisture detection data of the downstream node of the current process as the initial benchmark for the inlet moisture of the next process; use the product of the normalized moisture loss deviation rate, the moisture deviation transmission coefficient and the moisture tolerance range of the current process as the deviation transmission correction amount.

[0069] In this embodiment, the deviation transmission correction amount represents the magnitude of the impact of the current process's moisture control deviation on the next process. It is a correction value calculated based on the product of the normalized moisture loss deviation rate, the moisture deviation transmission coefficient, and the current process's moisture tolerance range. This correction amount reflects the adjustment range by which the moisture control deviation of the upstream process is transmitted to the downstream process, and is used to accurately correct the inlet moisture baseline of the downstream process. For example, when the actual moisture content of the upstream drying process is 0.5% lower than expected, and the moisture deviation transmission coefficient is 0.8, the calculated deviation transmission correction amount may be -0.4%, indicating that the inlet moisture baseline value of the downstream process needs to be lowered by 0.4% to adapt to the actual situation.

[0070] Specifically, firstly, real-time moisture detection data from downstream nodes of the current process is acquired. This data typically comes from moisture detection equipment installed in the process flow, directly recording the real-time moisture content of the material after passing through the current process. This real-time detection data is used as the initial benchmark for the inlet moisture of the next process, establishing data connectivity between processes. Next, the previously calculated normalized moisture loss deviation rate is obtained, a dimensionless index representing the degree of deviation between the predicted and measured moisture values. Simultaneously, a pre-determined moisture deviation transmission coefficient is obtained. This coefficient describes the proportion of the influence of the current process's moisture deviation on the next process; it is usually a value between 0 and 1, reflecting the correlation of moisture transmission between different processes. The moisture tolerance range of the current process is then obtained, which is the allowable range of moisture fluctuation during the process. The normalized moisture loss deviation rate, the moisture deviation transmission coefficient, and the current process's moisture tolerance range are multiplied together. Specifically, the normalized deviation rate is multiplied by the transmission coefficient, and then multiplied by the tolerance range. The result is the deviation transmission correction amount. This correction amount includes directionality and can be positive or negative, representing the magnitude of upward or downward adjustment, respectively.

[0071] Step 1052: Add the deviation transmission correction amount to the initial inlet moisture reference to obtain the corrected inlet moisture reference value for the next process.

[0072] In this embodiment, the corrected inlet moisture reference value for the next process represents the initial moisture reference value of the material in the next process after calibration through the deviation transmission mechanism. It is a more accurate moisture control target obtained by adding the deviation transmission correction amount to the initial inlet moisture reference value. This reference value comprehensively considers the deviation between the actual moisture state of the upstream process and the theoretical expectation, providing a more realistic moisture control starting point for the next process. For example, if the initial inlet moisture reference value for the next process was originally 13.5%, and the deviation transmission correction amount is calculated to be +0.3%, the corrected inlet moisture reference value is adjusted to 13.8% to adapt to the actual moisture deviation transmitted from the upstream process.

[0073] Specifically, a defined initial inlet moisture baseline is obtained, which is an initial reference value set based on real-time moisture detection data from downstream nodes of the current process. Simultaneously, a calculated deviation transmission correction is obtained, which is the product of the normalized moisture loss deviation rate, the moisture deviation transmission coefficient, and the current process's moisture tolerance range. A mathematical addition operation is performed, adding the deviation transmission correction to the initial inlet moisture baseline; that is, adding the initial baseline value to the correction. When the correction is positive, the baseline value is increased; when the correction is negative, the baseline value is decreased. Through this superposition operation, the corrected inlet moisture baseline value for the next process is obtained. This corrected baseline value incorporates the influence of actual moisture deviations from upstream processes and more accurately reflects the actual moisture state of the material entering the next process compared to the original baseline value.

[0074] Step 1053: Based on the modified inlet moisture baseline value of the next process and the target moisture loss mapping relationship of the next process, calculate the cascaded modified prediction value of the next node, and pass it down to the final process to obtain the cascaded modified prediction moisture value of the finished product node.

[0075] In this embodiment, the cascaded correction predicted moisture value represents the final moisture state of the finished product predicted through a step-by-step correction mechanism. It is a high-precision moisture prediction result obtained after considering the impact of moisture deviation transmission between processes. This predicted value is calculated based on the corrected mapping relationship between the moisture at the process inlet and the target moisture loss, and is accumulated step-by-step through the process chain to ultimately form a systematic prediction of the moisture at the finished product node. For example, when the cascaded correction mechanism calculates that the moisture deviation of the tobacco will be transmitted step-by-step between processes, the predicted final moisture content of the finished tobacco is 12.8%, instead of the uncorrected standard prediction of 13.2%, thus providing prediction data closer to the actual production results.

[0076] Specifically, the process begins by obtaining the calculated, corrected inlet moisture baseline value for the next process step. This moisture data already considers the impact of deviation propagation from the previous process. Simultaneously, the target moisture loss mapping relationship for the next process step is obtained; this is a pre-established mathematical model describing the moisture variation pattern of this process. The corrected inlet moisture baseline value for the next process step is substituted into the target moisture loss mapping relationship for that process step to calculate the expected moisture value for the next detection node. This expected value already includes the deviation correction impact from the previous process and is called the cascaded correction prediction value for the next node. A similar calculation process is repeated for all subsequent processes on the production line. The cascaded correction prediction value for the current process is used as the inlet moisture baseline for the next process step, and combined with the target moisture loss mapping relationship for that process step, the cascaded correction prediction value for the next step after that is calculated. This correction prediction mechanism is propagated step-by-step according to the process flow, starting from the first process step, passing through each intermediate process step, until the final process step on the production line. Finally, at the detection node after the final process step, the cascaded correction prediction moisture value for the finished product node is obtained. This moisture value integrates the deviation propagation and cumulative impact between each process step throughout the entire production process.

[0077] Step 106: Calculate the cumulative moisture deviation risk index based on the normalized moisture loss deviation rate of each completed process, determine the finished product prediction deviation by combining the cascaded correction prediction moisture value, and execute a moisture warning based on the cumulative moisture deviation risk index and the finished product prediction deviation.

[0078] In this embodiment, the "cumulative moisture deviation risk index" represents the degree of moisture abnormality risk accumulated in the completed processes of the current batch of materials. It is a quantitative index calculated by comprehensively considering the weighted cumulative moisture deviation of each process, and is used to assess the moisture abnormality risk level of the current production status. For example, a cumulative moisture deviation risk index of 85 points (out of 100) indicates that the current batch has a high risk of moisture abnormality; a finished product prediction deviation of +1.2% indicates that the predicted finished product moisture content will be 1.2 percentage points higher than the standard value.

[0079] Specifically, the process of implementing moisture warning first calculates the cumulative moisture deviation risk index. The normalized moisture loss deviation rate of each completed process is weighted and summed according to the weight of each process's impact on the finished product's moisture content. The weight values ​​are pre-set based on the importance of the process and its position in the production line. The calculated weighted sum is then non-linearly transformed into a risk index value of 0-100, with higher values ​​indicating greater risk. Simultaneously, based on the obtained cascaded correction predicted moisture value, the difference between it and the standard moisture value of the finished product is calculated to obtain the predicted deviation. By setting multi-level warning threshold rules, considering both the cumulative moisture deviation risk index and the predicted deviation, the warning level is determined. For example, when the risk index exceeds 70 and the absolute value of the predicted deviation exceeds 0.8%, a level two warning is triggered. Based on the warning level, corresponding warning information is generated, including anomaly description, affected processes, predicted impact, and recommended measures. Warning signals are then sent to operators through the system interface, audible and visual alarms, or message push notifications, and the warning event is recorded in the system log.

[0080] In the above embodiments, the transmission and cumulative calculation of moisture deviation between processes are realized through a cascaded correction prediction mechanism. To further improve the system's ability to predict abnormal moisture conditions and quantify the comprehensive impact of each process on the final product's moisture content, this application also provides an intelligent monitoring and early warning method for yarn moisture. This method constructs a process evaluation system with historical moisture loss ratios as weights, calculates two key indicators—the cumulative moisture deviation risk index and the predicted deviation of the finished product—in the production process, quantifies the potential risk of the current production state to the finished product's moisture quality, and distinguishes between normal, early warning, and emergency moisture states using a dual threshold judgment criterion, enabling the system to more accurately assess the stability of moisture control during production and the probability of the finished product's moisture content meeting standards. The following section combines... Figure 2 Another intelligent monitoring and early warning method for silk production moisture in the embodiments of this application is described below: Please see Figure 2 This is another flowchart illustrating a method for intelligent monitoring and early warning of moisture content in silk production, as described in this application.

[0081] Step 201: Determine the process weighting coefficient based on the historical average moisture loss ratio of each process; based on the process weighting coefficient, perform a weighted summation of the normalized moisture loss deviation rate of each completed process to obtain the cumulative moisture deviation risk index.

[0082] In this embodiment, the cumulative moisture deviation risk index represents the cumulative risk assessment value of moisture loss control deviation in each process. It is a comprehensive index calculated by weighting and summing the normalized moisture loss deviation rates of each completed process according to the process weight coefficient. This index quantifies the cumulative effect of moisture control deviation in the completed processes throughout the entire production process; the higher the value, the higher the moisture control risk. For example, if the normalized moisture loss deviation rates calculated for the three completed processes of drying fibers, moistening leaves, and blending leaves and fibers on the current production line are 0.2, 0.3, and 0.5, respectively, and the corresponding process weight coefficients are 0.5, 0.3, and 0.2, the calculated cumulative moisture deviation risk index is 0.29, reflecting the overall deviation risk level of moisture control in the current production process.

[0083] Specifically, the process begins by analyzing historical production data for each process to extract its historical average moisture loss. The proportion of moisture loss in the total moisture loss of each process is then calculated, by dividing the historical average moisture loss of a single process by the sum of the historical average moisture losses of all processes. Based on the calculated proportions of moisture loss for each process, corresponding process weighting coefficients are determined; the magnitude of these weighting coefficients is positively correlated with the proportion of moisture loss in each process. The normalized moisture loss deviation rate for each completed process in the current production process is then obtained. These deviation rates are typically dimensionless indicators calculated by comparing the actual moisture loss with the target moisture loss for each process. The normalized moisture loss deviation rate for each completed process is multiplied by its corresponding process weighting coefficient to obtain the weighted deviation value for each process. Finally, the weighted deviation values ​​for all completed processes are summed, calculated as the sum of the products of each process weighting coefficient and its corresponding normalized moisture loss deviation rate, yielding the cumulative moisture deviation risk index. This index comprehensively reflects the cumulative effect of moisture control deviations in completed processes, providing a quantitative basis for subsequent risk assessment.

[0084] Step 202: Calculate the difference between the cascaded correction predicted moisture value and the median value of the finished product process moisture standard, and divide the difference by the half-width of the finished product moisture qualified range to obtain the finished product prediction deviation.

[0085] In this embodiment, the predicted deviation of the finished product represents the degree of normalization of the predicted moisture content of the finished product from the standard median. It is a dimensionless index calculated by dividing the difference between the cascaded correction predicted moisture content and the standard median of the finished product's process moisture content by the half-width of the acceptable moisture content range. This deviation measures the relative position of the predicted moisture content within the acceptable process moisture content range. When the absolute value exceeds 1, it indicates that the predicted moisture content has exceeded the acceptable range. For example, when the standard range for the finished product's process moisture content is 12.5% ​​± 0.5%, and the cascaded correction predicted moisture content is 13.3%, the calculated predicted deviation is 1.6, indicating that the predicted moisture content exceeds the upper limit of the standard by 0.3 percentage points, which translates to a normalized deviation of 1.6 units.

[0086] Specifically, first, the predicted moisture content of the finished product node, calculated through a cascaded correction mechanism, is obtained; this value reflects the expected moisture state of the finished product after considering the transmission effect of moisture deviations between processes. Simultaneously, the median of the finished product's process moisture standard is obtained, which is the ideal moisture content specified in the product quality standard. The difference between the cascaded correction predicted moisture content and the median of the finished product's process moisture standard is calculated; that is, the cascaded correction predicted moisture content is subtracted from the median, resulting in a difference with a positive or negative sign. A positive value indicates that the predicted moisture content is higher than the standard, and a negative value indicates that the predicted moisture content is lower than the standard. The half-width of the acceptable moisture range for the finished product is obtained; this is half of the allowable moisture fluctuation range specified in the product quality standard. The previously calculated difference is divided by the half-width of the acceptable moisture range for the finished product. This division operation normalizes the difference, making the moisture control status comparable between different production lines or different products. Through this calculation, the predicted deviation of the finished product is obtained, a dimensionless relative indicator representing the degree to which the predicted moisture value deviates from the standard center.

[0087] Step 203: Determine whether the cumulative moisture deviation risk index or the finished product prediction deviation exceeds the corresponding critical warning threshold; if at least one of the cumulative moisture deviation risk index or the finished product prediction deviation exceeds the corresponding critical warning threshold, then trigger a moisture warning signal.

[0088] In this embodiment, the moisture warning signal represents a warning notification triggered by the system based on a moisture control risk assessment. It is an automatic reminder generated by the system when the cumulative moisture deviation risk index or the predicted deviation of the finished product exceeds the corresponding critical warning threshold. This signal is used to promptly alert operators to the risk of abnormal moisture control and the need to take appropriate control measures to prevent the finished product from failing the moisture test. For example, when the cumulative moisture deviation risk index reaches 0.8 (exceeding the set critical warning threshold of 0.75), the system will trigger a level-one moisture warning signal, indicating to the production line operators that there has been a significant deviation in moisture control and suggesting appropriate adjustments to subsequent process parameters.

[0089] Specifically, firstly, the calculated cumulative moisture deviation risk index is obtained, which is a quantitative indicator reflecting the cumulative effect of moisture control deviations in completed processes. Simultaneously, the calculated finished product prediction deviation is obtained, which is a standardized indicator representing the position of the predicted moisture value relative to the standard acceptable range. The system-preset critical warning threshold for the cumulative moisture deviation risk index is obtained, which is the upper limit of risk tolerance determined based on production experience or statistical analysis. Simultaneously, the system-preset critical warning threshold for the finished product prediction deviation is obtained, which is the upper limit of allowable deviation determined based on product quality requirements. The cumulative moisture deviation risk index is compared with its corresponding critical warning threshold to determine if any threshold is exceeded. The finished product prediction deviation is also compared with its corresponding critical warning threshold to determine if any threshold is exceeded. If the cumulative moisture deviation risk index exceeds the corresponding critical warning threshold, or the finished product prediction deviation exceeds the corresponding critical warning threshold, the conditions for triggering a warning are met. Under the condition of meeting the warning triggering conditions, a moisture warning signal is generated and sent. This signal may contain information such as the risk type, risk level, and involved processes to guide subsequent intervention measures.

[0090] In one possible implementation, a moisture warning signal is triggered. Following this step, steps 2031-2033 are further included, as follows: Step 2031: Obtain the measured moisture value of the finished product at the finished product node for the current batch, and calculate the terminal residual between the measured moisture value of the finished product and the cascaded correction predicted moisture value.

[0091] In this embodiment, the terminal residual represents the difference between the actual measured moisture value of the finished product and the moisture value predicted by the cascade correction mechanism. This residual reflects the accuracy of the moisture prediction model and the impact of uncertainties in the actual production process, serving as important reference data for adjusting and optimizing the moisture prediction model. For example, when the actual measured moisture value of the finished product is 13.2%, while the cascade correction predicts a moisture value of 13.5%, the terminal residual is -0.3 percentage points, indicating that the actual moisture value is lower than the predicted value.

[0092] Specifically, the actual moisture content data of the current batch at the final finished product node of the production line is first collected using sensors or detection equipment to obtain the measured moisture value of the finished product. Simultaneously, the predicted moisture value of the finished product, previously calculated through a cascaded correction mechanism, is extracted from the system; this is the cascaded corrected predicted moisture value. The difference between the measured moisture value and the cascaded corrected predicted moisture value is calculated, i.e., the measured moisture value is subtracted from the cascaded corrected predicted moisture value, yielding the terminal residual. The terminal residual can be positive or negative; a positive value indicates that the measured value is higher than the predicted value, and a negative value indicates that the measured value is lower than the predicted value. This terminal residual will serve as the basis for adjusting the moisture control parameters in subsequent steps.

[0093] Step 2032: Distribute the terminal residuals according to the proportion of the normalized moisture loss deviation rate of each completed process to obtain the residual components of each process; use the residual components to compensate for the target moisture loss mapping relationship of each process to generate the updated moisture loss mapping relationship.

[0094] In this application embodiment, the moisture loss mapping relationship represents the correspondence model between the process parameter settings and the actual moisture loss in each process step. It is the core data model for predicting and controlling moisture changes in each process step. This mapping relationship describes the expected moisture loss under specific process parameter conditions and can be used to guide the adjustment of process parameters to achieve target moisture control. For example, the moisture loss mapping relationship in the yarn drying process may indicate that when the yarn drying temperature is 85°C and the air velocity is 2.5 m / s, the expected moisture loss rate is 2.3%. Based on this, the process parameter settings required to achieve the target moisture level can be derived in reverse.

[0095] Specifically, firstly, the proportion of the normalized moisture loss deviation rate of each completed process in all completed processes is calculated, that is, the normalized moisture loss deviation rate of a single process is divided by the sum of the normalized moisture loss deviation rates of all completed processes. Based on the calculated proportion of the normalized moisture loss deviation rate of each process, the obtained terminal residuals are proportionally allocated to each completed process. The calculation method is to multiply the terminal residual by the corresponding proportion of the normalized moisture loss deviation rate of the process to obtain the residual component of each process. For each completed process, the target moisture loss mapping relationship currently in use is obtained, which is model data describing the relationship between process parameters and moisture loss. The residual components of each process are used as compensation amounts to adjust the target moisture loss mapping relationship of the corresponding process. The adjustment method can be to add or subtract the moisture value corresponding to the residual component based on the original mapping relationship, or to modify it through a more complex functional relationship. Through the above adjustments, an updated moisture loss mapping relationship is generated, which will more accurately reflect the actual correspondence between process parameters and moisture loss in each process.

[0096] Step 2033: Based on the cumulative moisture deviation risk index when the moisture warning is triggered and the updated moisture loss mapping relationship, dynamically adjust the moisture deviation transmission coefficient between adjacent processes.

[0097] In this embodiment, the moisture deviation transmission coefficient represents a quantitative parameter indicating the transmission effect of moisture control deviations between adjacent processes. It describes the degree of influence of moisture deviations in upstream processes on moisture control in downstream processes. This coefficient is a key parameter in the cascaded moisture control model and is typically expressed as a scaling factor, reflecting the amplification or attenuation effect of moisture deviations between processes. For example, when the moisture deviation transmission coefficient between the leaf-moistening process and the yarn-drying process is 1.2, it means that a 1% moisture deviation in the leaf-moistening process will lead to a 1.2% moisture deviation in the yarn-drying process. A coefficient greater than 1 indicates that the deviation is amplified, while a coefficient less than 1 indicates that the deviation is attenuated.

[0098] Specifically, firstly, the cumulative moisture deviation risk index recorded when a moisture warning is triggered is obtained. This is a comprehensive indicator reflecting the overall deviation risk of moisture control in the current production process. Simultaneously, the updated moisture loss mapping relationships for each process are obtained. These updated mapping relationships reflect the correspondence between the latest process parameters and moisture loss. The magnitude and trend of the cumulative moisture deviation risk index are analyzed to determine the severity and distribution of moisture control deviations. Based on the updated moisture loss mapping relationships, the moisture deviation transmission characteristics between adjacent processes are reassessed, including transmission direction and intensity. Based on the above analysis results, a new moisture deviation transmission coefficient between adjacent processes is calculated. The adjustment method can be to proportionally adjust the original transmission coefficient according to the magnitude of the cumulative moisture deviation risk index, or to recalculate the transmission coefficient based on a more complex algorithm. The calculated new moisture deviation transmission coefficient is updated in the system for subsequent cascaded moisture prediction and control processes.

[0099] The following describes a silk-making moisture intelligent monitoring and early warning system according to an embodiment of the present invention from the perspective of hardware processing. Please refer to [link / reference]. Figure 3 This is a schematic diagram of the structure of a silk-making moisture intelligent monitoring and early warning system in an embodiment of this application.

[0100] It should be noted that, Figure 3 The structure of the intelligent monitoring and early warning system for moisture content in silk production shown is merely an example and should not impose any limitations on the functionality and scope of application of the embodiments of the present invention.

[0101] like Figure 3As shown, a silk-making moisture intelligent monitoring and early warning system includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage section 308 into a random access memory (RAM) 303, such as executing the methods described in the above embodiments. The RAM 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0102] The following components are connected to I / O interface 305: input section 306 including audio input devices, push-button switches, etc.; output section 307 including a liquid crystal display (LCD) and audio output devices, indicator lights, etc.; storage section 308 including a hard disk, etc.; and communication section 309 including a network interface card such as a LAN (Local Area Network) card, modem, etc. Communication section 309 performs communication processing via a network such as the Internet. Drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0103] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by central processing unit (CPU) 301, it performs the various functions defined in the present invention.

[0104] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0105] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0106] Specifically, the intelligent monitoring and early warning system for silk production moisture in this embodiment includes a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the intelligent monitoring and early warning method for silk production moisture provided in the above embodiment.

[0107] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the intelligent monitoring and early warning system for silk production moisture described in the above embodiments; or it may exist independently and not assembled into the intelligent monitoring and early warning system for silk production moisture. The storage medium carries one or more computer programs, which, when executed by a processor of the intelligent monitoring and early warning system for silk production moisture, enable the system to implement the intelligent monitoring and early warning method for silk production moisture based on encrypted data transmission via the Internet of Things provided in the above embodiments.

Claims

1. A method for intelligent monitoring and early warning of moisture content in silk processing, characterized in that, The method includes: Acquire sample moisture detection data from multiple adjacent moisture detection nodes on the silk production line and sample process parameters for each process. Based on the sample moisture detection data and the sample process parameters, determine the process parameter set for each process and the moisture deviation transfer coefficient between adjacent processes; Based on the set of process parameters, an initial moisture loss mapping relationship for each process is constructed, and the initial moisture loss mapping relationship is corrected by combining the historical storage time and historical environmental parameters between processes to obtain the target moisture loss mapping relationship for each process. The real-time moisture detection data, real-time process parameters, real-time storage time and real-time environmental parameters of the current batch of materials are obtained, and the real-time process parameters, real-time storage time and real-time environmental parameters are substituted into the target moisture loss mapping relationship, and the normalized moisture loss deviation rate of the current process is calculated in combination with the real-time moisture detection data. Based on the normalized moisture loss deviation rate and the moisture deviation transmission coefficient, the inlet moisture benchmark of the downstream unfinished process is corrected and transmitted step by step to obtain the cascaded corrected predicted moisture value of the finished product node. The cumulative moisture deviation risk index is calculated based on the normalized moisture loss deviation rate of each completed process. The finished product prediction deviation is determined by combining the cascaded correction prediction moisture value. Moisture warning is then executed based on the cumulative moisture deviation risk index and the finished product prediction deviation.

2. The method according to claim 1, characterized in that, The step of determining the process parameter set for each process and the moisture deviation transfer coefficient between adjacent processes based on the sample moisture detection data and the sample process parameters includes: Based on the sample moisture detection data, calculate the water loss between adjacent moisture detection nodes, and calculate the correlation coefficient between the water loss and each sample process parameter; The process parameters of the samples whose correlation coefficient values ​​are greater than a preset correlation threshold are extracted and combined to form the sample process parameter set. Extract the measured moisture values ​​of the same production batch at adjacent moisture detection nodes, and calculate the moisture deviation between the measured moisture values ​​and the historical overall mean of the corresponding nodes; Calculate the ratio of moisture deviation values ​​between adjacent processes, and use the ratio as the moisture deviation transfer coefficient between the adjacent processes.

3. The method according to claim 2, characterized in that, The initial moisture loss mapping relationship for each process is constructed based on the process parameter set, and the initial moisture loss mapping relationship is corrected by combining the historical storage time and historical environmental parameters between processes to obtain the target moisture loss mapping relationship for each process, including: Using the parameters in the process parameter set as independent variables and the sample water loss as the dependent variable, regression fitting is performed to obtain the initial water loss mapping relationship; The temporary storage correction term is obtained by fitting the historical temporary storage time. The environmental correction term is obtained by fitting the historical environmental parameters. The temporary correction term and the environmental correction term are superimposed on the initial water loss mapping relationship to obtain the target water loss mapping relationship.

4. The method according to claim 1, characterized in that, The step of substituting the real-time process parameters, the real-time storage time, and the real-time environmental parameters into the target moisture loss mapping relationship, and calculating the normalized moisture loss deviation rate of the current process in conjunction with the real-time moisture detection data, includes: Substituting the real-time process parameters, the real-time storage time, and the real-time environmental parameters into the target moisture loss mapping relationship, the downstream node moisture prediction benchmark value of the current process is calculated. The instantaneous flow rate of the material is extracted from the real-time process parameters, and the measurement delay of the moisture detection probe is compensated based on the instantaneous flow rate of the material to correct the real-time moisture detection data to the target measured moisture value. Extract the temperature and humidity fluctuation characteristics from the real-time environmental parameters, calculate the environmental penalty coefficient based on the temperature and humidity fluctuation characteristics, and use the environmental penalty coefficient to dynamically scale the initial moisture tolerance range of the current process to obtain the dynamic moisture tolerance range. The deviation between the predicted moisture baseline value of the downstream node and the measured target moisture value is calculated, and the deviation is divided by the dynamic moisture tolerance range to obtain the normalized moisture loss deviation rate.

5. The method according to claim 1, characterized in that, The step of progressively correcting and propagating the inlet moisture baseline of downstream incomplete processes based on the normalized moisture loss deviation rate and the moisture deviation propagation coefficient to obtain the cascaded corrected predicted moisture value of the finished product node includes: The real-time moisture detection data of the downstream node of the current process is used as the initial benchmark for the inlet moisture of the next process. The product of the normalized moisture loss deviation rate, the moisture deviation transmission coefficient, and the moisture tolerance range of the current process is used as the deviation transmission correction amount. The deviation transmission correction amount is superimposed on the initial inlet moisture reference to obtain the corrected inlet moisture reference value for the next process. Based on the modified inlet moisture baseline value of the next process and the target moisture loss mapping relationship of the next process, the cascaded modified prediction value of the next node is calculated and passed down to the final process to obtain the cascaded modified prediction moisture value of the finished product node.

6. The method according to claim 1, characterized in that, The step of calculating the cumulative moisture deviation risk index based on the normalized moisture loss deviation rate of each completed process, determining the finished product prediction deviation by combining the cascaded correction predicted moisture value, and executing a moisture warning based on the cumulative moisture deviation risk index and the finished product prediction deviation includes: The process weighting coefficient is determined based on the historical average moisture loss percentage of each process. The cumulative moisture deviation risk index is obtained by weighting and summing the normalized moisture loss deviation rates of each completed process based on the process weight coefficients. Calculate the difference between the cascaded correction predicted moisture value and the median of the finished product process moisture standard, and divide the difference by the half-width of the finished product moisture acceptable range to obtain the finished product prediction deviation. Determine whether the cumulative moisture deviation risk index or the finished product prediction deviation exceeds the corresponding critical warning threshold. If at least one of the cumulative moisture deviation risk index or the finished product prediction deviation exceeds the corresponding critical warning threshold, a moisture warning signal is triggered.

7. The method according to claim 6, characterized in that, After triggering the moisture warning signal, the following is also included: Obtain the measured moisture value of the finished product at the finished product node for the current batch, and calculate the terminal residual between the measured moisture value of the finished product and the cascaded correction predicted moisture value; The terminal residuals are allocated according to the proportion of the normalized moisture loss deviation rate of each completed process to obtain the residual components of each process. The residual components are used to compensate for the target moisture loss mapping relationship of each process, and an updated moisture loss mapping relationship is generated. Based on the cumulative moisture deviation risk index when the moisture warning is triggered and the updated moisture loss mapping relationship, the moisture deviation transmission coefficient between adjacent processes is dynamically adjusted.

8. A smart monitoring and early warning system for moisture content in silk production, characterized in that, The intelligent monitoring and early warning system for silk production moisture includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the intelligent monitoring and early warning system for silk production moisture to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the intelligent monitoring and early warning system for silk production moisture, the intelligent monitoring and early warning system for silk production moisture performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on the intelligent monitoring and early warning system for silk production moisture, the intelligent monitoring and early warning system for silk production moisture performs the method as described in any one of claims 1-7.

Images