A tobacco curing room intelligent temperature and humidity adjusting method and system

Abstract

The present application relates to the field of tobacco processing and baking technology, and discloses a method and system for intelligently adjusting temperature and humidity in a tobacco baking room. The method comprises collecting temperature and humidity data in multiple regions of the baking room and external weather data, removing outliers and performing smoothing filtering to obtain a standardized distribution data set; constructing a temperature and humidity interaction matrix to calculate a coupling index and classify, quantifying regional deviation and determining the degree of imbalance when the threshold value is exceeded; fusing environmental disturbance data to generate global imbalance distribution information and generate partition compensation instructions to dynamically control the parameters of ventilation equipment; combining real-time feedback and equipment delay to predict humidity trends, and closing the loop to optimize parameters to achieve stable and balanced temperature and humidity. The present application solves the problems of traditional single-point control, poor adaptability of fixed parameters, and temperature and humidity coupling imbalance. The present method can achieve precise partition control of the global baking room, effectively ensuring the consistency of tobacco baking quality and the stability of working conditions.

Description

Technical Field

[0001] This invention relates to the field of tobacco processing and curing technology, and in particular to an intelligent temperature and humidity control method and system for tobacco curing chambers. Background Technology

[0002] Currently, in the field of tobacco processing and curing technology, precise control of temperature and humidity in the curing environment is a key aspect of ensuring tobacco leaf quality and avoiding curing losses, and has become a major research focus in the field.

[0003] In one existing technology, temperature and humidity regulation in the curing chamber is mainly achieved through manual adjustment or single-parameter control. This relies primarily on fixed parameter settings or simple threshold triggers, such as using preset temperature and humidity values ​​to control the start and stop of equipment, and judging the curing status through single-point monitoring data. However, this approach is clearly insufficient in complex operating environments. Fixed parameters cannot adapt to dynamic fluctuations in external weather, and single control struggles to balance the coupled effects of temperature and humidity with overall distribution balance. Affected by regional environmental differences, changes in tobacco leaf stages, and airflow disturbances, it is prone to insufficient control precision and imbalanced temperature and humidity distribution during the curing process. Furthermore, it lacks in-depth analysis of coupling relationships and the ability to dynamically adjust zones, especially in complex scenarios involving multi-zone curing and continuous operations. It is difficult to quickly correct temperature and humidity deviations and stabilize the curing status, resulting in uneven tobacco leaf curing quality.

[0004] In summary, existing technologies cannot achieve precise control of temperature and humidity throughout the tobacco curing chamber, and cannot meet the core requirements of tobacco processing and curing technology for stability of curing process control and consistency of tobacco leaf quality when facing complex working conditions. Summary of the Invention

[0005] This invention provides a method and system for intelligent temperature and humidity control in tobacco curing rooms, so as to achieve precise control of temperature and humidity throughout the tobacco curing room and meet the core requirements of tobacco processing and curing technology for stability of curing process control and consistency of tobacco quality when facing complex working conditions.

[0006] In a first aspect, to solve the above-mentioned technical problems, the present invention provides an intelligent temperature and humidity control method for a tobacco curing chamber, comprising:

[0007] Acquire raw temperature and humidity data for multiple areas within the baking chamber, as well as external weather fluctuation data;

[0008] The raw temperature and humidity data are filtered to obtain a temperature and humidity distribution dataset.

[0009] A temperature and humidity interaction matrix is ​​constructed based on the distributed dataset. The coupling index of the temperature and humidity interaction matrix is ​​calculated. The coupling index is divided by a preset grading threshold to obtain the coupling strength level.

[0010] If the coupling strength level exceeds the preset coupling judgment threshold, then regional deviation data within the distributed dataset is extracted, the regional deviation data is weighted and calculated to obtain a comprehensive deviation value, and the comprehensive deviation value is compared with the preset imbalance judgment threshold to obtain the degree of temperature and humidity imbalance.

[0011] By integrating the external weather fluctuation data, the regional deviation data, and the degree of temperature and humidity imbalance, temperature and humidity imbalance distribution information is generated.

[0012] Based on the temperature and humidity imbalance distribution information, a compensation instruction for the humidity of each zone is generated. The ventilation equipment parameters of the baking room are adjusted according to the compensation instruction to obtain a temperature and humidity balance state.

[0013] Real-time feedback data is extracted based on the temperature and humidity balance state, and the parameters of the ventilation equipment are updated according to the real-time feedback data to obtain a stable temperature and humidity distribution.

[0014] Secondly, the present invention provides an intelligent temperature and humidity control system for a tobacco curing chamber, comprising:

[0015] The data acquisition module is used to acquire raw temperature and humidity data from multiple areas within the baking chamber, as well as external weather fluctuation data.

[0016] The data filtering module is used to filter the raw temperature and humidity data to obtain a temperature and humidity distribution dataset.

[0017] The coupling analysis module is used to construct a temperature and humidity interaction matrix based on the distributed dataset, calculate the coupling index of the temperature and humidity interaction matrix, and divide the coupling index by a preset grading threshold to obtain the coupling strength level.

[0018] The imbalance determination module is used to extract regional deviation data within the distributed dataset if the coupling strength level exceeds a preset coupling determination threshold, perform weighted calculation on the regional deviation data to obtain a comprehensive deviation value, and compare the comprehensive deviation value with the preset imbalance determination threshold to obtain the degree of temperature and humidity imbalance.

[0019] The data fusion module is used to fuse the external weather fluctuation data, the regional deviation data, and the degree of temperature and humidity imbalance to generate temperature and humidity imbalance distribution information;

[0020] The parameter control module is used to generate a compensation instruction for the humidity of the zone based on the temperature and humidity imbalance distribution information, and adjust the ventilation equipment parameters of the baking room according to the compensation instruction to obtain a temperature and humidity balance state.

[0021] The feedback optimization module is used to extract real-time feedback data based on the temperature and humidity balance state, update the ventilation equipment parameters according to the real-time feedback data, and obtain a stable temperature and humidity distribution.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] (1) This invention collects raw temperature and humidity data from multiple areas of the baking room and external weather fluctuation data, aligns time-series nodes to complete data synchronization, and generates a temperature and humidity distribution dataset through anomaly removal and smoothing filtering. It breaks through the limitations of traditional single-point monitoring and fixed parameter control, explores the spatiotemporal distribution characteristics of temperature and humidity in the entire baking room, eliminates the noise impact of data collection jumps and airflow disturbances, provides high-precision basic data support for precise control, effectively improves the reliability of multi-area temperature and humidity data collection, and solves the problems of data distortion and incomplete monitoring coverage in traditional methods.

[0024] (2) This invention calculates the coupling index and divides the coupling strength level by constructing a temperature and humidity interaction matrix. When the threshold is exceeded, regional deviation data is extracted to determine the degree of imbalance. External weather interference is integrated to generate temperature and humidity imbalance distribution information. This invention breaks through the limitation that traditional single parameter control cannot adapt to the influence of temperature and humidity coupling. It accurately captures the correlation characteristics between temperature and humidity interaction and environmental disturbance, provides multi-dimensional basis for imbalance judgment, significantly improves the accuracy of temperature and humidity imbalance identification under complex working conditions, and makes up for the lack of targeted control of existing technologies.

[0025] (3) Based on the imbalance distribution information, the present invention generates zonal compensation instructions to adjust the parameters of the ventilation equipment. It combines real-time feedback data and equipment delay to predict humidity trends and dynamically update parameters, thereby achieving temperature and humidity balance and stable distribution throughout the curing room. This breaks through the limitations of traditional methods that lack zonal dynamic control and closed-loop feedback optimization, providing a basis for balanced temperature and humidity control throughout the curing room. It solves the problems of regional temperature and humidity imbalance and uneven curing quality, while taking into account both control accuracy and operational stability, and meeting the core requirements of tobacco processing and curing for consistent tobacco quality. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of a method for intelligent temperature and humidity control in a tobacco curing room according to the first embodiment of the present invention;

[0027] Figure 2 This is a schematic diagram of an intelligent temperature and humidity control system for a tobacco curing room provided in the second embodiment of the present invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Reference Figure 1 The first embodiment of the present invention provides an intelligent temperature and humidity control method for a tobacco curing room, comprising the following steps:

[0030] S101, raw temperature and humidity data for multiple areas inside the room and external weather fluctuation data;

[0031] S102 filters the raw temperature and humidity data to obtain a temperature and humidity distribution dataset;

[0032] S103 constructs a temperature and humidity interaction matrix based on the distributed dataset, calculates the coupling index of the temperature and humidity interaction matrix, and divides the coupling index by a preset grading threshold to obtain the coupling strength level;

[0033] S104 If the coupling strength level exceeds the preset coupling judgment threshold, then extract the regional deviation data in the distributed dataset, perform weighted calculation on the regional deviation data to obtain a comprehensive deviation value, and compare the comprehensive deviation value with the preset imbalance judgment threshold to obtain the degree of temperature and humidity imbalance.

[0034] S105, integrate the external weather fluctuation data, the regional deviation data and the degree of temperature and humidity imbalance to generate temperature and humidity imbalance distribution information;

[0035] S106, Based on the temperature and humidity imbalance distribution information, generate a compensation instruction for the humidity of the zone, and adjust the ventilation equipment parameters of the baking room according to the compensation instruction to obtain a temperature and humidity balance state;

[0036] S107, extract real-time feedback data based on the temperature and humidity balance state, update the ventilation equipment parameters according to the real-time feedback data, and obtain a stable temperature and humidity distribution.

[0037] In step S101, acquiring raw temperature and humidity data for multiple areas within the baking chamber and external weather fluctuation data includes:

[0038] Collect temperature and humidity data at various monitoring points inside the baking oven;

[0039] Align the timeline nodes of the temperature acquisition data and the humidity acquisition data to obtain time-synchronized data;

[0040] By integrating the aforementioned time-series synchronized data, raw temperature and humidity data for multiple areas within the baking chamber are generated.

[0041] Collect ambient temperature and humidity data outside the baking room, and integrate the ambient temperature and humidity data to obtain external weather fluctuation data.

[0042] It should be noted that, firstly, temperature and humidity data are collected from various monitoring points within the baking chamber. Data acquisition is accomplished using an integrated temperature and humidity sensor array. The arrangement of this sensor array follows a pre-defined quantitative specification: For conventional baking chambers with a volume of 50 cubic meters or less, no fewer than 11 integrated temperature and humidity sensors are configured. Nine of these sensors are positioned at heights of 0.5 meters, 1.5 meters, and 2.5 meters above the ground, forming three monitoring layers (lower, middle, and upper). Each layer has three monitoring points evenly distributed along the longitudinal direction of the chamber, with the horizontal spacing between sensors within the same layer maintained between 1.5 meters and 2.0 meters. For large baking chambers with a volume exceeding 50 cubic meters, an additional row (containing three sensors) is added for every 20 cubic meters increase in volume, proportionally increasing the space ratio. In addition, one sensor is independently installed at the center of each air inlet and outlet cross-section to monitor boundary airflow. The sampling frequency was set based on statistical analysis of environmental changes during the tobacco curing process over the past year. The basic sampling frequency was set at 5 seconds per sampling, increased to 2 seconds per sampling during the high-precision curing stage, and decreased to 10 seconds per sampling during the stable curing stage. For example, multiple sets of temperature and humidity sensors were deployed in the curing chamber, and the temperature and humidity values ​​at each point were collected synchronously every 5 seconds.

[0043] Next, the timeline nodes of the temperature and humidity data acquisitions are aligned to obtain time-synchronized data. Time alignment uses a timestamp matching method, based on the system's unified clock, to pair temperature and humidity values ​​acquired at the same time, eliminating time-disordered data caused by transmission delays. For example, the temperature value acquired at the 5th second is paired with the humidity value acquired at the same time to generate time-synchronized data.

[0044] Subsequently, the time-series synchronized data is integrated to generate raw temperature and humidity data for multiple areas within the baking chamber. The integration process is categorized and summarized according to the sensor's location, combining time-series synchronized data from the same area into regional data blocks. After combining all regional data blocks, raw temperature and humidity data covering the entire baking chamber is generated.

[0045] The system collects ambient temperature and humidity data from outside the baking oven and integrates them to obtain external weather fluctuation data. External data is acquired by outdoor temperature and humidity sensors deployed outside the baking oven (such as on the roof or surrounding walls). These outdoor sensors collect the temperature and humidity values ​​of the environment surrounding the baking oven in real time, with a collection frequency consistent with that of the sensors inside the oven, ensuring that the data from inside and outside correspond in time.

[0046] In step S102, the filtering process performed on the raw temperature and humidity data to obtain a temperature and humidity distribution dataset includes:

[0047] The abnormal values ​​in the original temperature and humidity data are removed to obtain the cleaned data;

[0048] The cleaned data is smoothed and integrated to obtain a temperature and humidity distribution dataset.

[0049] It should be noted that, firstly, outliers in the original temperature and humidity data are removed to obtain cleaned data. Outlier identification employs the 3σ statistical discrimination method. The outlier threshold is set based on the statistical analysis of temperature and humidity data from the steady-state operation of tobacco curing barns over the past year. The basic threshold corresponds to a three-standard-deviation interval; for high-precision curing scenarios involving tobacco leaf color fixing, this can be tightened to a two-standard-deviation interval; and for routine curing scenarios involving yellowing tobacco leaves, it can be widened to a four-standard-deviation interval. This threshold adapts to the data characteristics of different curing stages, stably filtering out data with abrupt changes in acquisition and transmission distortion. For missing data points resulting from outlier removal, the system uses a timestamp-based linear interpolation method for continuity filling. This involves calculating the slope of change using two adjacent valid data points before and after the missing location, and then estimating the replacement value at the missing time to ensure the integrity and physical continuity of the time series. The filled valid values ​​are mapped to the 0-1 interval using the min-max normalization method. For example, by filtering out humidity values ​​with sudden changes at a certain point and performing linear interpolation filling, continuous and stable valid data is retained, completing the outlier cleaning of the entire temperature and humidity data.

[0050] The cleaned data was smoothed and integrated to obtain a dataset of temperature and humidity distributions. The smoothing process employed a simple moving average (SMA) filtering algorithm with a single-step sliding step. This means the window moves forward one sampling point along the time axis each time, and the arithmetic mean of the data within the window is calculated to generate a new smoothed value. The filter window size was set based on the airflow disturbance characteristics of the drying chamber. The basic window was set to 3 sampling points, which could be expanded to 5 sampling points in the strong airflow region at the air inlet, and reduced to 2 sampling points in the closed steady-state region. For boundary issues at the beginning and end of the data sequence where direct calculation was impossible due to insufficient window coverage, an edge value duplication and filling rule was adopted. This involved copying and filling the missing preceding data at the beginning of the sequence with the first valid value, and copying and filling the missing following data at the end of the sequence with the last valid value. The smoothing operation weakened random noise fluctuations while preserving the true trend of temperature and humidity changes.

[0051] Finally, all regional data were integrated according to spatial layout to form a distributed dataset that is unified in time and space. The specific rules for data integration are as follows: using a unified sampling timestamp as the global alignment standard, at each time segment, the discrete smooth data of each point are sequentially spliced ​​and arranged into a spatial feature vector according to the three-dimensional physical coordinates of the sensor in the baking chamber (in the physical order from the bottom layer to the top layer and from the air inlet side to the air outlet side); then, all spatial feature vectors at continuous time steps are combined in time sequence to finally construct a feature matrix containing time dimension and three-dimensional spatial dimension, which serves as a complete temperature and humidity distribution dataset.

[0052] In step S103, the step of constructing a temperature and humidity interaction matrix based on the distributed dataset, calculating the coupling index of the temperature and humidity interaction matrix, and dividing the coupling index by a preset grading threshold to obtain the coupling strength level includes:

[0053] Temperature and humidity feature data are extracted from the aforementioned distribution dataset.

[0054] A temperature and humidity interaction matrix is ​​constructed based on the temperature feature data and the humidity feature data;

[0055] The coupling index is obtained by analyzing the characteristic values ​​of the temperature and humidity interaction matrix and normalizing the characteristic values.

[0056] The coupling index is compared with a preset grading threshold to obtain the coupling strength level.

[0057] It should be noted that temperature and humidity feature data are extracted from the distributed dataset. Feature extraction covers two types of indicators: real-time values ​​and short-term rates of change at each monitoring point. All feature values ​​use the previously completed minimum-maximum normalization results, keeping the data range uniform between 0 and 1. The feature data is arranged by region and time sequence, and can be directly used for matrix construction operations. For example, the real-time temperature values ​​and rates of change at the middle layer points of the baking room are extracted, and the corresponding humidity features are extracted simultaneously, forming standardized two-dimensional feature data.

[0058] Next, a temperature and humidity interaction matrix is ​​constructed based on the temperature and humidity feature data. The matrix construction employs a triangular membership function for fuzzification. Specifically, the temperature and humidity feature data, normalized to the 0-1 interval, are first divided into fuzzy subsets: low (L), moderate (M), and high (H). The parameters of the triangular membership function are strictly limited as follows: the vertex coordinates for low (L) are [0, 0, 0.5], for moderate (M) are [0, 0.5, 1.0], and for high (H) are [0.5, 1.0, 1.0].

[0059] Subsequently, feature association calculations are performed using a pre-defined fuzzy rule base. The core IF-THEN rule entries contain nine items, with representative rules such as: if the temperature is high (H) and the humidity is low (L), the interaction is strong (S); if the temperature is moderate (M) and the humidity is moderate (M), the interaction is weak (W); and if the temperature is high (H) and the humidity is high (H), the interaction is extremely strong (VS). The fuzzy inference process uses the Mamdani Max-Min inference method to calculate the activation intensity of each rule. Finally, the centroid method is used for defuzzification, transforming the inferred fuzzy set into specific, clear, quantified values, which are then filled into the corresponding matrix elements. The generated matrix rows and columns correspond to the temperature and humidity feature dimensions, respectively, and the matrix elements accurately represent the degree of interaction between the two types of parameters. The structure is fully adapted to the coupling analysis requirements of multiple areas in the baking chamber.

[0060] Subsequently, the feature values ​​of the temperature and humidity interaction matrix are analyzed, and the coupling index is obtained after normalization. Feature extraction employs a principal eigenvalue algorithm, extracting the largest eigenvalue of the matrix as the basic data. Then, a min-max normalization method is used to map the values ​​to the interval between 0 and 1, ultimately yielding a coupling index that quantifies the degree of temperature and humidity correlation; a higher value indicates a stronger coupling effect. For example, solving for the principal eigenvalue of the interaction matrix and performing normalization yields a coupling index of 0.78, quantifying the current temperature and humidity coupling strength.

[0061] It is worth noting that the specific process for setting the preset coupling grading threshold is as follows: extract 10,000 sets of temperature and humidity coupled operation samples and corresponding system stability adjustment times from the comprehensive historical database of tobacco curing over the past year (it should be noted that this database is the only historical data source that is uniformly called by the system globally, and it is the same database object as the historical database mentioned below used for feature importance assessment, steady-state operation determination, etc.).

[0062] The construction specifications for the comprehensive historical database of tobacco curing throughout the entire cycle are as follows: In terms of sample selection rules, the isolated forest algorithm is used to remove abnormal missing samples caused by sensor disconnection or power failure, and only complete closed-loop operation data with a single curing duration of no less than 120 hours are retained; In terms of database field composition, each sample record must include six major categories of fields: timestamp, measured temperature and humidity sequence in multiple regions, external environmental meteorological parameters, output power of ventilation and heating equipment at the corresponding time, system stabilization adjustment time, and the final comprehensive score of tobacco leaf quality; In terms of sample operating conditions, it is mandatory that the extracted samples must evenly cover the three core stages of tobacco curing, namely the yellowing stage (low temperature and high humidity environment, 35-42℃ / 80-90%RH), the color fixing stage (medium temperature and low humidity environment, 42-54℃ / 60-70%RH), and the dry rib stage (high temperature and low humidity environment, 60-68℃ / 30-40%RH), to ensure the global representativeness of the statistical results.

[0063] A nonlinear mapping curve between the coupling index and the system's settling-in time was constructed. The first and second derivatives of the mapping curve were calculated, and two physical abrupt inflection points where the rate of change of the curve showed a significant jump were extracted. Through massive data fitting analysis, it was found that when the coupling index was below the first inflection point of 0.4, the cross-interference between temperature and humidity was slight, and the system's settling-in time showed a linear and gradual change. When it exceeded the first inflection point of 0.4 and was below the second inflection point of 0.7, the settling-in time showed a significant nonlinear increase. After exceeding the second inflection point of 0.7, the settling-in time experienced an exponential jump, and the system faced an extremely high risk of local water loss or condensation.

[0064] Therefore, based on the aforementioned physical inflection points, the basic grading intervals are strictly defined as follows: a coupling index greater than or equal to 0 and less than 0.4 corresponds to weak coupling; a coupling index greater than or equal to 0.4 and less than 0.7 corresponds to medium coupling; and a coupling index greater than or equal to 0.7 and less than or equal to 1.0 corresponds to strong coupling. For high-precision baking scenarios (such as the color-fixing period of upper tobacco leaves, which is extremely sensitive to temperature and humidity fluctuations), the interval can be tightened (e.g., lowering the lower limit of strong coupling to 0.6 to trigger strong interference compensation earlier), while for conventional baking scenarios, the interval can be widened.

[0065] The coupling index calculated in real time is compared with a preset grading threshold to obtain the coupling strength level. The level classification can intuitively distinguish the degree of influence of temperature and humidity interaction. For example, if the coupling index calculated at a certain moment is 0.78, since it is within the strong coupling range of greater than or equal to 0.7 and less than or equal to 1.0, the coupling strength level of the current baking chamber area is determined to be strong coupling.

[0066] In step S104, if the coupling strength level exceeds a preset coupling judgment threshold, then regional deviation data within the distributed dataset is extracted, the regional deviation data is weighted to obtain a comprehensive deviation value, and the comprehensive deviation value is compared with a preset imbalance judgment threshold to obtain the degree of temperature and humidity imbalance, including:

[0067] If the coupling strength level exceeds the preset coupling judgment threshold, the actual temperature and humidity values ​​in the distributed dataset are extracted, and the difference between the actual temperature and humidity values ​​and the preset target values ​​is calculated to obtain the regional deviation data.

[0068] The regional deviation data are weighted and calculated to generate a comprehensive deviation value;

[0069] The overall deviation value is compared with a preset imbalance judgment threshold to determine the degree of temperature and humidity imbalance.

[0070] It should be noted that the coupling judgment threshold is set based on the historical data statistics of temperature and humidity coupling control in tobacco curing rooms over the past year. The basic threshold is set to the strong coupling level. In high-precision tobacco curing scenarios, the judgment level can be lowered to medium coupling, while in conventional curing scenarios, the strong coupling judgment standard can be maintained.

[0071] The preset target values ​​are set according to the "three-stage" basic process of tobacco curing, and dynamically evolved in combination with the characteristics of tobacco leaves and real-time disturbances. First, the reference range of the basic target values ​​for each stage of tobacco curing is defined: during the yellowing stage, the target temperature is set at 35℃ to 42℃, and the target relative humidity is maintained at 80% to 90% to promote full degradation of chlorophyll; during the color-fixing stage, the target temperature is steadily increased to 42℃ to 54℃ at a preset slope, and the target relative humidity is smoothly reduced to 60% to 70% to lock in the quality and color of the tobacco leaves; during the drying stage, the target temperature is set at 60℃ to 68℃, and the target relative humidity is reduced to 30% to 40% to achieve complete dehydration.

[0072] Secondly, the system implements differentiated adjustment rules based on tobacco variety and growing part: for lower leaves with high water content and easy water loss, the system automatically lowers the target humidity during the yellowing period by 2% to 3% to accelerate early dehumidification; for upper leaves with dense tissue and resistance to baking, the system automatically raises the target humidity during the color fixing period by 1% to 2% and extends the temperature stabilization time to prevent ash accumulation or greening.

[0073] Finally, a dynamic update logic for preset target values ​​is introduced during the baking process: the system continuously monitors the aforementioned external weather fluctuation data and the rate of change of the actual temperature and humidity inside the room. When a cold wave or strong wind disturbance causes a jump in the system's heat load, and the actual temperature and humidity deviate from the predetermined process curve within three consecutive collection cycles, the system automatically calls the internal trend prediction model and superimposes an adaptive compensation increment (e.g., dynamically increasing the target temperature by 0.5℃ to offset the penetrating effect of external cooling) on ​​the original basic target value curve, generating a preset target value that is updated in real time, ensuring that the control benchmark accurately matches the current physical needs.

[0074] If the coupling strength level exceeds the preset coupling judgment threshold, the actual temperature and humidity values ​​of each monitoring area in the distributed dataset are extracted. The absolute difference between the actual value and the dynamically updated preset target value of the baking process is calculated point by point. All differences are mapped to the interval between 0 and 1 using the minimum-maximum normalization method to form standardized regional deviation data. For example, the actual temperature and humidity values ​​of the middle layer area of ​​the baking chamber are extracted, the difference between them and the dynamically updated process target value at the current moment is calculated, and after normalization, the deviation data of that area is generated.

[0075] Next, the regional deviation data are weighted to generate a comprehensive deviation value. The weighted combination is obtained by summing the temperature deviation and humidity deviation, with a weight of 0.6 for temperature deviation and 0.4 for humidity deviation. This weighting combination is based on historical control data of the tobacco curing process. Temperature changes have a more direct impact on tobacco quality and are given a higher weight, while the impact of humidity is relatively milder and is given a lower weight. For example, combining a temperature deviation of 0.4 and a humidity deviation of 0.3 to complete the weighted calculation yields the comprehensive deviation value for the corresponding region.

[0076] It is worth noting that the imbalance judgment thresholds are set based on statistical data of effective temperature and humidity imbalance control in the baking chamber over the past year. The basic grading intervals are set as follows: 0 to 0.3 corresponds to slight imbalance, 0.3 to 0.5 corresponds to moderate imbalance, and 0.5 to 1.0 corresponds to severe imbalance. The intervals can be tightened during the critical color-fixing baking stage and widened during the normal yellowing stage. By comparing the comprehensive deviation value with the preset imbalance judgment thresholds, the degree of temperature and humidity imbalance in the corresponding area is directly determined. The grading results can be directly used for subsequent control decisions. For example, if the comprehensive deviation value falls into the moderate imbalance range, the temperature and humidity in that baking area are determined to be in a state of moderate imbalance.

[0077] In step S105, the process of fusing the external weather fluctuation data, the regional deviation data, and the degree of temperature and humidity imbalance to generate temperature and humidity imbalance distribution information includes:

[0078] The external weather fluctuation data is normalized and weighted to obtain the environmental impact value.

[0079] The environmental impact values, the regional deviation data, and the degree of temperature and humidity imbalance are linearly weighted and superimposed to generate temperature and humidity imbalance distribution information.

[0080] It should be noted that, firstly, the environmental impact values ​​are calculated based on external weather fluctuation data. The numerical calculation employs a weighted fusion algorithm, with the calculation weights determined statistically through historical baking environment correlation data. The weight for external temperature fluctuations is set at 0.6, and the weight for external humidity fluctuations is set at 0.4. This combination appropriately matches the impact patterns of weather factors on the indoor environment; temperature fluctuations, with their stronger transmission effect, are given a higher weight, while humidity fluctuations, with their relatively milder impact, are given a lower weight. All weather data are mapped to the 0-1 range using the min-max normalization method, and the calculation results are simultaneously standardized. For example, by combining the external temperature and humidity fluctuation values ​​for weighted calculation, a quantifiable environmental impact value representing the intensity of environmental disturbance is obtained.

[0081] Finally, environmental impact data, regional deviation data, and the degree of temperature and humidity imbalance are integrated to generate temperature and humidity imbalance distribution information. Multi-source data fusion employs a linear weighted superposition method. The fusion weights are set based on the actual needs of baking control: regional deviation data has a weight of 0.5, the degree of temperature and humidity imbalance has a weight of 0.3, and environmental impact data has a weight of 0.2. This combination highlights the dominant role of the oven's own deviations while taking into account the corrective effects of the external environment. The fusion results are integrated according to the spatial regions of the baking oven, forming temperature and humidity imbalance distribution information covering the entire area and with time-series identification, which can be directly used for subsequent control command generation.

[0082] In step S106, generating a compensation instruction for zoned humidity based on the temperature and humidity imbalance distribution information, and adjusting the ventilation equipment parameters of the baking chamber according to the compensation instruction to obtain a temperature and humidity balance state, includes:

[0083] The humidity control zones of the baking room are divided according to the temperature and humidity imbalance distribution information;

[0084] Calculate the humidity compensation value for each humidity control zone and convert the humidity compensation value into a zone humidity compensation command.

[0085] The humidity compensation command is issued according to the preset execution priority to adjust the operating parameters of the ventilation equipment;

[0086] The temperature and humidity fluctuations in the baking chamber are monitored. If the fluctuation values ​​are lower than a preset stable threshold, a temperature and humidity balance state is obtained.

[0087] It should be noted that, firstly, the humidity control zones of the curing barn are divided based on the distribution information of temperature and humidity imbalance. The zone division is based on the degree of imbalance and spatial layout, dividing the curing barn into three categories: key control zone, regular control zone, and steady-state maintenance zone. This division rule closely matches the regional sensitivity characteristics of different stages of tobacco curing, with no spatial overlap between the control zones, allowing for precise adaptation to differentiated humidity regulation needs. For example, areas with moderate and severe imbalance are designated as key control zones, areas with slight imbalance as regular control zones, and areas without imbalance as steady-state maintenance zones.

[0088] The specific process for setting the preset compensation weights is as follows: 10,000 production verification samples containing dynamic humidity adjustment records for different zones and the final comprehensive tobacco quality score (covering the proportion of orange-yellow tobacco, oil content grade, and dry matter loss rate) are extracted from the historical tobacco curing database of the past year. A nonlinear correlation model between the zone humidity evolution characteristics and the final tobacco quality is constructed using a random forest algorithm. Specifically, the input features of the random forest algorithm are defined as the initial humidity deviation amplitude, average adjustment rate, and time-domain cumulative error of each control zone (key control zone, regular control zone, and steady-state maintenance zone), and the output label is the corresponding final comprehensive tobacco quality score. The core hyperparameters of the model are set as follows: the number of decision trees is set to 200, the maximum tree depth is set to 15 layers, and the minimum number of samples required for internal node subdivision is set to 5 to prevent overfitting. In the training process, the 10,000 sets of validation samples were randomly divided into training and test sets at an 8:2 ratio. Multiple training subsets were generated from the training set using the Bootstrap sampling with replacement method to construct independent decision trees. At each node split, a feature subset was randomly selected to find the optimal split point. The convergence condition for training was set as follows: the fluctuation rate of the out-of-bag mean square error in 10 consecutive incremental iterations of the decision tree was less than 0.001, or the upper limit of 200 decision trees was reached. After the model training converged, the contribution of humidity deviation in each zone to the overall baking quality was extracted by calculating the reduction in Gini impurity. Verification through massive data fitting and global normalization revealed that the feature contribution ratio of key control areas (such as the densely packed smoke area with the most intense temperature and humidity gradients) remained stable at around 60%, the ratio of regular control areas (such as the airflow circulation area in the middle of the duct) was approximately 30%, and the ratio of steady-state maintenance areas (such as the area near the exhaust vent) was approximately 10%. Therefore, strictly following the above-mentioned characteristic contribution ratio rules, the weight of the key control area is set to 0.6, the weight of the regular control area is set to 0.3, and the weight of the steady-state maintenance area is set to 0.1.

[0089] After numerical conversion, standardized humidity compensation instructions for the corresponding region are generated. For example, the humidity deficit value of the key control area is calculated, and after normalization conversion, a precise humidity compensation instruction for that region is generated by combining it with an allocation weight of 0.6.

[0090] Subsequently, humidity compensation commands are issued according to the preset execution priority to adjust the operating parameters of the ventilation equipment. The execution priority is set based on the degree of influence of each region on the quality of tobacco curing, with the priority order being: key control areas take precedence over regular control areas, and regular control areas take precedence over steady-state maintenance areas. After the command is issued, two types of operating parameters, namely the speed of the ventilation fan and the opening of the air valve, are adjusted simultaneously to meet the needs of airflow distribution optimization.

[0091] For example, compensation instructions are issued to key control areas first, and the operating speed of ventilation fans and the opening of dampers in the corresponding areas are increased simultaneously.

[0092] The fluctuation values ​​of temperature and humidity throughout the curing chamber are continuously monitored, and the fluctuation values ​​are normalized to unify the evaluation criteria. In this implementation case, the specific process for setting the preset stability threshold is as follows: 10,000 sets of temperature and humidity sample sequences labeled by process experts in steady-state operation and dynamic adjustment are extracted from the historical tobacco curing database of the past year; the fluctuation variance of the normalized temperature and humidity data in each sample sequence is calculated using a 3-minute sliding window; and the probability density function (PDF) distribution curves of the fluctuation variance of steady-state samples and dynamic samples are constructed respectively. Since an excessively high threshold will lead to premature misjudgment of steady state (resulting in uneven tobacco drying), and an excessively low threshold will cause the system to fall into continuous oscillation (failing to determine convergence and increasing ineffective energy consumption), a comprehensive cost function including the steady-state misjudgment rate and the system convergence time is constructed, and the derivative of this cost function is used to find the minimum point. After verification through massive data fitting, when the split point corresponding to the minimum value is 0.1, the system achieves a globally optimal balance between preventing premature misjudgment and ensuring rapid convergence. Therefore, the basic threshold is set to 0.1.

[0093] Meanwhile, considering the biochemical reaction characteristics of tobacco leaves at different curing stages: for the high-precision curing scenario during the color-fixing period (when tobacco leaves are extremely sensitive to sudden changes in temperature and humidity, and even slight fluctuations can easily lead to irreversible quality damage such as green curing or ash formation), the threshold can be lowered to 0.05; for the conventional curing scenario during the yellowing period (when the main goal is to maintain a high-humidity environment to promote internal enzyme activity, allowing for a certain range of humidity fluctuations), the threshold can be raised to 0.15. Those skilled in the art can adjust this threshold within the range of 0.05 to 0.15 (inclusive) according to the actual curing stage.

[0094] If the calculated fluctuation value is less than or equal to the preset stability threshold, the baking chamber is determined to have reached a temperature and humidity balance state. For example, if the monitored temperature and humidity fluctuation value is less than 0.1 and the duration meets the judgment period, the baking chamber is determined to have reached a stable temperature and humidity balance state.

[0095] In step S107, the step of extracting real-time feedback data based on the temperature and humidity balance state, and updating the ventilation equipment parameters according to the real-time feedback data to obtain a stable temperature and humidity distribution includes:

[0096] Extract the real-time humidity data under the temperature and humidity equilibrium state, and determine the real-time humidity data as real-time feedback data;

[0097] Obtain preset device response delay data, and combine the real-time feedback data with the device response delay data to perform trend prediction and obtain the humidity change trend;

[0098] The parameters of the ventilation equipment are updated according to the humidity change trend.

[0099] After continuous monitoring of the temperature and humidity deviation values ​​of the entire baking room after parameter updates, if the temperature and humidity deviation values ​​are less than the preset deviation judgment threshold, a stable temperature and humidity distribution is obtained.

[0100] It should be noted that, firstly, real-time humidity data under temperature and humidity equilibrium is extracted and used as real-time feedback data. To achieve precise zoned control, the system pre-constructs a ventilation system architecture based on dense smoke loading and stratified circulation. Specifically, the baking chamber is divided vertically into three independent humidity control zones: an upper color-sensing zone, a middle steady-state transition zone, and a lower heating-driven zone. In terms of airflow organization and hardware mapping, each control zone is equipped with an independent set of adjustable-speed centrifugal fans and electric airflow regulating valves, forming a one-to-one physical controlled unit (for example, the upper color-sensing zone has a dedicated No. 1 fan and No. 1 valve for independent local airflow circulation control). The system calculates specific equipment adjustment parameters based on the humidity compensation values ​​of each zone using a preset quantitative conversion model. This quantitative conversion model clearly stipulates that: the fan frequency adjustment increment equals the humidity compensation amount multiplied by the fan dehumidification sensitivity coefficient, plus a fan offset correction term; simultaneously, the valve opening step value equals the humidity compensation amount multiplied by the valve dehumidification sensitivity coefficient, plus a valve offset correction term.

[0101] The sensitivity coefficient is calibrated based on the equipment's rated dehumidification capacity, while the bias correction term is dynamically calculated based on the real-time temperature and humidity baseline deviation. Pre-set equipment response delay data is acquired. This delay data comes from historical operational test statistics of ventilation and humidification equipment, covering two types of parameters: fan speed adjustment delay and valve opening response delay. Parameter values ​​can be adaptively adjusted according to the equipment's aging level to meet the control requirements of equipment at different service lifespans. Trend prediction is performed by combining real-time feedback data and equipment response delay data to obtain the humidity change trend. The trend prediction uses a discrete Kalman filter model with time-delay compensation logic.

[0102] The complete structure and parameter settings of the model are based on the following: First, the state variables and observation variables are defined. A two-dimensional vector containing the predicted humidity value and the rate of change of humidity at the current moment is set as the state variable, and the real-time humidity measurement value collected by the sensor is set as the observation variable. Second, state equations and observation equations are constructed. The state equation is obtained by multiplying the state transition matrix by the state variable at the previous moment and adding process noise. The observation equation is obtained by multiplying the observation matrix by the current state variable and adding measurement noise. Specifically, when constructing the state transition matrix, the acquired equipment response delay data is introduced as a time constant into the Taylor expansion term, enabling the transition matrix to predict the equipment's lag state. Next, the numerical range of the core covariance matrix is ​​set. Based on the turbulence intensity caused by the fan speed fluctuation, the range of the process noise covariance matrix is ​​set within a certain range. to Between; based on the sensor's factory signal-to-noise ratio characteristics, the range of values ​​for the measurement noise covariance matrix is ​​set within... to Finally, the system executes the equipment delay compensation logic. It updates the optimal state estimate for the current moment by calculating the Kalman gain and performs advance prediction using a corrected state transition matrix incorporating the equipment response delay time. The prediction time step is strictly set to the corresponding equipment response delay duration, thereby accurately predicting the expected humidity pattern at the moment the equipment actually takes effect. For example, by substituting the feedback humidity data and equipment delay data into the filtering algorithm, the future humidity decrease trend can be predicted.

[0103] The parameters of the ventilation equipment are updated based on humidity change trends. Parameter updates are implemented in conjunction with regional control weights, with larger adjustments made to parameters in key control areas compared to regular control areas. Updates include two main indicators: fan speed and vent opening. The adjustment magnitude closely matches the strength of the humidity change trend to avoid secondary fluctuations caused by over-control. For example, based on a decreasing humidity trend, the fan speed and damper opening in key control areas are slightly increased to complete the equipment parameter update.

[0104] In this implementation case, the deviation judgment threshold is set based on statistically measured data from the stable operation of the entire tobacco curing area over the past year. The basic threshold is set to 0.05, which can be lowered to 0.03 during the high-precision tobacco color fixing stage and raised to 0.08 during the regular curing stage. This threshold has been verified under multiple operating conditions to stably distinguish between stable and slightly fluctuating states. After continuous monitoring of parameter updates, the temperature and humidity deviation values ​​across the entire curing chamber are recorded. All deviation values ​​are normalized. If the temperature and humidity deviation values ​​are less than the preset deviation judgment threshold, a uniform and stable temperature and humidity distribution is obtained across the entire curing chamber. For example, if the monitored temperature and humidity deviation value is below 0.05, it is determined that the curing chamber has formed a uniform and stable temperature and humidity distribution.

[0105] In summary, this invention discloses an intelligent temperature and humidity control method for tobacco curing barns. The method includes collecting temperature and humidity data from multiple areas of the curing barn along with external weather data; obtaining a standardized distribution dataset through anomaly removal and smoothing filtering; constructing a temperature and humidity interaction matrix to calculate and classify the coupling index; quantifying regional deviations and determining the degree of imbalance when thresholds are exceeded; fusing environmental disturbance data to generate global imbalance distribution information; generating zonal compensation commands to dynamically adjust ventilation equipment parameters; and combining real-time feedback and equipment delay prediction of humidity trends to achieve closed-loop parameter optimization and stable temperature and humidity balance. This invention solves the problems of traditional single-point control, poor adaptability of fixed parameters, and temperature and humidity coupling imbalance, achieving precise zonal control of the entire curing barn and effectively ensuring the consistency of tobacco curing quality and operational stability.

[0106] Reference Figure 2 The second embodiment of the present invention provides an intelligent temperature and humidity control system for a tobacco curing room, comprising:

[0107] The data acquisition module is used to acquire raw temperature and humidity data from multiple areas within the baking chamber, as well as external weather fluctuation data.

[0108] The data filtering module is used to filter the raw temperature and humidity data to obtain a temperature and humidity distribution dataset.

[0109] The coupling analysis module is used to construct a temperature and humidity interaction matrix based on the distributed dataset, calculate the coupling index of the temperature and humidity interaction matrix, and divide the coupling index by a preset grading threshold to obtain the coupling strength level.

[0110] The imbalance determination module is used to extract regional deviation data within the distributed dataset if the coupling strength level exceeds a preset coupling determination threshold, perform weighted calculation on the regional deviation data to obtain a comprehensive deviation value, and compare the comprehensive deviation value with the preset imbalance determination threshold to obtain the degree of temperature and humidity imbalance.

[0111] The data fusion module is used to fuse the external weather fluctuation data, the regional deviation data, and the degree of temperature and humidity imbalance to generate temperature and humidity imbalance distribution information;

[0112] The parameter control module is used to generate a compensation instruction for the humidity of the zone based on the temperature and humidity imbalance distribution information, and adjust the ventilation equipment parameters of the baking room according to the compensation instruction to obtain a temperature and humidity balance state.

[0113] The feedback optimization module is used to extract real-time feedback data based on the temperature and humidity balance state, update the ventilation equipment parameters according to the real-time feedback data, and obtain a stable temperature and humidity distribution.

[0114] It should be noted that the intelligent temperature and humidity control system for tobacco curing room provided in this embodiment of the invention is used to execute all the process steps of the intelligent temperature and humidity control method for tobacco curing room in the above embodiment. The working principle and beneficial effects of the two are one-to-one, so they will not be described again.

[0115] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0116] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for intelligent temperature and humidity control in a tobacco curing chamber, characterized in that, include: Acquire raw temperature and humidity data for multiple areas within the baking chamber, as well as external weather fluctuation data; The raw temperature and humidity data are filtered to obtain a temperature and humidity distribution dataset. A temperature and humidity interaction matrix is ​​constructed based on the distributed dataset. The coupling index of the temperature and humidity interaction matrix is ​​calculated. The coupling index is divided by a preset grading threshold to obtain the coupling strength level. If the coupling strength level exceeds the preset coupling judgment threshold, then regional deviation data within the distributed dataset is extracted, the regional deviation data is weighted and calculated to obtain a comprehensive deviation value, and the comprehensive deviation value is compared with the preset imbalance judgment threshold to obtain the degree of temperature and humidity imbalance. By integrating the external weather fluctuation data, the regional deviation data, and the degree of temperature and humidity imbalance, temperature and humidity imbalance distribution information is generated. Based on the temperature and humidity imbalance distribution information, a compensation instruction for the humidity of the zone is generated. The ventilation equipment parameters of the baking room are adjusted according to the compensation instruction to obtain a temperature and humidity balance state. Real-time feedback data is extracted based on the temperature and humidity balance state, and the parameters of the ventilation equipment are updated according to the real-time feedback data to obtain a stable temperature and humidity distribution.

2. The intelligent temperature and humidity control method for tobacco curing chamber according to claim 1, characterized in that, The acquisition of raw temperature and humidity data for multiple areas within the baking chamber and external weather fluctuation data includes: Collect temperature and humidity data at various monitoring points inside the baking room; Align the timeline nodes of the temperature acquisition data and the humidity acquisition data to obtain time-synchronized data; By integrating the aforementioned time-series synchronized data, raw temperature and humidity data for multiple areas within the baking chamber are generated. Collect ambient temperature and humidity data outside the baking room, and integrate the ambient temperature and humidity data to obtain external weather fluctuation data.

3. The intelligent temperature and humidity control method for tobacco curing chamber according to claim 1, characterized in that, The filtering process performed on the raw temperature and humidity data to obtain a temperature and humidity distribution dataset includes: The abnormal values ​​in the original temperature and humidity data are removed to obtain the cleaned data; The cleaned data is smoothed and integrated to obtain a temperature and humidity distribution dataset.

4. The intelligent temperature and humidity control method for a tobacco curing room according to claim 1, characterized in that, The step involves constructing a temperature and humidity interaction matrix based on the distributed dataset, calculating the coupling index of the temperature and humidity interaction matrix, and classifying the coupling index using a preset grading threshold to obtain a coupling strength level, including: Temperature and humidity feature data are extracted from the distributed dataset. A temperature and humidity interaction matrix is ​​constructed based on the temperature feature data and the humidity feature data; The coupling index is obtained by analyzing the characteristic values ​​of the temperature and humidity interaction matrix and normalizing the characteristic values. The coupling index is compared with a preset grading threshold to obtain the coupling strength level.

5. The intelligent temperature and humidity control method for tobacco curing chamber according to claim 1, characterized in that, If the coupling strength level exceeds a preset coupling judgment threshold, then regional deviation data within the distributed dataset is extracted, the regional deviation data is weighted to obtain a comprehensive deviation value, and the comprehensive deviation value is compared with a preset imbalance judgment threshold to obtain the degree of temperature and humidity imbalance, including: If the coupling strength level exceeds the preset coupling judgment threshold, the actual temperature and humidity values ​​in the distributed dataset are extracted, and the difference between the actual temperature and humidity values ​​and the preset target values ​​is calculated to obtain the regional deviation data. The regional deviation data are weighted and calculated to generate a comprehensive deviation value; The overall deviation value is compared with a preset imbalance judgment threshold to determine the degree of temperature and humidity imbalance.

6. The intelligent temperature and humidity control method for a tobacco curing chamber according to claim 1, characterized in that, The process of integrating the external weather fluctuation data, the regional deviation data, and the degree of temperature and humidity imbalance to generate temperature and humidity imbalance distribution information includes: The external weather fluctuation data is normalized and weighted to obtain the environmental impact value. The environmental impact values, the regional deviation data, and the degree of temperature and humidity imbalance are linearly weighted and superimposed to generate temperature and humidity imbalance distribution information.

7. The intelligent temperature and humidity control method for tobacco curing chamber according to claim 1, characterized in that, The process of generating a zoned humidity compensation instruction based on the temperature and humidity imbalance distribution information, and adjusting the ventilation equipment parameters of the baking chamber according to the compensation instruction to obtain a temperature and humidity balance state, includes: The humidity control zones of the baking room are divided according to the temperature and humidity imbalance distribution information; Calculate the humidity compensation value for each humidity control zone and convert the humidity compensation value into a zone humidity compensation command. The humidity compensation command is issued according to the preset execution priority to adjust the operating parameters of the ventilation equipment; The temperature and humidity fluctuations in the baking chamber are monitored. If the fluctuation values ​​are lower than a preset stable threshold, a temperature and humidity balance state is obtained.

8. The intelligent temperature and humidity control method for a tobacco curing room according to claim 1, characterized in that, The step of extracting real-time feedback data based on the temperature and humidity balance state, and updating the ventilation equipment parameters according to the real-time feedback data to obtain a stable temperature and humidity distribution includes: Extract the real-time humidity data under the temperature and humidity equilibrium state, and determine the real-time humidity data as real-time feedback data; Obtain preset device response delay data, and combine the real-time feedback data with the device response delay data to perform trend prediction and obtain the humidity change trend; The parameters of the ventilation equipment are updated according to the humidity change trend. After continuous monitoring of the temperature and humidity deviation values ​​of the entire baking room after parameter updates, if the temperature and humidity deviation values ​​are less than the preset deviation judgment threshold, a stable temperature and humidity distribution is obtained.

9. An intelligent temperature and humidity control system for a tobacco curing room, characterized in that, include: The data acquisition module is used to acquire raw temperature and humidity data and external weather fluctuation data in multiple areas of the baking room; The data filtering module is used to filter the raw temperature and humidity data to obtain a temperature and humidity distribution dataset. The coupling analysis module is used to construct a temperature and humidity interaction matrix based on the distributed dataset, calculate the coupling index of the temperature and humidity interaction matrix, and divide the coupling index by a preset grading threshold to obtain the coupling strength level. The imbalance determination module is used to extract regional deviation data within the distributed dataset if the coupling strength level exceeds a preset coupling determination threshold, perform weighted calculation on the regional deviation data to obtain a comprehensive deviation value, and compare the comprehensive deviation value with the preset imbalance determination threshold to obtain the degree of temperature and humidity imbalance. The data fusion module is used to fuse the external weather fluctuation data, the regional deviation data, and the degree of temperature and humidity imbalance to generate temperature and humidity imbalance distribution information; The parameter control module is used to generate a compensation instruction for the humidity of the zone based on the temperature and humidity imbalance distribution information, and adjust the ventilation equipment parameters of the baking room according to the compensation instruction to obtain a temperature and humidity balance state. The feedback optimization module is used to extract real-time feedback data based on the temperature and humidity balance state, update the ventilation equipment parameters according to the real-time feedback data, and obtain a stable temperature and humidity distribution.

Images