A kiln control simulation method based on computational heat transfer model

Abstract

The application provides a kiln control simulation method based on a computational heat transfer model, and the method comprises the following steps: determining a parameter domain of a PID control model and membership rules; introducing a fuzzy controller to construct a fuzzy PID control model in combination with an optimized membership rule table, and correcting parameter values according to a fuzzy calculation result; constructing a computational heat transfer model of a kiln and performing reliability verification, coupling the computational heat transfer model with the fuzzy PID control model, simulating a furnace temperature field by means of the computational heat transfer model, and updating boundary conditions by calling the fuzzy PID control model when the calculation converges; performing multiple iterations, taking a correlation coefficient reaching a preset value as an optimal solution of parameters, and performing kiln control simulation. The application realizes dynamic simulation of a kiln temperature control process, makes the simulation more consistent with actual operation of the kiln, and improves the accuracy of the kiln control simulation.

Description

Technical Field

[0001] This invention relates to a kiln control simulation method, specifically a kiln control simulation method based on a computational heat transfer model, belonging to the field of industrial kiln control simulation technology. Background Technology

[0002] Industrial kilns have various structures, and the kiln temperature determines the quality of the fired products. Kiln temperature monitoring is mainly achieved through thermocouples installed on the kiln top.

[0003] The primary task of kiln temperature control is to maintain the temperature of the controlled object within the range required by the firing process. Current temperature control systems mainly employ classical PID control systems and fuzzy PID control systems. Fuzzy PID control systems derive fuzzy control rules based on experimental results and experience, and obtain a fuzzy control table through fuzzy inference, allowing the interaction of comprehensive parameters to fit to the optimal state. Fuzzy PID control systems have advantages such as fast response, small overshoot, and short transition time. As a complex thermodynamic system, the temperature of a kiln is affected by surrounding environmental conditions and equipment operation, exhibiting significant uncertainty. In the past, control system simulations often used simple first- or second-order inertial systems with pure time delays as a substitute. Therefore, directly applying the fuzzy PID control parameters obtained from simulations to actual kiln operation often results in significantly reduced effectiveness.

[0004] Computational heat transfer models (NHT models) have significant advantages in describing the heat transfer characteristics of a system. By establishing an appropriate kiln model, detailed temperature distribution inside the furnace and the influence of different thermodynamic parameters and boundary conditions on the heat transfer process can be obtained. Using this model for fuzzy PID control system simulation can yield reliable control parameters and effectively improve the kiln's operating performance. However, there is currently no research combining computational heat transfer models and fuzzy PID control systems. Summary of the Invention

[0005] Based on the above background, the purpose of this invention is to provide a kiln control simulation method based on a computational heat transfer model, so as to realize the dynamic simulation of the kiln temperature control process, make the simulation more consistent with the actual operation of the kiln, and improve the accuracy of kiln control simulation.

[0006] To achieve the above-mentioned objectives, the present invention provides the following technical solution:

[0007] A kiln control simulation method based on a computational heat transfer model, comprising the following steps:

[0008] Determine the proportional parameter K of the PID control model p Integral parameter K i and differential parameter K dThe domain of discourse and membership rules are optimized based on the gas valve flow regulation rules recorded during the actual firing process in the kiln.

[0009] A PID control model is constructed, and the proportional parameter K of the PID control model is obtained through a self-tuning method, based on the transfer function of the kiln. p Integral parameter K i and differential parameter K d The value of is determined, and a fuzzy PID control model is constructed by introducing a fuzzy controller based on the optimized membership rule table. Combined with the kiln's transfer function, the deviation e between the measured furnace temperature and the set furnace temperature, as well as the rate of change of deviation ec, are used as inputs to the fuzzy controller for fuzzy calculation. The proportional parameter K is then adjusted based on the fuzzy calculation results. p Integral parameter K i and differential parameter K d The possible values ​​of ;

[0010] A computational heat transfer model of the kiln is constructed based on parameters such as the actual kiln structure dimensions, gas valve flow rate, and thermal properties of the object being fired, and its reliability is verified. The computational heat transfer model is coupled with a fuzzy PID control model. The computational heat transfer model simulates the furnace temperature field. When the calculation converges, the fuzzy PID control model is called to update the boundary conditions.

[0011] By comparing the actual furnace temperature field and the simulated furnace temperature field, the proportional parameter K was optimized. p Integral parameter K i and differential parameter K d The value of is selected and iterated multiple times until the correlation coefficient between the simulated furnace temperature field and the actual furnace temperature field calculated based on the optimized value reaches a preset value, which is then used as the proportional parameter K. p Integral parameter K i and differential parameter K d The optimal solution;

[0012] Kiln control simulation was performed using a computational heat transfer model based on a coupled fuzzy PID control model with the aforementioned optimal solution.

[0013] Preferably, the transfer function of the kiln is constructed based on a second-order inertial system with pure time delay, and the transfer function of the kiln is:

[0014]

[0015] Where i is the thermocouple number, K mi T is the magnification factor. mi T is the time constant, τ is the pure time delay, and T is the time delay. i For the control system output, M i For input to the control system.

[0016] Preferably, the convergence criterion for the calculation is:

[0017] When t-t1=2△t, the calculation is considered convergent;

[0018] Where t is the total simulation time, t1 is the time of the previous convergence result, the initial value of t1 is set to 0, and Δt is the time step; when t-t1 = Δt, it means that the first step has not converged, and the iterative calculation continues; when t-t1 = 2Δt, it means that the first step of iteration has converged, let t = t1 + Δt, and perform the second iteration, and so on. t1 is only updated at the time of convergence, that is, it increments by one time step Δt.

[0019] As a preferred method, when determining the convergence of the calculation, a Boolean value S is set to indicate whether the previous convergence temperature result has called the fuzzy PID control model. When S=0, the previous convergence temperature result has not called the fuzzy PID control model, and when S=1, the previous convergence temperature result has called the fuzzy PID control model.

[0020] Preferably, the operating logic of the fuzzy PID control model is as follows:

[0021] Set the single control cycle t T Set the flow rate change rate q, and let the sampling interval t s Equal to the time step Δt, the simulation time reaches t. T When the flow rate is an integer multiple of the value, the fuzzy PID control result is output once, which is the gas valve flow rate value Q1 for the next control cycle. Then, based on the valve flow rate Q0 at the beginning of the current control cycle, the time point t for starting to adjust the gas valve flow rate within a control cycle is calculated using the following formula. c :

[0022] Q1 = Fuzzy_PID(t T )

[0023]

[0024] Set the time parameter t within the period h This represents the time elapsed within one cycle of the control system, and it increments by Δt when the calculation converges, i.e., t h =t h +△t, and reset to △t at the start of the next cycle, i.e., t h = Δt; In each fuzzy PID control cycle, when t h <t c At time t, the valve flow rate Q = Q0; when t h >t c At that time, the valve flow rate Q = Q0 ± (t h -t c)×q, where the plus sign is taken when Q1≥Q0 and the minus sign is taken when Q1<Q0 in the above formula;

[0025] According to t h and t c Output the gas valve flow value according to the results.

[0026] Preferably, the output of the gas valve flow value according to t h and t c The results specifically include the following steps:

[0027] When the calculation does not converge, judge and compare t h with t s and calculate the gas valve flow value;

[0028] When the calculation converges and the previous convergence temperature result does not call the fuzzy PID control model, judge and compare t h with t s and calculate the gas valve flow value;

[0029] When the calculation converges and the previous convergence temperature result calls the fuzzy PID control model and the current control time does not exceed the control period, that is, t h ≤t T The fuzzy PID control model calculates the gas valve flow value for the next control period, but does not output the time point t c at which the valve flow starts to be adjusted. Let t h increment by an iterative time step △t. Finally, judge and compare t h with t s and calculate the gas valve flow value;

[0030] When the calculation converges and the previous convergence temperature result calls the fuzzy PID control model and the current control time has exceeded the control period, that is, t h >t T The fuzzy PID control model calculates the gas valve flow value for the next control period and outputs the time point t c at which the valve flow starts to be adjusted. Let t h be initialized to the iterative time step △t. Finally, judge and compare t h with t s and calculate the gas valve flow value.

[0031] Preferably, the preset value of the correlation coefficient is 0.996, and the calculation formula of the correlation coefficient is

[0032]

[0033] where is the actual average furnace temperature, is the simulated average furnace temperature.

[0034] Preferably, the steps of the kiln control simulation are as follows:

[0035] Variable initialization: Set the initial temperature T0, and let the current temperature T m =T0, target temperature T t Fuzzy PID control period t p The previous convergence time t1 indicates whether the previous convergence temperature result used the Boolean value S of the fuzzy PID control model, and the self-tuning proportional parameter K. p Integral parameter K i and differential parameter K d ;

[0036] Get the current time t;

[0037] To determine if the calculation is convergent, update T when t - t1 = 2Δt. m Then determine whether the current time t exceeds the total simulation time t. end If the value exceeds the limit, the simulation ends; otherwise, proceed to the next step. When t-t1≠2△t, proceed directly to the next step.

[0038] Determine the current temperature T m The temperature range in which it is located, if T m Below the fuzzy PID control range, keep the gas nozzle continuously operating and increase T. m Set the Boolean value S = 0; if T m Within the fuzzy PID control range, the fuzzy PID control model is invoked to calculate the gas valve flow rate, and the Boolean value S = 1 is set; if T m If the value exceeds the fuzzy PID control range, shut off the gas nozzle and set the Boolean value S = 0;

[0039] The gas nozzle opening parameters are adjusted based on the calculated gas valve flow rate.

[0040] Check the convergence again. If t - t1 = 2Δt, then set t1 = t1 + Δt; otherwise, proceed to the next step.

[0041] Call the computational heat transfer model to calculate the simulated furnace temperature field until the calculation converges, then obtain the current time t and repeat the subsequent steps;

[0042] Repeat the above steps until the current time t exceeds the total simulation time t. end ,Finish.

[0043] Preferably, in the construction of the fuzzy PID control model, the proportional parameter K... p Integral parameter K i and differential parameter K dThe value setting and correction are implemented through a MATLAB Simulink program.

[0044] As a preferred approach, the computational heat transfer model is coupled with the fuzzy PID control model, which is implemented using a user-defined function in the Fluent program.

[0045] Compared with the prior art, the present invention has the following advantages:

[0046] This invention discloses a kiln control simulation method based on a computational heat transfer model, applicable to temperature field simulation of various kilns and development of kiln temperature control systems. Based on the computational heat transfer model, this invention utilizes the coupling of the temperature field and control system to intuitively demonstrate the control effect, facilitating the optimization of control parameters during the development of the kiln temperature control system. Furthermore, this invention does not require accurate solution of the kiln system's transfer function during simulation. Compared to the simulation results using simple transfer functions in existing technologies, this method better reflects the actual operating state of the kiln, improving the control effect during actual kiln operation. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0048] Figure 1 This is a schematic diagram illustrating the principle of a kiln control simulation method based on a computational heat transfer model according to the present invention.

[0049] Figure 2 This is a flowchart of the fuzzy PID control model implemented by the Fluent program in the form of user-defined functions in this invention.

[0050] Figure 3 This is a flowchart of the kiln control simulation in this invention;

[0051] Figure 4 This is a flowchart of the optimization process for control parameters in this invention;

[0052] Figure 5 These are the fuzzy subsets and fuzzification level diagrams in this invention;

[0053] Figure 6 The parameter ΔK in this invention p Fuzzy rule table;

[0054] Figure 7 The parameter ΔK in this invention iFuzzy rule table;

[0055] Figure 8 The parameter ΔK in this invention d Fuzzy rule table;

[0056] Figure 9 This is a diagram of the fuzzy PID control system implemented using the MATLAB Simulink program in this invention;

[0057] Figure 10 This is a diagram showing the results of the fuzzy PID control system in this invention;

[0058] Figure 11 This is an NHT model diagram of the shuttle kiln in this invention;

[0059] Figure 12 This is a temperature comparison chart of the NHT model with different control parameters in this invention. Detailed Implementation

[0060] The technical solution of the present invention will be further described in detail below through specific embodiments and in conjunction with the accompanying drawings. It should be understood that the implementation of the present invention is not limited to the following embodiments, and any modifications and / or alterations made to the present invention will fall within the protection scope of the present invention.

[0061] In this invention, unless otherwise specified, all parts and percentages are by weight, and the equipment and raw materials used are commercially available or commonly used in the art. Unless otherwise specified, the methods in the following embodiments are conventional methods in the art. Unless otherwise specified, the components or equipment in the following embodiments are general standard parts or components known to those skilled in the art, and their structures and principles can be learned by those skilled in the art through technical manuals or conventional experimental methods.

[0062] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings. In the following detailed description, many specific details are set forth to facilitate explanation and provide a full understanding of the embodiments of the present invention.

[0063] refer to Figures 1-4 The present invention discloses a kiln control simulation method based on a computational heat transfer model, the method comprising the following steps:

[0064] Determine the proportional parameter K of the PID control model p Integral parameter K i and differential parameter K d The domain of discourse and membership rules are optimized based on the gas valve flow regulation rules recorded during the actual firing process in the kiln.

[0065] A PID control model is constructed, and the proportional parameter K of the PID control model is obtained through a self-tuning method, based on the transfer function of the kiln. p Integral parameter K i and differential parameter K d The value of is determined, and a fuzzy PID control model is constructed by introducing a fuzzy controller based on the optimized membership rule table. Combined with the kiln's transfer function, the deviation e between the measured furnace temperature and the set furnace temperature, as well as the rate of change of deviation ec, are used as inputs to the fuzzy controller for fuzzy calculation. The proportional parameter K is then adjusted based on the fuzzy calculation results. p Integral parameter K i and differential parameter K d The value of ; where, in the construction of the fuzzy PID control model, the proportional parameter K p Integral parameter K i and differential parameter K d The value setting and correction are implemented through a MATLAB Simulink program;

[0066] A computational heat transfer model of the kiln is constructed based on parameters such as the actual kiln structure dimensions, gas valve flow rate, and thermal properties of the fired object, and its reliability is verified. The computational heat transfer model is coupled with a fuzzy PID control model. Specifically, the fuzzy PID control model is implemented in the form of a user-defined function (UDF) through the Fluent program. The computational heat transfer model simulates the furnace temperature field, and the fuzzy PID control model is called to update the boundary conditions when the calculation converges.

[0067] By comparing the actual furnace temperature field and the simulated furnace temperature field, the proportional parameter K was optimized. p Integral parameter K i and differential parameter K d The value of is selected and iterated multiple times until the correlation coefficient between the simulated furnace temperature field and the actual furnace temperature field calculated based on the optimized value reaches a preset value, which is then used as the proportional parameter K. p Integral parameter K i and differential parameter K d The optimal solution;

[0068] Kiln control simulation was performed using a computational heat transfer model based on a coupled fuzzy PID control model with the aforementioned optimal solution.

[0069] In the above method steps, the transfer function of the kiln is constructed based on a second-order inertial system with pure time delay. The transfer function of the kiln is:

[0070]

[0071] Where i is the thermocouple number, K mi T is the magnification factor. miT is the time constant, τ is the pure time delay, and T is the time delay. i For the control system output, M i For input to the control system.

[0072] In the above method steps, the convergence criterion is calculated as follows:

[0073] When t-t1=2△t, the calculation is considered convergent;

[0074] Where t is the total simulation time, t1 is the time of the previous convergence result, the initial value of t1 is set to 0, and Δt is the time step; when t-t1 = Δt, it means that the first step has not converged, and the iterative calculation continues; when t-t1 = 2Δt, it means that the first step of iteration has converged, let t = t1 + Δt, and perform the second iteration, and so on. t1 is only updated at the time of convergence, that is, it increments by one time step Δt.

[0075] When determining convergence, a Boolean value S is set to indicate whether the previous convergence temperature result called the fuzzy PID control model. When S=0, the previous convergence temperature result did not call the fuzzy PID control model, and when S=1, the previous convergence temperature result called the fuzzy PID control model.

[0076] In the above method steps, the operating logic of the fuzzy PID control model is as follows:

[0077] Set the single control cycle t T Set the flow rate change rate q, and let the sampling interval t s Equal to the time step Δt, the simulation time reaches t. T When the flow rate is an integer multiple of the value, the fuzzy PID control result is output once, which is the gas valve flow rate value Q1 for the next control cycle. Then, based on the valve flow rate Q0 at the beginning of the current control cycle, the time point t for starting to adjust the gas valve flow rate within a control cycle is calculated using the following formula. c :

[0078] Q1 = Fuzzy_PID(t T )

[0079]

[0080] Set the time parameter t within the period h This represents the time elapsed within one cycle of the control system, and it increments by Δt when the calculation converges, i.e., t h =t h +△t, and reset to △t at the start of the next cycle, i.e., t h = Δt; In each fuzzy PID control cycle, when t h <t c At time t, the valve flow rate Q = Q0; when t h >tc When, the valve flow rate Q = Q0 ± (t h - t c ) × q. In the above formula, when Q1 ≥ Q0, take +, and when Q1 < Q0, take -;

[0081] According to t h and t c Output the gas valve flow rate value, which specifically includes the following steps:

[0082] When the calculation does not converge, judge and compare t h with t s to calculate the gas valve flow rate value;

[0083] When the calculation converges and the previous convergence temperature result does not call the fuzzy PID control model, judge and compare t h with t s to calculate the gas valve flow rate value;

[0084] When the calculation converges and the previous convergence temperature result calls the fuzzy PID control model and the current control time does not exceed the control period, i.e., t h ≤ t T , the fuzzy PID control model calculates the gas valve flow rate value for the next control period, but does not output the time point t c when starting to adjust the valve flow rate. Let t h increment by an iterative time step △t. Finally, judge and compare t h with t s to calculate the gas valve flow rate value;

[0085] When the calculation converges and the previous convergence temperature result calls the fuzzy PID control model and the current control time has exceeded the control period, i.e., t h > t T , the fuzzy PID control model calculates the gas valve flow rate value for the next control period and outputs the time point t c when starting to adjust the valve flow rate. Let t h be initialized to the iterative time step △t. Finally, judge and compare t h with t s to calculate the gas valve flow rate value.

[0086] In the above method steps, the preset value of the correlation coefficient is 0.996, and the calculation formula of the correlation coefficient is

[0087]

[0088] where, is the actual average furnace temperature, is the simulated average furnace temperature.

[0089] In the above method and steps, the kiln control simulation steps are as follows:

[0090] Variable initialization: Set the initial temperature T0, and let the current temperature T m =T0, target temperature T t Fuzzy PID control period t p The previous convergence time t1 indicates whether the previous convergence temperature result used the Boolean value S of the fuzzy PID control model, and the self-tuning proportional parameter K. p Integral parameter K i and differential parameter K d ;

[0091] Get the current time t;

[0092] To determine if the calculation is convergent, update T when t - t1 = 2Δt. m Then determine whether the current time t exceeds the total simulation time t. end If the value exceeds the limit, the simulation ends; otherwise, proceed to the next step. When t-t1≠2△t, proceed directly to the next step.

[0093] Determine the current temperature T m The temperature range in which it is located, if T m Below the fuzzy PID control range, keep the gas nozzle continuously operating and increase T. m Set the Boolean value S = 0; if T m Within the fuzzy PID control range, the fuzzy PID control model is invoked to calculate the gas valve flow rate, and the Boolean value S = 1 is set; if T m If the value exceeds the fuzzy PID control range, shut off the gas nozzle and set the Boolean value S = 0;

[0094] The gas nozzle opening parameters are adjusted based on the calculated gas valve flow rate.

[0095] Check the convergence again. If t - t1 = 2Δt, then set t1 = t1 + Δt; otherwise, proceed to the next step.

[0096] Call the computational heat transfer model to calculate the simulated furnace temperature field until the calculation converges, then obtain the current time t and repeat the subsequent steps;

[0097] Repeat the above steps until the current time t exceeds the total simulation time t. end ,Finish.

[0098] The following section uses a shuttle kiln as an example to explain in detail how to construct a shuttle kiln control simulation method based on a computational heat transfer model according to the above steps. The shuttle kiln structure is like a shuttle, and it is an intermittently pulled kiln. After the kiln car is filled with celadon, it is pushed into the kiln. After burning, the fired celadon is taken out in the reverse direction.

[0099] A simulation method for shuttle kiln control based on a computational heat transfer model mainly includes three parts: the formulation of expert rules, the establishment of a fuzzy PID control model, and the coupling of the shuttle kiln NHT model with the fuzzy PID control model.

[0100] The formulation of expert rules requires determining the K of the PID. p K i K d The domain of discourse and membership degree of the three parameters are optimized based on the existing classical expert rules and the gas valve pressure regulation rules recorded during the actual firing process.

[0101] The deviation between the actual measured kiln temperature T and the set temperature Td is e = TT d The rate of change of its deviation, ec, is used as the input parameter of the fuzzy controller, ΔK. p ΔK i ΔK d As the output of a fuzzy PID controller.

[0102] To achieve the initial and fine-tuning of the above parameters, they are transformed into a fuzzy universe of discourse using the following method:

[0103]

[0104]

[0105] Similarly, the fuzzy universes of discourse for the other input and output parameters can be obtained, and the fuzzy universes of discourse after fuzzy quantization are:

[0106] e∈[-3, 3], ec∈[-3, 3], ΔK p ∈[-3, 3],

[0107] ΔK i ∈[-3, 3], ΔK d ∈[-3, 3]

[0108] Define e, ec, ΔK p ,ΔK i and ΔK d The fuzzy subset is [NB, NM, NS, Z, PS, PM, PB], where NB represents negative large, NM represents negative medium, NS represents negative small, Z represents zero, PS represents positive small, PM represents positive medium, and PB represents positive large.

[0109] Let e, ec, ΔK p ΔK i and ΔK dAll fuzzy subsets follow a normal distribution, and their quantization factors are all 1. The membership function of each fuzzy subset on the corresponding fuzzy universe can have multiple representations; here, the triangle (trimf) representation is used. The expression for the triangle membership function is:

[0110]

[0111] Where: x is the domain of the variable; a, b, and c specify the shape of the triangular function. The corresponding fuzzy subsets and fuzzification levels are referenced. Figure 5 As shown.

[0112] The control rules are as follows:

[0113] When the deviation |e| is large, in order to improve the system's response speed, K should be made larger. p To take a larger value, and to avoid differential saturation of e in the initial stage, K should be made larger. d The value is relatively small. However, to prevent large overshoot, the integral should be limited, and K should be chosen. i =0;

[0114] When |e| and |ec| are of moderate magnitude, a smaller K is chosen to minimize response overshoot. p In this case, K d It has a significant impact on the system response, so the value should be moderate to ensure the speed of response;

[0115] When |e| is small and the set value is close to the target value, K should be increased. p and K i The value of K is chosen to improve the steady-state performance of the system, but to avoid oscillations and consider anti-interference capabilities, when the value of |ec| is small, K... d The value should be appropriately large; when the value of |ec| is large, K d The value should be appropriately small.

[0116] According to the changes in e and ec above, K p K i and K d The self-tuning principle can be used to derive a reference. Figures 6-8 The ΔK shown p ΔK i and ΔK d A fuzzy rule table.

[0117] The expert rule is:

[0118] Because the heating control system exhibits significant lag, output control is necessary in the later stages of the heating and stabilization phases to ensure subsequent deviations remain within acceptable limits. In this example, a knowledge base is used to obtain this data to analyze the need for pre-control and the appropriate timing for it. The design can be based on the following four scenarios:

[0119] When |e|≥m, where m is the set error threshold, it indicates that the error is large at this time, and the controller output should be adjusted to the maximum.

[0120] When e×ec < 0 and e×ec(k-1) < 0, it indicates that the absolute value of the error is decreasing. In this case, we should consider keeping the controller output constant.

[0121] When e×ec≥0, it indicates that the error is increasing; if |e|≥n (where n is the set error threshold) and m>n, it indicates that the error is relatively large, and the controller output is:

[0122] u(k)=u(k-1)+k1{k p [e(k)-e(k-1)]+k i e(k)+k d [e(k)-2e(k-1)+e(k-2)]}

[0123] If |e| < n, it indicates that the error is increasing, but the error value is still within the threshold. In this case, the output is:

[0124] u(k)=u(k-1)+k p [e(k)-e(k-1)]+k i e(k)+k d [e(k)-2e(k-1)+e(k-2)

[0125] When e×ec < 0 and e×ec(k-1) < 0, if the error is large at this time, the controller output should be kept large. The controller output at this time is:

[0126] u(k)=u(k-1)+k1k p e(k)

[0127] If the error is small at this point, the controller output should be kept relatively weak. The controller output at this time would be:

[0128] u(k)=u(k-1)+k2k p e(k)

[0129] Where k1 is the gain amplification factor and k2 is the suppression factor, i.e., k1 > 1 and 0 < k2 < 1.

[0130] Based on the actual measured kiln temperature characteristic curve, the parameters of the transfer function can be further determined. The final transfer function between the kiln gas flow rate and the furnace temperature can be expressed as:

[0131]

[0132] A classic PID control model was built on Matlab Simulink software, and K was obtained through self-tuning. p K i K d Three key parameters. A reference model is constructed based on the classic PID control model and combined with a membership rule table. Figure 9 The fuzzy PID control model shown is simulated using the actual temperature characteristic curve of a kiln as input, and K is finally set. p =10, K i =0.1, K d =5 as a preliminary result. From Figure 10 The results of the fuzzy PID control system show that the initially set parameter values ​​can roughly meet the control requirements of the kiln temperature curve.

[0133] Based on the actual measured parameters such as the structural dimensions of the shuttle kiln, valve gas flow rate, and ceramic thermophysical properties, a system is established as follows: Figure 11 The NHT model of the shuttle kiln is shown. The changes in the temperature and flow fields within the shuttle kiln during firing are calculated, and the reliability of the NHT model is verified using actual measured data. Table 1 shows the parameters of the kiln in this embodiment:

[0134] Table 1 Kiln Parameters

[0135] shuttle kiln parameters unit value Furnace body length m 4.1 Furnace body width m 3.0 Furnace height m 3.2 burner diameter mm 200 Maximum furnace temperature ℃ 1280-1290 Gas flow <![CDATA[m 3 / h]]> 64 Natural gas import pressure MPa 0.027 Flue gas outlet velocity m / s 9 flue gas outlet length m 0.36 flue gas outlet width m 0.28

[0136] The calculation process of the NHT model is as follows: According to this embodiment, the single control period t of the fuzzy PID is set. T =10s, set the flow rate change rate q = 0.001, and let the sampling interval t s =0.5s, therefore the control frequency of the fuzzy PID in one cycle is 20 times. Before Fluent calculation, the initial temperature T0 is set to 27℃, and the current temperature T is set to... m =T0, target temperature T t =1290℃, time step Δt = t s =0.5s, total calculation time 10000s, upper limit of temperature control band T supper =1300℃, lower limit of temperature control zone is T lower =10℃.

[0137] During the calculation process, the temperature monitoring point is set at the top center of the kiln. When the iteration converges, the temperature is output and saved.

[0138] K was gradually optimized through multiple NHT model simulations. p K i and K dThe temperature curve obtained from the first simulation was not optimized using the NHT model and is recorded as the zeroth iteration. In the second simulation, the control parameters were adjusted based on the results of the first simulation, and this process was repeated thereafter until the K value was obtained after the third simulation. p =12.1, K i =0.09, K d =6. The final temperature curve will match the expected result at K. p K i and K d As the optimal value.

[0139] refer to Figure 12 The temperature comparison chart of the NHT model with different control parameters shows that the simulated temperature curve after adjusting the control parameters of the NHT model is more consistent with the temperature curve in the actual production process of the kiln. The correlation coefficient R is calculated by comparing the temperature obtained after each iteration of optimization with the target temperature, and the results are shown in Table 2.

[0140] Table 2 Correlation coefficients

[0141] Number of iterations Correlation coefficient R <![CDATA[For the zero - th time, K p = 10, K i = 0.1, K d = 5]]> 0.993 <![CDATA[For the first time, K p = 12, K i = 0.1, K d = 5]]> 0.994 <![CDATA[Second time, K p = 12, K i = 0.09, K d = 5]]> 0.995 <![CDATA[Third time, K p = 12.6, K i = 0.09, K d = 6]]> 0.997

[0142] The method in this embodiment is more accurate and better meets actual production needs compared to traditional control system simulation methods.

[0143] This article uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. It should be noted that those skilled in the art can make several improvements and modifications to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A kiln control simulation method based on a computational heat transfer model, characterized in that: The kiln control simulation method based on a computational heat transfer model includes the following steps: Determine the proportional parameter K of the PID control model p Integral parameter K i and differential parameter K d The domain of discourse and membership rules are optimized based on the gas valve flow regulation rules recorded during the actual firing process in the kiln. A PID control model is constructed, and the proportional parameter K of the PID control model is obtained through a self-tuning method, based on the transfer function of the kiln. p Integral parameter K i and differential parameter K d The value of is determined, and a fuzzy PID control model is constructed by introducing a fuzzy controller based on the optimized membership rule table. Combined with the kiln's transfer function, the deviation e between the measured furnace temperature and the set furnace temperature, as well as the rate of change of deviation ec, are used as inputs to the fuzzy controller for fuzzy calculation. The proportional parameter K is then adjusted based on the fuzzy calculation results. p Integral parameter K i and differential parameter K d The possible values ​​of ; A computational heat transfer model of the kiln is constructed based on parameters such as the actual kiln structure dimensions, gas valve flow rate, and thermal properties of the object being fired, and its reliability is verified. The computational heat transfer model is coupled with a fuzzy PID control model. The computational heat transfer model simulates the furnace temperature field. When the calculation converges, the fuzzy PID control model is called to update the boundary conditions. By comparing the actual furnace temperature field and the simulated furnace temperature field, the proportional parameter K was optimized. p Integral parameter K i and differential parameter K d The value of is selected and iterated multiple times until the correlation coefficient between the simulated furnace temperature field and the actual furnace temperature field calculated based on the optimized value reaches a preset value, which is then used as the proportional parameter K. p Integral parameter K i and differential parameter K d The optimal solution; Kiln control simulation was performed using a computational heat transfer model based on a coupled fuzzy PID control model with the aforementioned optimal solution. The operating logic of the fuzzy PID control model is as follows: Set the single control cycle t T Set the flow rate change rate q, and let the sampling interval t s Equal to the time step Δt, the simulation time reaches t. T When the flow rate is an integer multiple of the value, the fuzzy PID control result is output once, which is the gas valve flow rate value Q1 for the next control cycle. Then, based on the valve flow rate Q0 at the beginning of the current control cycle, the time point t for starting to adjust the gas valve flow rate within a control cycle is calculated using the following formula. c : ， ， Time parameter t within a set period h Indicates the time elapsed within one cycle of the control system. When the calculation converges, it increments by △t, i.e., t h =t h +△t. When reaching the next cycle, it is reset to △t, i.e., t h =△t; In each fuzzy PID control cycle, when t h <t c , the valve flow rate Q = Q0; When t h >t c , the valve flow rate Q = Q0 ± (t h -t c ) × q. In the above formula, + is taken when Q1 ≥ Q0, and - is taken when Q1 < Q0; According to t h and t c The output of the gas valve flow rate value includes the following steps: When the computation fails to converge, compare t. h With t s Calculate the gas valve flow rate based on the size of the valve. When the calculation converges and the previous convergence temperature result did not call the fuzzy PID control model, the comparison t is determined. h With t s Calculate the gas valve flow rate based on the size of the valve. When the calculation converges, the previous convergent temperature result calls the fuzzy PID control model, and the current control time has not exceeded the control cycle (t), the fuzzy PID control model is invoked. h ≤t T At that time, the fuzzy PID control model calculates the gas valve flow rate for the next control cycle, but does not output the time point t at which the valve flow rate adjustment begins. c , let t h Increment the iteration time step by Δt, and finally compare t with the result. h With t s Calculate the gas valve flow rate based on the size of the valve. When the calculation converges and the previous convergent temperature result calls the fuzzy PID control model, and the current control time has exceeded the control cycle (t),... h >t T At that time, the fuzzy PID control model calculates the gas valve flow rate value for the next control cycle and outputs the time point t at which the valve flow rate adjustment begins. c , let t h Initialize to the iteration time step Δt, and finally compare t. h With t s Calculate the flow rate of the gas valve based on the size of the valve.

2. The kiln control simulation method based on a computational heat transfer model according to claim 1, characterized in that: The transfer function of the kiln is constructed based on a second-order inertial system with pure time delay. The transfer function of the kiln is as follows: ， Where i is the thermocouple number, K mi T is the magnification factor. mi Let T be the time constant, τ be the pure time delay, and T be the time delay. i For the control system output, M i For input to the control system.

3. The kiln control simulation method based on a computational heat transfer model according to claim 1, characterized in that: The convergence criterion for the calculation is as follows: When t - t1 = 2Δt, the calculation is considered convergent; Where t is the total simulation time, t1 is the time of the previous convergence result, the initial value of t1 is set to 0, and Δt is the time step; when t-t1=Δt, it means that the first step has not converged, and the iterative calculation continues; when t-t1=2Δt, it means that the first step of iteration has converged, let t=t1+Δt, and perform the second step of iteration, and so on. t1 is only updated at the time of convergence, that is, it increments by one time step Δt.

4. The kiln control simulation method based on a computational heat transfer model according to claim 1, characterized in that: When determining convergence, a Boolean value S is set to indicate whether the previous convergence temperature result called the fuzzy PID control model. When S=0, the previous convergence temperature result did not call the fuzzy PID control model, and when S=1, the previous convergence temperature result called the fuzzy PID control model.

5. The kiln control simulation method based on a computational heat transfer model according to claim 1, characterized in that: The preset value for the correlation coefficient is 0.996, and the formula for calculating the correlation coefficient is as follows: ,in, This represents the actual average temperature of the furnace. To simulate the average temperature of the furnace.

6. The kiln control simulation method based on a computational heat transfer model according to claim 1, characterized in that: The steps for kiln control simulation are as follows: Variable initialization: Set the initial temperature T0, and let the current temperature T m =T0, target temperature T t Fuzzy PID control period t p The previous convergence time t1, the time step Δt, the Boolean value S indicating whether the previous convergence temperature result called the fuzzy PID control model, and the self-tuning proportional parameter K. p Integral parameter K i and differential parameter K d ; Get the current time t; To determine if the calculation is convergent, update T when t - t1 = 2Δt. m Then determine whether the current time t exceeds the total simulation time t. end If the value exceeds the limit, the simulation ends; otherwise, proceed to the next step. When t-t1≠2△t, proceed directly to the next step. Determine the current temperature T m The temperature range in which it is located, if T m Below the fuzzy PID control range, keep the gas nozzle continuously operating and increase T. m Set the Boolean value S=0; if T m Within the fuzzy PID control range, the fuzzy PID control model is invoked to calculate the gas valve flow rate, and the Boolean value S=1 is set; if T m If the value exceeds the fuzzy PID control range, shut off the gas nozzle and set the Boolean value S=0; The gas nozzle opening parameters are adjusted based on the calculated gas valve flow rate. The calculation is checked again to determine if it converges. If t - t1 = 2Δt, then set t1 = t1 + Δt; otherwise, proceed to the next step. Call the computational heat transfer model to calculate the simulated furnace temperature field until the calculation converges, then obtain the current time t and repeat the subsequent steps; Repeat the above steps until the current time t exceeds the total simulation time t. end ,Finish.

7. The kiln control simulation method based on a computational heat transfer model according to claim 1, characterized in that: The construction of the fuzzy PID control model, the proportional parameter K p Integral parameter K i and differential parameter K d The value setting and correction are implemented through a MATLAB Simulink program.

8. The kiln control simulation method based on a computational heat transfer model according to claim 1, characterized in that: The method of coupling the computational heat transfer model with the fuzzy PID control model is that the fuzzy PID control model is implemented in the form of user-defined functions through the Fluent program.

Images