Cut tobacco dryer temperature control method and system, electronic equipment and medium

Abstract

The invention provides a cut tobacco dryer temperature control method and system, electronic equipment and a medium, and the method comprises the steps: collecting the current inlet cut tobacco moisture, the current cut tobacco outlet temperature and the current cylinder wall temperature of a cut tobacco dryer, and obtaining a cut tobacco outlet temperature prediction value at the next moment based on a prediction model; obtaining a preset target cut tobacco quality evaluation value and the current cut tobacco flow, and obtaining a reference value of the cut tobacco outlet temperature based on the trained deep learning model; inputting the predicted value of the tobacco shred outlet temperature and the reference value of the tobacco shred outlet temperature into an MPC algorithm model, and outputting a cylinder wall temperature adjustment amount; and adjusting the temperature of the cut tobacco dryer based on the cylinder wall temperature adjustment amount. Compared with the prior art, the final quality of the tobacco shreds is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of tobacco processing, and in particular to a temperature control method, system, electronic equipment and medium for a tobacco drying machine. Background Art

[0002] The traditional tobacco drying machine temperature control method mainly relies on the detection results of the outlet tobacco moisture content. That is, the outlet tobacco moisture content is compared with the preset value. If the outlet moisture content is smaller than the preset value, the tobacco drying machine temperature is reduced, otherwise the tobacco drying machine temperature is increased.

[0003] However, when the moisture content, flow rate, or physical properties of the incoming tobacco change, the system may be unable to adjust in time, resulting in unstable tobacco quality. Furthermore, because the drying process involves chemical reactions that are sensitive to temperature and time, traditional control methods may not be able to accurately predict and regulate these reactions, thus affecting the final quality of the tobacco. Summary of the Invention

[0004] In view of this, one of the objectives of the embodiments of the present application is to provide a tobacco drying machine temperature control method, which can improve the problem that the existing tobacco drying temperature control method affects the final quality of tobacco.

[0005] To achieve the above technical objectives, the technical solutions adopted in this application are as follows:

[0006] In a first aspect, an embodiment of the present application provides a method for controlling the temperature of a tofu drying machine, the method comprising:

[0007] Collect the current inlet moisture, outlet temperature and wall temperature of the tobacco at the tobacco drying machine;

[0008] Inputting the current inlet tobacco moisture, current tobacco outlet temperature and current barrel wall temperature into the prediction model to obtain a predicted value of the tobacco outlet temperature at the next moment;

[0009] Obtain the preset target tobacco quality evaluation value and current tobacco flow rate, and obtain the baseline value of tobacco outlet temperature based on the trained deep learning model;

[0010] Inputting the predicted value of the cut tobacco outlet temperature and the reference value of the cut tobacco outlet temperature into the MPC algorithm model, and outputting the cylinder wall temperature adjustment value;

[0011] The temperature of the tofu drying machine is adjusted based on the cylinder wall temperature adjustment amount.

[0012] Furthermore, obtaining a predicted value of the cut tobacco outlet temperature based on the prediction model includes:

[0013] The moisture content, flow rate and wall temperature of the cut tobacco are input into a prediction model, which outputs a predicted value of the cut tobacco outlet temperature. The prediction model is:

[0014] T out (k1+1)=a·T out (k1)+b·T wall (k1)+c·W in (k1)

[0015] Among them, T out (k1+1) is the predicted value of tobacco outlet temperature;

[0016] T out (k1) is the actual tobacco outlet temperature;

[0017] T wall (k1) is the current cylinder wall temperature;

[0018] W in (k1) is the current moisture content of the tobacco;

[0019] a, b, c are all model parameters.

[0020] Furthermore, the base value of the tobacco outlet temperature is obtained based on the trained deep learning model, including:

[0021] Initializing the deep learning model;

[0022] Based on the training data, the initialized deep learning model is trained to obtain a trained deep learning model, wherein the training data includes independent variable data and corresponding dependent variable data, the independent variable data includes training cut tobacco moisture, training cut tobacco flow rate, training barrel wall temperature, and training cut tobacco outlet temperature, and the dependent variable data includes a training cut tobacco quality evaluation value;

[0023] The current cut tobacco moisture, current cut tobacco flow, current barrel wall temperature and target cut tobacco quality evaluation value are input into the trained deep learning model to obtain a reference value of the cut tobacco outlet temperature.

[0024] Furthermore, the objective function of the MPC algorithm model is:

[0025]

[0026] Among them, N p represents the prediction time domain;

[0027] represents the predicted value of the tobacco outlet temperature;

[0028] T set Indicates the reference value of the outlet temperature of the cut tobacco;

[0029] N c represents the control time domain;

[0030] k2 represents the time;

[0031] i represents step;

[0032] Δu(k2+i) represents the cylinder wall temperature adjustment in step i at time k2;

[0033] represents the predicted value of tobacco outlet temperature at step i at time k2;

[0034] Q represents the weight of Δu(k2+i).

[0035] Furthermore, adjusting the temperature of the tofu drying machine based on the cylinder wall temperature adjustment amount includes:

[0036] Sending the cylinder wall temperature adjustment amount to a PID controller;

[0037] The PID controller is used to obtain a proportional coefficient, an integral coefficient, and a differential coefficient based on the cylinder wall temperature adjustment amount;

[0038] An opening value is obtained based on the proportional coefficient, the integral coefficient, and the differential coefficient, and the valve opening of the actuator is controlled to the opening value.

[0039] Furthermore, the PID controller obtains a proportional coefficient, an integral coefficient, and a differential coefficient based on the cylinder wall temperature adjustment amount, including:

[0040] The cylinder wall temperature adjustment amount is transferred to the PID parameter algorithm model to obtain the proportional coefficient, integral coefficient and differential coefficient. The PID parameter algorithm model is:

[0041] K p =K p0 +α·|Δu(k2+i)|

[0042] Among them: K p represents the proportionality coefficient;

[0043] K p0 Represents the initialization scale factor;

[0044] α represents the sensitivity coefficient of the proportional term;

[0045] Δu represents the cylinder wall temperature adjustment amount;

[0046]

[0047] Where: T i represents the integral coefficient;

[0048] T i0 Represents the initialized integral coefficient;

[0049] β represents the integral adaptation coefficient;

[0050] Δu(k2+i) represents the cylinder wall temperature adjustment in step i at time k2;

[0051]

[0052] Among them, T d represents the differential coefficient;

[0053] T d0 Represents the initialized differential coefficient;

[0054] γ represents the differential sensitivity coefficient.

[0055] In a second aspect, the present application provides a temperature control system for a tofu drying machine, comprising:

[0056] The first calculation module is configured to collect the moisture content, tobacco flow rate and cylinder wall temperature of the tobacco at the inlet of the tobacco drying machine, and obtain a predicted value of the tobacco outlet temperature based on the prediction model;

[0057] The second calculation module is configured to obtain a preset target cut tobacco quality evaluation value and obtain a reference value of the cut tobacco outlet temperature based on the trained deep learning model;

[0058] a third calculation module configured to input the predicted value of the cut tobacco outlet temperature and the reference value of the cut tobacco outlet temperature into an MPC algorithm model and output a barrel wall temperature adjustment value;

[0059] The control module is configured to adjust the temperature of the tofu drying machine based on the drum wall temperature adjustment amount.

[0060] In a third aspect, an embodiment of the present application proposes an electronic device, which includes a processor and a memory coupled to each other, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the electronic device executes the above method.

[0061] In a fourth aspect, an embodiment of the present application proposes a computer-readable storage medium, in which a computer program is stored. When the computer program runs on a computer, the computer executes the above method.

[0062] The invention adopting the above technical solution has the following advantages:

[0063] In the technical solution provided in the present application, the predicted value of the tobacco outlet temperature is predicted by the current inlet tobacco moisture, the current tobacco outlet temperature and the current barrel wall temperature, and then the target tobacco quality evaluation value is used as the target to obtain the baseline value of the tobacco outlet temperature. Then, the predicted value of the tobacco outlet temperature and the baseline value of the tobacco outlet temperature are applied, and the MPC algorithm is applied to obtain the barrel wall temperature adjustment amount, and the temperature of the tobacco drying machine is adjusted based on the barrel wall temperature adjustment amount. This solution predicts the predicted value of the tobacco outlet temperature at the next moment by considering multi-source data including the current inlet tobacco moisture, the current tobacco outlet temperature and the current barrel wall temperature, thereby improving the prediction accuracy. At the same time, the target tobacco quality evaluation value is used as a guide to make the baseline value of the tobacco outlet temperature more accurate. At the same time, the MPC algorithm can consider future states and optimize the control strategy to achieve precise regulation of the tobacco drying process, thereby improving the final quality of the tobacco compared to the existing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The present application may be further illustrated by the non-limiting embodiments provided in the accompanying drawings. It should be understood that the following drawings illustrate only certain embodiments of the present application and are therefore not to be construed as limiting the scope of the present application. It is understood that a person skilled in the art can derive other relevant drawings from these drawings without inventive effort.

[0065] Figure 1 This is a flowchart of S110-S140 provided in an embodiment of the present application.

[0066] Figure 2 This is a sub-flowchart of S140 provided in an embodiment of the present application.

[0067] Figure 3 This is a block diagram of the temperature control system of the tofu drying machine provided in an embodiment of the present application.

[0068] Icon: 100 - first calculation module; 200 - second calculation module; 300 - third calculation module; 400 - control module. DETAILED DESCRIPTION

[0069] The present application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that similar or identical parts in the drawings or descriptions are numbered the same. Implementations not shown or described in the drawings are known to those of ordinary skill in the art. In the description of this application, the terms "first," "second," etc. are used solely to distinguish descriptions and are not to be construed as indicating or implying relative importance.

[0070] An embodiment of the present application provides an electronic device including a processing module and a storage module. The storage module stores a computer program, and when the computer program is executed by the processing module, the electronic device can perform corresponding steps in the following method for controlling the temperature of a tofu drying machine.

[0071] Please refer to Figure 1 The present application also provides a method for controlling the temperature of a tofu drying machine. The method for controlling the temperature of a tofu drying machine may include the following steps:

[0072] S110, collecting the current inlet moisture, current outlet temperature, and current wall temperature of the tobacco at the tobacco drying machine, and obtaining a predicted value of the tobacco outlet temperature at the next moment based on a prediction model;

[0073] S120, obtaining a preset target tobacco quality evaluation value and a current tobacco flow rate, and obtaining a reference value of the tobacco outlet temperature based on the trained deep learning model;

[0074] S130, inputting the predicted value of the cut tobacco outlet temperature and the reference value of the cut tobacco outlet temperature into the MPC algorithm model, and outputting the cylinder wall temperature adjustment value;

[0075] S140: Adjust the temperature of the tofu drying machine based on the cylinder wall temperature adjustment amount.

[0076] The following will describe in detail the various steps of the temperature control method of the tofu drying machine, as follows:

[0077] In S110, the moisture content of incoming cut tobacco can be measured using a moisture meter. Specifically, a moisture meter can be installed on the electronic scale belt before drying to measure and collect the moisture content of the cut tobacco in real time. This method directly and accurately obtains the moisture content of the cut tobacco, providing key data for subsequent prediction models.

[0078] The tobacco flow rate can be collected using an electronic belt scale.

[0079] The cylinder wall temperature can be collected using a temperature sensor. The temperature sensor can be installed on the cylinder wall condensate discharge pipe to measure and collect cylinder wall temperature data in real time.

[0080] In S110, the prediction model is implemented based on a mathematical algorithm and is used to output a predicted value of the tobacco outlet temperature based on the inlet tobacco moisture, tobacco flow rate and barrel wall temperature as input values. In this embodiment, the prediction model can be expressed as:

[0081] The prediction model is:

[0082] T out (k1+1)=a·T out (k1)+b·T wall(k1)+c·W in (k1)

[0083] Among them, T out (k1+1) is the predicted value of tobacco outlet temperature at the next moment;

[0084] T out (k1) is the current tobacco outlet temperature;

[0085] T wall (k1) is the current cylinder wall temperature;

[0086] W in (k1) is the current moisture content of the tobacco;

[0087] a, b, c are all model parameters.

[0088] Among them, a, b, c can represent T respectively out (k1), T wall (k1) and W in The weight value of (k1) and the specific values of a, b, and c can be obtained based on actual bench experiments.

[0089] In S120, the deep learning model can be a neural network model, an LSTM model, etc. When using the deep learning model, the collected preset target tobacco quality evaluation value, the current tobacco flow rate, the current inlet tobacco moisture and the current barrel wall temperature are used as input to obtain the benchmark value of the tobacco outlet temperature.

[0090] In this embodiment, the method for training the deep learning model is:

[0091] Initializing the deep learning model;

[0092] Based on the training data, the initialized deep learning model is trained to obtain a trained deep learning model, wherein the training data includes independent variable data and corresponding dependent variable data, the independent variable data includes training tobacco moisture, training tobacco flow, training barrel wall temperature and training tobacco outlet temperature, and the dependent variable data includes training tobacco quality evaluation value.

[0093] Among them, training data can come from the following aspects:

[0094] 1. Public Datasets: Many research institutions and companies publicly release the datasets they collect or organize for use by other researchers and developers. These datasets typically contain a large number of samples and cover a wide range of fields. In the tobacco industry, there may also be relevant public datasets that include various parameters and quality evaluation values during the tobacco production process.

[0095] 2. Internal company data: Tobacco companies typically collect extensive data during the production process, including tobacco moisture, flow rate, barrel wall temperature, and outlet temperature. This data can be directly used to train deep learning models. Companies may also have evaluation data on tobacco quality, such as expert scoring or quality testing results, which can serve as dependent variable data.

[0096] 3. Synthetic Data: In some cases, to increase the diversity and quantity of training data, synthetic data can be generated through computer programs. This data can be generated based on known physical or statistical models to simulate the data distribution in real-world scenarios. However, it is important to note that synthetic data should be as close to real data as possible to ensure model accuracy.

[0097] 4. Experimental Data: To obtain data under specific conditions, experiments can be conducted and relevant data collected. For example, tobacco outlet temperature and quality evaluation values can be recorded under different conditions of tobacco moisture, flow rate, and barrel wall temperature. This data can be used to train deep learning models and help them learn relationships under different conditions.

[0098] Therefore, after training the deep learning model, it will determine the relationship between the dependent and independent variables. Based on the preset target cut tobacco quality evaluation value, a reference value for the cut tobacco outlet temperature is obtained. In this embodiment, the reference value for the cut tobacco outlet temperature refers to a standard or reference value that the outlet temperature of the cut tobacco after being processed by the cut tobacco drying machine should reach under specific production conditions. In this embodiment, by calculating the reference value for the cut tobacco outlet temperature and using it as a standard, the adjustment amount of the drum wall temperature is limited, which is conducive to improving the quality of the cut tobacco.

[0099] In this embodiment, the target tobacco quality evaluation value is a comprehensive reflection of the sensory qualities (such as aroma, pungency, and odor) and physical qualities (such as moisture content, fill value, and whole-cut percentage) that the tobacco should achieve after drying. This evaluation value provides a benchmark for the MPC (Model Predictive Control) algorithm, enabling the system to adjust the barrel wall temperature to ensure that the tobacco maintains both sensory comfort and physical processing properties (such as fill capacity and processing resistance) during the drying process.

[0100] In this embodiment, the calculation of the target tobacco quality evaluation value requires the following parameters:

[0101] Physical quality parameters, including outlet moisture content (to be controlled within 12.0% to 14.0% (drum dryer) or a narrower range (such as ±0.1% tolerance)); filling value (cm 3 / g), which reflects the expansion effect of tobacco shreds and directly affects the weight and draw resistance of a single cigarette; whole shred rate and broken shred rate: the whole shred rate must be ≥ the standard value, and the broken shred rate must be ≤2.0%.

[0102] Sensory quality parameters include: pungency (the degree of improvement after high-temperature treatment to volatilize free nicotine and ammonia); smoke removal effect (a quantitative evaluation of smoke permeability enhancement); aroma style (e.g., the amount of browning reaction products between sugars and amino acids);

[0103] In this embodiment, the sensory quality parameters can be assigned values to each sensory quality parameter using an assignment method, for example, 9 points: excellent (such as "sufficient aroma"), 5 points: medium (such as "moderate stimulation"), and 1 point: extremely poor (such as "heavy smell").

[0104] Process stability parameters include the dry head and dry tail rate: the proportion of low-quality tobacco produced during abnormal stages (such as the startup period after preheating) must be ≤0.3% to 0.6%; temperature fluctuation: the tolerance range is usually ±3°C.

[0105] In this embodiment, the target cut tobacco quality evaluation value can be calculated using the following method.

[0106] Q = w1·f (moisture content) + w2·f (filling value) + w3·f (whole yarn rate) + w4·f (sensory score)

[0107] Among them, w1, w2, w3, and w4 are weight coefficients, and f(·) is the standardization or normalization function of each parameter.

[0108] In this embodiment, the target cut tobacco quality evaluation value is used as a standard and a reference value is obtained based on a deep learning model.

[0109] In S130, the MPC algorithm (Model Predictive Control) uses a mathematical model of the system to predict future states and optimize control decisions based on these predictions. It aims to achieve control objectives by optimizing system behavior over a period of time in the future.

[0110] The MPC algorithm establishes a mathematical model of the controlled object, which can be a state space model, a transfer function, etc. This model is used to predict the future behavior of the system. The objective function of the mathematical model in this embodiment can be expressed as:

[0111]

[0112] Among them, N p Represents the prediction time domain, the system predicts the future N at each step p Outlet temperature change in a time step (e.g., predict the next 30 seconds);

[0113] represents the predicted value of the tobacco outlet temperature;

[0114] T setIndicates the reference value of the outlet temperature of the cut tobacco;

[0115] N c Represents the control time domain, and calculates the future N at each step c The control amount is adjusted in time steps (usually N c ≤N p );

[0116] k2 represents the time;

[0117] i represents step;

[0118] Δu(k2+i) represents the cylinder wall temperature adjustment in step i at time k2;

[0119] represents the predicted value of tobacco outlet temperature at step i at time k2;

[0120] Q represents the weight of Δu(k2+i).

[0121] The essence of the objective function mentioned above is to achieve a balance between temperature tracking accuracy and control action smoothness within the prediction time domain. The formula can be broken down into two parts:

[0122] Represents the tracking error term, which makes the predicted tobacco outlet temperature As close to the reference value T as possible set .

[0123] It represents the penalty term for the control variable change, which limits the sharp fluctuation of the cylinder wall temperature adjustment amount Δu and avoids frequent operation of the actuator.

[0124] Optimization goal: Find the optimal balance between the weights Q and R.

[0125] In this embodiment, the operating logic of the above objective function in the control of the tofu drying machine can be:

[0126] 1. Rolling Optimization: In each control cycle k2, MPC performs the following steps:

[0127] Based on the current state (inlet moisture, cylinder wall temperature, etc.) and the prediction model, the future N p The outlet temperature prediction sequence of the step;

[0128] In N c In the step control time domain, solve the above objective function and obtain the optimal control sequence Δu(k2), Δu(k2+1),..., Δu(k2+N c -1);

[0129] Only the first step control amount Δu(k2) is implemented and the optimization is repeated in the next cycle.

[0130] 2. Dynamic constraint processing (implicit in the objective function):

[0131] Temperature limit: For example, the cylinder wall temperature must not exceed the upper limit of the material tolerance (such as 180°C);

[0132] Adjustment rate limit: for example, single-step temperature adjustment amount |Δu|≤2℃.

[0133] Flexible constraints: implemented indirectly through weights Q and R, rather than explicit hard constraints.

[0134] In this example, Q represents the weight matrix for tracking error, reflecting the importance of temperature tracking. As Q increases, the system adjusts the temperature more aggressively to approach the setpoint. R represents the weight matrix for control variable changes, reflecting the focus on actuator energy consumption or wear. As R increases, the system tends to be more conservative.

[0135] In this embodiment, the technical effects of using the MPC (Model Predictive Control) algorithm model are mainly reflected in the following aspects:

[0136] 1. Improved predictability and control accuracy: The MPC algorithm model collects the current state of the tobacco dryer (such as inlet tobacco moisture, current tobacco outlet temperature, and current drum wall temperature) and uses a predictive model to predict the tobacco outlet temperature at the next moment. This predictive control strategy enables the system to make adjustments in advance, avoiding the control inaccuracies caused by lag in traditional control methods. Furthermore, the MPC algorithm model, based on a rolling optimization strategy, re-predicts and optimizes within each control cycle, ensuring accurate and real-time control.

[0137] 2. Optimize production parameters and improve product quality: In step S120, the system obtains the preset target tobacco quality evaluation value and the current tobacco flow rate. Based on the trained deep learning model, it determines a baseline value for the tobacco outlet temperature. This baseline value reflects the tobacco outlet temperature required to achieve the target quality. After inputting the predicted and baseline values into the MPC algorithm model, the model comprehensively considers multiple factors and calculates the optimal barrel wall temperature adjustment to optimize production parameters and ensure that the tobacco outlet temperature remains within an appropriate range, thereby improving product quality.

[0138] 3. Enhanced system robustness and adaptability: The MPC algorithm model's ability to handle multiple variables and constraints makes it exceptionally effective in complex systems. In the tobacco drying machine temperature control system, the MPC algorithm comprehensively considers multiple influencing factors, such as inlet tobacco moisture, tobacco flow rate, and drum wall temperature, and flexibly adjusts based on actual conditions. This flexibility enables the system to better cope with various changes and uncertainties in the production process, enhancing its robustness and adaptability.

[0139] 4. Intelligent control improves production efficiency: Combining a deep learning model with the MPC algorithm model enables more intelligent control. The deep learning model predicts a baseline value for cut tobacco outlet temperature, providing accurate prediction information for the MPC algorithm. The MPC algorithm then performs rolling optimization based on the predicted and baseline values, outputting the barrel wall temperature adjustment. This intelligent control approach enables the system to more effectively address various production challenges and improve production efficiency.

[0140] In S140, if Figure 2 As shown, S140 may specifically include the following steps:

[0141] S141: Sending the cylinder wall temperature adjustment value to the PID controller;

[0142] S142: The PID controller obtains a proportional coefficient, an integral coefficient, and a differential coefficient based on the cylinder wall temperature adjustment amount;

[0143] S143: The PID controller obtains an opening value based on a proportional coefficient, an integral coefficient, and a differential coefficient, and controls the valve opening of the actuator to the opening value.

[0144] The PID controller obtains a proportional coefficient, an integral coefficient, and a differential coefficient based on the cylinder wall temperature adjustment amount, including:

[0145] The cylinder wall temperature adjustment amount is transferred to the PID parameter algorithm model to obtain the proportional coefficient, integral coefficient and differential coefficient. The PID parameter algorithm model is:

[0146] K p =K p0 +α·|Δu(k2+i)|

[0147] Among them: K p represents the proportionality coefficient;

[0148] K p0 Represents the initialization scale factor;

[0149] α represents the sensitivity coefficient of the proportional term;

[0150] Δu represents the cylinder wall temperature adjustment amount;

[0151]

[0152] Where: T i represents the integral coefficient;

[0153] T i0 Represents the initialized integral coefficient;

[0154] β represents the integral adaptation coefficient;

[0155] Δu(k2+i) represents the cylinder wall temperature adjustment in step i at time k2;

[0156]

[0157] Among them, T d represents the differential coefficient;

[0158] T d0 Represents the initialized differential coefficient;

[0159] γ represents the differential sensitivity coefficient.

[0160] In this embodiment, the valve in S143 may be a fuel valve, an air intake valve, a steam valve, etc.

[0161] For airflow tobacco drying machines, PID mainly controls the oil-gas ratio valve (fuel valve, air intake valve), taking into account both combustion efficiency and temperature stability; for drum tobacco drying machines: PID adjusts the steam valve and moisture exhaust valve, and controls the moisture content of tobacco by synergistically controlling the drum wall temperature and the hot air temperature.

[0162] In this embodiment, after obtaining the proportional coefficient, integral coefficient and differential coefficient, the PID controller outputs a control signal based on the proportional coefficient, integral coefficient and differential coefficient. After receiving the control signal, the actuator converts the control signal into a valve opening instruction, thereby adjusting the valve opening.

[0163] Please refer to Figure 3 The present application also provides a tow-bread drying machine temperature control system, which includes at least one software function module that can be stored in a storage module in the form of software or firmware, or embedded in an operating system (OS). The processing module is configured to execute the executable modules stored in the storage module, such as the software function modules and computer programs included in the tow-bread drying machine temperature control system.

[0164] like Figure 3 As shown, the functions of each module of the tofu drying machine temperature control system can be as follows:

[0165] The first calculation module 100 is configured to collect the moisture content, tobacco flow rate and cylinder wall temperature of the tobacco at the inlet of the tobacco drying machine, and obtain a predicted value of the tobacco outlet temperature based on the prediction model;

[0166] The second calculation module 200 is configured to obtain a preset target tobacco quality evaluation value and obtain a reference value of the tobacco outlet temperature based on the trained deep learning model;

[0167] The third calculation module 300 is configured to input the predicted value of the cut tobacco outlet temperature and the reference value of the cut tobacco outlet temperature into the MPC algorithm model and output the barrel wall temperature adjustment value;

[0168] The control module 400 is configured to adjust the temperature of the tofu drying machine based on the drum wall temperature adjustment amount.

[0169] The first calculation module 100 calculates the predicted value of the tobacco outlet temperature based on the following method.

[0170] The moisture content, flow rate and wall temperature of the cut tobacco are input into a prediction model, which outputs a predicted value of the cut tobacco outlet temperature. The prediction model is:

[0171] T out (k1+1)=α·T out (k1)+b·T wall (k1)+c·W in (k1)

[0172] Among them, T out (k1+1) is the predicted value of tobacco outlet temperature;

[0173] T out (k1) is the actual tobacco outlet temperature;

[0174] T wall (k1) is the current cylinder wall temperature;

[0175] W in (k1) is the current moisture content of the tobacco;

[0176] a, b, c are all model parameters.

[0177] The second calculation module 200 obtains a reference value of the outlet temperature of the cut tobacco based on the following method.

[0178] Initializing the deep learning model;

[0179] Based on the training data, the initialized deep learning model is trained to obtain a trained deep learning model, wherein the training data includes independent variable data and corresponding dependent variable data, the independent variable data includes training cut tobacco moisture, training cut tobacco flow rate, training barrel wall temperature, and training cut tobacco outlet temperature, and the dependent variable data includes a training cut tobacco quality evaluation value;

[0180] The current cut tobacco moisture, current cut tobacco flow, current barrel wall temperature and target cut tobacco quality evaluation value are input into the trained deep learning model to obtain a reference value of the cut tobacco outlet temperature.

[0181] The objective function of the MPC algorithm in the third calculation module 300 is:

[0182]

[0183] Among them, N p represents the prediction time domain;

[0184] represents the predicted value of the tobacco outlet temperature;

[0185] T set Indicates the reference value of the outlet temperature of the cut tobacco;

[0186] N c represents the control time domain;

[0187] k2 represents the time;

[0188] i represents step;

[0189] Δu(k2+i) represents the cylinder wall temperature adjustment in step i at time k2;

[0190] represents the predicted value of tobacco outlet temperature at step i at time k2;

[0191] Q represents the weight of Δu(k2+i).

[0192] The control module 400 adjusts the temperature of the tofu drying machine based on the following method.

[0193] Sending the cylinder wall temperature adjustment amount to a PID controller;

[0194] The PID controller obtains a proportional coefficient, an integral coefficient, and a differential coefficient based on the cylinder wall temperature adjustment amount;

[0195] The PID controller obtains an opening value based on a proportional coefficient, an integral coefficient, and a differential coefficient, and controls the valve opening of the actuator to the opening value.

[0196] Wherein: The PID controller obtains the proportional coefficient, integral coefficient and differential coefficient based on the cylinder wall temperature adjustment amount, including:

[0197] The cylinder wall temperature adjustment amount is transferred to the PID parameter algorithm model to obtain the proportional coefficient, integral coefficient and differential coefficient. The PID parameter algorithm model is:

[0198] K p =K p0 +α·|Δu(k2+i)|

[0199] Among them: K p represents the proportionality coefficient;

[0200] K p0 Represents the initialization scale factor;

[0201] α represents the sensitivity coefficient of the proportional term;

[0202] Δu represents the cylinder wall temperature adjustment amount;

[0203]

[0204] e(t)=

[0205] Where: T i represents the integral coefficient;

[0206] T i0 Represents the initialized integral coefficient;

[0207] β represents the integral adaptation coefficient;

[0208] Δu(k2+i) represents the cylinder wall temperature adjustment in step i at time k2;

[0209]

[0210] Among them, T d represents the differential coefficient;

[0211] T d0 Represents the initialized differential coefficient;

[0212] γ represents the differential sensitivity coefficient.

[0213] In this embodiment, the processing module can be an integrated circuit chip with signal processing capabilities. The above-mentioned processing module can be a general-purpose processor. For example, the processor can be a central processing unit (CPU), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, and can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application.

[0214] The storage module may be, but is not limited to, a random access memory, a read-only memory, a programmable read-only memory, an erasable programmable read-only memory, an electrically erasable programmable read-only memory, etc. In this embodiment, the storage module may be used to store a program, and the processing module executes the program after receiving an execution instruction.

[0215] It should be noted that those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working process of the electronic device described above can refer to the corresponding processes of each step in the aforementioned method, and will not be elaborated here.

[0216] The present application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed on a computer, causes the computer to execute the temperature control method for a tofu drying machine as described in the above embodiment.

[0217] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented through hardware or by means of software plus a necessary general hardware platform. Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, an electronic device, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0218] In the embodiments provided in the present application, it should be understood that the disclosed systems and methods can also be implemented in other ways. The system and method embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of code, and a part of the module, program segment or code includes one or more executable instructions for implementing the specified logical function. It should also be noted that each box in the block diagram and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart can be implemented by a dedicated hardware-based system that performs the specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions. In addition, the functional modules in the various embodiments of the present application can be integrated together to form an independent part, or each module can exist alone, or two or more modules can be integrated to form an independent part.

[0219] The above description is only an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification,

[0220] Equivalent replacements, improvements, etc. should all be included in the protection scope of this application.

Claims

1. A temperature control method for a tofu drying machine, characterized in that: The method comprises: Collect the current inlet moisture, outlet temperature and wall temperature of the tobacco at the tobacco drying machine; Inputting the current inlet tobacco moisture, current tobacco outlet temperature and current barrel wall temperature into the prediction model to obtain a predicted value of the tobacco outlet temperature at the next moment; Obtain the preset target tobacco quality evaluation value and current tobacco flow rate, and obtain the baseline value of tobacco outlet temperature based on the trained deep learning model; Inputting the predicted value of the cut tobacco outlet temperature and the reference value of the cut tobacco outlet temperature into the MPC algorithm model, and outputting the cylinder wall temperature adjustment value; The temperature of the tofu drying machine is adjusted based on the cylinder wall temperature adjustment amount.

2. The method according to claim 1, characterized in that The method of obtaining a predicted value of the cut tobacco outlet temperature based on the prediction model includes: The moisture content, flow rate and wall temperature of the cut tobacco are input into a prediction model, which outputs a predicted value of the cut tobacco outlet temperature. The prediction model is: T out (k1+1)=a·T out (k1)+b·T wall (k1)+c·W in (k1) Among them, T out (k1+1) is the predicted value of tobacco outlet temperature; T out (k1) is the actual tobacco outlet temperature; T wall (k1) is the current cylinder wall temperature; W in (k1) is the current moisture content of the tobacco; a, b, c are all model parameters.

3. The method according to claim 1, characterized in that The method of obtaining a reference value of the tobacco outlet temperature based on the trained deep learning model includes: Initializing the deep learning model; Based on the training data, the initialized deep learning model is trained to obtain a trained deep learning model, wherein the training data includes independent variable data and corresponding dependent variable data, the independent variable data includes training cut tobacco moisture, training cut tobacco flow rate, training barrel wall temperature, and training cut tobacco outlet temperature, and the dependent variable data includes a training cut tobacco quality evaluation value; The current cut tobacco moisture, current cut tobacco flow, current barrel wall temperature and target cut tobacco quality evaluation value are input into the trained deep learning model to obtain a reference value of the cut tobacco outlet temperature.

4. The method according to claim 1, wherein The objective function of the MPC algorithm model is: Among them, N p represents the prediction time domain; represents the predicted value of the tobacco outlet temperature; T set Indicates the reference value of the outlet temperature of the cut tobacco; N c represents the control time domain; k2 represents the time; i represents step; Δu(k2+i) represents the cylinder wall temperature adjustment in step i at time k2; represents the predicted value of tobacco outlet temperature at step i at time k2; Q represents the weight of Δu(k2+i).

5. The method according to claim 1, wherein The adjusting the temperature of the tow dryer based on the cylinder wall temperature adjustment amount includes: Sending the cylinder wall temperature adjustment amount to a PID controller; The PID controller is used to obtain a proportional coefficient, an integral coefficient, and a differential coefficient based on the cylinder wall temperature adjustment amount; An opening value is obtained based on the proportional coefficient, the integral coefficient, and the differential coefficient, and the valve opening of the actuator is controlled to the opening value.

6. The method according to claim 5, characterized in that The PID controller obtains a proportional coefficient, an integral coefficient, and a differential coefficient based on the cylinder wall temperature adjustment amount, including: The cylinder wall temperature adjustment amount is transferred to the PID parameter algorithm model to obtain the proportional coefficient, integral coefficient and differential coefficient. The PID parameter algorithm model is: K p =K p0 +α·|Δu(k2+i)| Among them: K p represents the proportionality coefficient; K p0 Represents the initialization scale factor; α represents the sensitivity coefficient of the proportional term; Δu represents the cylinder wall temperature adjustment amount; Where: T i represents the integral coefficient; T i0 Represents the initialized integral coefficient; β represents the integral adaptation coefficient; Δu(k2+i) represents the cylinder wall temperature adjustment in step i at time k2; Among them, T d represents the differential coefficient; T d0 Represents the initialized differential coefficient; γ represents the differential sensitivity coefficient.

7. A temperature control system for a tofu drying machine, characterized in that: include: The first calculation module is configured to collect the moisture content, tobacco flow rate and cylinder wall temperature of the tobacco at the inlet of the tobacco drying machine, and obtain a predicted value of the tobacco outlet temperature based on the prediction model; The second calculation module is configured to obtain a preset target cut tobacco quality evaluation value and obtain a reference value of the cut tobacco outlet temperature based on the trained deep learning model; a third calculation module configured to input the predicted value of the cut tobacco outlet temperature and the reference value of the cut tobacco outlet temperature into an MPC algorithm model and output a barrel wall temperature adjustment value; The control module is configured to adjust the temperature of the tofu drying machine based on the drum wall temperature adjustment amount.

8. An electronic device, characterized in that: The electronic device includes a processor and a memory coupled to each other, wherein a computer program is stored in the memory. When the computer program is executed by the processor, the electronic device executes the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is run on a computer, the computer is caused to execute the method according to any one of claims 1 to 6.

Images