Temperature prediction control method for oxidation kettle

Abstract

The invention provides an oxidation kettle temperature prediction control method. The method comprises the following steps: S1, establishing a differential equation model; s2, converting the differential equation model into a discretized space prediction model, enabling the space prediction model to be equivalent to a first-order inertia plus lag system, determining state transition matrix parameters through system identification, and generating an oxidation kettle temperature prediction model; s3, obtaining a target function of a model prediction controller according to an error between the predicted temperature and the reference temperature; s4, improving an artificial bee colony algorithm by adopting a variation strategy of a differential evolution algorithm; s5, in the rolling optimization process of model prediction control, the improved artificial bee colony algorithm is adopted to update the optimal control input sequence in a rolling mode until the target function is minimized, and the optimal control input sequence is obtained; and S6, applying the optimal control input sequence to an oxidation kettle control system in real time, and dynamically adjusting the temperature of the oxidation kettle. The selection of the control input sequence is optimized, and the temperature control in the reaction process is more accurate and efficient.

Description

Technical Field

[0001] This application relates to the field of automatic control technology, and particularly to a method for predicting and controlling the temperature of an oxidation kettle. Background Art

[0002] An oxidation kettle is a common device in industries such as chemical engineering and pharmaceuticals, used for various chemical reactions and material mixing. However, in the process of controlling the temperature of the oxidation kettle, due to factors such as heat transfer and material reactions, there are often significant time-delay problems. This time-delay not only affects the accuracy and stability of the temperature control in the oxidation kettle, but also may lead to problems such as a decline in product quality and an increase in energy consumption. Summary of the Invention

[0003] This application proposes a method for predicting and controlling the temperature of an oxidation kettle, which improves the accuracy and reliability of model predictive control in the temperature control of the reaction kettle, significantly optimizes the selection of the control input sequence, and ensures that the temperature control of the reaction process is more accurate and efficient.

[0004] This application provides a method for predicting and controlling the temperature of an oxidation kettle, including the following steps:

[0005] S1, collect the historical operating parameters of the oxidation kettle and establish a differential equation model; S2, transform the differential equation model into a discretized space prediction model, equivalent the space prediction model to a first-order inertial plus lag system, determine the state transition matrix parameters through system identification, and generate an oxidation kettle temperature prediction model; S3, obtain the objective function of the model predictive controller according to the error between the predicted temperature and the reference temperature; S4, improve the artificial bee colony algorithm using the mutation strategy of the differential evolution algorithm; S5, in the rolling optimization process of model predictive control, use the improved artificial bee colony algorithm to roll and update the optimal control input sequence until the objective function is minimized to obtain the optimal control input sequence; S6, apply the optimal control input sequence to the oxidation kettle control system in real time to dynamically adjust the temperature of the oxidation kettle.

[0006] Specifically, the operating parameters include at least one of the oxidation kettle temperature T(t), the jacket hot water flow rate F hoting , the jacket hot water temperature T hot , the jacket coolant flow rate F cooling , and the jacket coolant temperature T cool .

[0007] Specifically, the differential equation model includes a thermal dynamic model of the oxidation kettle established based on the principle of energy conservation, and the thermal dynamic model includes:

[0008] The heat balance equation, the formula is as follows:

[0009]

[0010] Among them, m is the mass of the material, C p is the specific heat capacity, Q in , Q out are the input heat flow rate and the output heat flow rate respectively, ΔH r is the reaction heat release amount, and r(t) is the reaction rate;

[0011] The cooling system equation is as follows:

[0012] Q out (t) = U·F cooling ·(T(t) - T cool (t)), or,

[0013] Q in (t) = U·F hoting ·(T hot (t) - T(t)),

[0014] Among them, U is the heat transfer coefficient.

[0015] Specifically, the reaction rate is expressed by the Arrhenius formula equation as follows:

[0016]

[0017] Among them, k is the reaction rate constant, C A (t) is the concentration of the reactant, E a is the activation energy of the reaction, and R is the molar gas constant.

[0018] Specifically, the differential equation model is transformed into a discretized space prediction model, and the formula of the space prediction model is as follows:

[0019] T(t + 1) = A·T(t) + B·u(t),

[0020] Among them, T(t + 1) is the temperature at the prediction time t + 1, T(t) is the temperature at the current time t, u(t) is the control input at the current time, and A and B are the state transition matrix parameters.

[0021] Specifically, the temperature prediction model of the oxidation kettle is generated, and the formula is as follows:

[0022]

[0023] Among them, is the temperature predicted for the next k steps starting from the current time t; u(t + i) is the control input for the next k steps.

[0024] Specifically, the objective function of the model predictive controller is obtained according to the error between the predicted output and the reference trajectory, and the formula is as follows:

[0025]

[0026] , where J is the objective function, T(t + k|t) is the predicted temperature, T ref (t + k) is the reference temperature, u(t + k) is the control input, and λ is the penalty factor for the change of the control input.

[0027] Specifically, in the rolling optimization process of model predictive control, the improved artificial bee colony algorithm is used to roll-update the optimal control input sequence until the objective function is minimized. The specific steps are as follows:

[0028] S81. Take the control input sequence as the search individual of the artificial bee colony algorithm and initialize the search space; S82. Generate a new solution through differential evolution mutation and update the individual position by combining local random perturbation; S83. Calculate the cost function value of each individual, and the cost function includes the temperature tracking error and the control input change penalty term; S84. Iteratively optimize until the convergence condition is satisfied, and output the optimal control input sequence.

[0029] Specifically, the differential evolution mutation strategy for improving the artificial bee colony algorithm includes:

[0030] S91. Perform differential mutation operation on the current population individual x i to generate a mutant vector v i , and the formula is as follows:

[0031] v i = x r1 + F·(x r2 - x r3 )

[0032] where x r1 , x r2 , x r3 are randomly selected distinct individuals, and F is the scaling factor;

[0033] S92. Perform a crossover operation on the mutant vector v i and the original individual x i to generate a new solution u i ;

[0034] S93. Adopt a greedy selection strategy to retain the solution with better fitness and update the population.

[0035] Compared with the prior art, the present application improves the artificial bee colony algorithm by using the mutation strategy of the differential evolution algorithm, which can perform a wider search in the multi-dimensional control input space and overcome the limitations of traditional optimization methods. The differential evolution algorithm enhances the global search ability through differential mutation operations, avoiding the phenomenon that the traditional artificial bee colony algorithm stagnates in local optimal traps, thus making the optimization process more accurate and stable. Especially in the practical application of complex reactor temperature control systems, it can effectively improve the robustness and response speed of the system. The application of this method improves the accuracy and reliability of model predictive control in reactor temperature control, significantly optimizes the selection of control input sequences, and ensures more accurate and efficient temperature control during the reaction process. Brief Description of the Drawings

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0037] Figure 1 is a schematic flowchart of the oxidation reactor temperature prediction control method;

[0038] Figure 2 is a schematic flowchart of optimizing the parameters during the rolling optimization process using the improved artificial bee colony algorithm. Detailed Embodiments

[0039] The embodiments of the present application will be described in detail below with reference to the drawings.

[0040] The following specific examples illustrate the implementation manners of the present application. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The present application can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application.

[0041] It should be noted that the following description relates to various aspects of embodiments within the scope of the appended claims. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on this application, those skilled in the art should understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number and aspects set forth herein can be used to implement an apparatus and / or practice a method. Additionally, this apparatus and / or method can be implemented using other structures and / or functionality in addition to one or more of the aspects set forth herein.

[0042] It should also be noted that the diagrams provided in the following embodiments merely illustrate the basic concept of this application schematically. Only the components related to this application are shown in the diagrams, rather than being drawn according to the number, shape, and size of the components in actual implementation. The type, quantity, and proportion of each component in actual implementation can be arbitrarily changed, and the component layout type may also be more complex.

[0043] In addition, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that these examples can be practiced without these specific details.

[0044] As Figure 1 shown, this embodiment provides a method for predicting and controlling the temperature of an oxidation kettle, including the following steps:

[0045] S1, Collect the historical operating parameters of the oxidation kettle and establish a differential equation model; the differential equation model includes the mathematical model of the cooling system and the mathematical model of the heat balance equation.

[0046] S2, Convert the differential equation model into a discretized space prediction model, equivalent the space prediction model to a first-order inertia plus lag system, determine the state transition matrix parameters through system identification, and generate an oxidation kettle temperature prediction model. In this embodiment, system identification is performed through a unit step, and the historical operating data is brought into the space prediction model to obtain the oxidation kettle temperature prediction model.

[0047] S3, Obtain the objective function of the model predictive controller based on the error between the predicted temperature and the reference temperature. Take improving the prediction accuracy as the goal of the improved artificial bee colony algorithm.

[0048] S4, Improve the artificial bee colony algorithm using the mutation strategy of the differential evolution algorithm.

[0049] S5. During the rolling optimization process of model predictive control, the improved artificial bee colony algorithm is used to roll - update the optimal control input sequence until the objective function is minimized to obtain the optimal control input sequence until a preset termination condition is reached.

[0050] S6. Apply the optimal control input sequence to the oxidation kettle control system in real - time to dynamically adjust the temperature of the oxidation kettle.

[0051] In this embodiment, the mutation strategy of the differential evolution algorithm is used to improve the artificial bee colony algorithm, which can perform a wider search in the multi - dimensional control input space and overcome the limitations of traditional optimization methods. The differential evolution algorithm enhances the global search ability through differential mutation operations, avoiding the phenomenon that the traditional artificial bee colony algorithm stagnates in local optimal traps, thus making the optimization process more accurate and stable. Especially in the practical application of complex reactor temperature control systems, it can effectively improve the robustness and response speed of the system. The application of this method improves the accuracy and reliability of model predictive control in reactor temperature control, significantly optimizes the selection of the control input sequence, and ensures more accurate and efficient temperature control during the reaction process.

[0052] The operating parameters include at least one of the oxidation kettle temperature T(t), the jacket hot water flow rate F hoting , the jacket hot water temperature T hot , the jacket coolant flow rate F cooling and the jacket coolant temperature T cool . And denoising and normalization pre - processing are performed on the operating parameters.

[0053] The temperature control of the oxidation kettle involves multiple factors such as energy balance, reaction heat, cooling system, and feed flow rate. Therefore, the accuracy of the model directly affects the control performance.

[0054] The temperature change in the oxidation kettle generally follows the balance of heat input and output. To establish the model, we can first start from the perspective of energy conservation and construct the thermal dynamic equation of the oxidation kettle.

[0055] Assume that the heat in the oxidation kettle is only input through the hot water flow rate and output through the cooling water flow rate, and the reaction heat can be handled through a simple assumption. The heat balance equation is usually expressed as:

[0056]

[0057] The chemical reaction in the oxidation kettle is usually exothermic, so the heat release amount needs to be described according to the reaction rate model.

[0058] Assume that the reaction heat Q reaction is proportional to the reaction rate r(t) and can be expressed as:

[0059] Qreaction q(t) = ΔH·r(t);

[0060] Where ΔH is the heat release of the reaction, that is, the energy released per mole of reactant during the reaction. r(t) is the reaction rate, which is usually related to the temperature and concentration of the reactants.

[0061] The reaction rate is expressed by the Arrhenius equation, and the formula is as follows:

[0062]

[0063] Where k is the reaction rate constant, C A (t) is the concentration of the reactant, E a is the activation energy of the reaction, and R is the molar gas constant.

[0064] In summary, the heat balance equation is as follows:

[0065]

[0066] Where m is the mass of the material, C p is the specific heat capacity, Q in , Q out are the input heat flow rate and output heat flow rate respectively, ΔH r is the reaction heat release, and r(t) is the reaction rate;

[0067] The heat output of the jacket cooling system is usually proportional to the cooling flow rate and the cooling temperature difference. Assuming that the heat exchange capacity of the cooling system is U, the cooling power, that is, the cooling system equation can be expressed as:

[0068] Q out (t) = U·F cooling ·(T(t) - T cool (t)), or,

[0069] Q in (t) = U·F hoting ·(T hot (t) - T(t)),

[0070] Where U is the heat exchange coefficient.

[0071] Combining the above energy balance equation with the relationship between the reaction heat and the cooling system, we can obtain the dynamic model of the oxidation kettle temperature.

[0072] In one embodiment, for temperature prediction, MPC (Model Predictive Control) usually adopts a discretized state space model. Assuming at time t,

[0073] Convert the differential equation model into a discretized spatial prediction model. The prediction model of temperature can be expressed in the following form:

[0074] T(t + 1) = A·T(t) + B·u(t),

[0075] where T(t + 1) is the temperature at the prediction time t + 1, T(t) is the temperature at the current time t, u(t) is the control input at the current time, and A and B are state transition matrix parameters, which reflect the relationship between temperature changes and the current temperature and control input. Usually, we need to perform system identification on A and B and fit these parameters through experimental data.

[0076] Based on the above thermal dynamics, a prediction model of the oxidation kettle can be established. Assuming that the control time domain is N, at each time t, we need to predict the temperature changes in the next N steps. Through the optimization objective of MPC, we embed the prediction model into the optimization problem to calculate the control input. The temperature prediction in the next N steps can be expressed as follows:

[0077]

[0078] where is the temperature predicted k steps into the future starting from the current time t; u(t + i) is the control input in the next k steps.

[0079] In model predictive control, we need to optimize the control input sequence in each control time domain to achieve a certain goal, such as temperature control.

[0080] The optimization problem of MPC is generally expressed as minimizing the objective function:

[0081] The objective function of the model predictive controller is obtained based on the error between the predicted output and the reference trajectory. The formula is as follows:

[0082]

[0083] , where J is the objective function, T(t + k|t) is the predicted temperature, T ref (t + k) is the reference temperature, u(t + k) is the control input, and λ is the penalty factor for the change of the control input.

[0084] The goal is to minimize the above cost function by adjusting the control input u(t + k), while considering the constraint conditions.

[0085] As Figure 2 shown, the steps for the improved artificial bee colony algorithm to select the optimal control input are as follows:

[0086] The algorithm starts.

[0087] Initialize a population at the beginning.

[0088] Calculate the fitness value of each individual.

[0089] The employed bees search for new food sources near the food sources according to the information: According to the current information, the employed bees search for new food sources near the food sources.

[0090] Calculate the fitness value and perform greedy selection to generate candidate solutions: Calculate the new fitness value and perform greedy selection to generate candidate solutions.

[0091] The onlooker bees search for food sources according to the algorithm and the information of the employed bees: The onlooker bees search for food sources according to the algorithm and the information of the employed bees.

[0092] Judge whether the onlooker bees find good food sources. If so, record the positions of the good food sources. If not, continue to the next step.

[0093] Perform crossover and mutation on the bee colonies at poor food sources.

[0094] Adjust the scaling factor according to the solution search process or the actual situation.

[0095] Generate a new onlooker bee colony.

[0096] Judge whether scout bees are generated. If so, generate new food sources to replace the original food sources. If not, continue to the next step.

[0097] Judge whether the end condition is reached. If so, output the optimal solution and the algorithm ends. If not, return to the step of calculating the fitness value and continue the iteration.

[0098] In this embodiment, an improved artificial bee colony algorithm is introduced to select the optimal control input. The specific steps are as follows:

[0099] It is necessary to initialize the search space of the artificial bee colony. In this embodiment, the search space is the parameter space of the MPC optimal control input sequence. Assuming that the control input sequence is u(t), u(t + 1),.....u(t + N - 1), we can regard these control inputs as the parameters of the bee individuals. Each individual represents a possible control input sequence.

[0100] For each individual (control input sequence), calculate its corresponding cost function. During the calculation process, a model is needed to predict the temperature change, and the future temperature values are predicted according to the control input sequence, and then compared with the reference temperature to obtain the cost function.

[0101] Each bee (control input sequence) will adjust its search position according to its neighbors and the current optimal solution. Specifically, the bees will perform local search and global search according to the following rules:

[0102] For each individual bee, a new solution is generated by making a small random perturbation near its current solution.

[0103] If the objective function value of the new solution is better, the new solution is accepted as the current solution. Assume the current solution is u i , and a new solution u′ is generated near it i .

[0104] u′ i = u i + φ · (u i - u best ),

[0105] where φ is a random perturbation factor and u best is the optimal solution of the current population.

[0106] During the local search process, the mutation strategy of the differential evolution algorithm is borrowed to enhance the search ability of individuals and avoid falling into local optima.

[0107] Perform differential mutation operation on the current population individual x i to generate a mutant vector v i , and the formula is as follows:

[0108] v i = x r1 + F · (x r2 - x r3 ),

[0109] where x r1 , x r2 , x r3 are randomly selected distinct individuals and F is a scaling factor;

[0110] S92. Perform crossover operation on the mutant vector υ i and the original individual x i to generate a new solution u i ;

[0111] Generate a new solution v i according to the above differential evolution formula, and then compare it with the current solution u i .

[0112] Compare the objective function values of the two and select the one with higher fitness as the optimal solution.

[0113] Repeat the above steps until the maximum number of iterations is reached or the objective function value converges.

[0114] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A temperature prediction and control method for an oxidation kettle, characterized in that, It includes the following steps: S1. Collect the historical operating parameters of the oxidation kettle and establish a differential equation model; S2. Convert the differential equation model into a discretized spatial prediction model, equivalent the spatial prediction model to a first-order inertia plus lag system, determine the state transition matrix parameters through system identification, and generate an oxidation kettle temperature prediction model; S3. Obtain the objective function of the model predictive controller based on the error between the predicted temperature and the reference temperature; S4. Improve the artificial bee colony algorithm using the mutation strategy of the differential evolution algorithm; S5. In the rolling optimization process of model predictive control, use the improved artificial bee colony algorithm to roll update the optimal control input sequence until the objective function is minimized to obtain the optimal control input sequence; S6. Apply the optimal control input sequence to the oxidation kettle control system in real time to dynamically adjust the temperature of the oxidation kettle.

2. The oxidation kettle temperature prediction control method according to claim 1, characterized in that The operating parameters include: the temperature T(t) of the oxidation kettle, the flow rate F of the jacket hot water hoting , the temperature T of the jacket hot water hot , the flow rate F of the jacket coolant cooling and the temperature T of the jacket coolant cool at least one of them.

3. The temperature prediction and control method of the oxidation kettle according to claim 2, wherein The differential equation model includes a thermal dynamic model of the oxidation kettle established based on the principle of energy conservation, and the thermal dynamic model includes: The heat balance equation, the formula is as follows: Where m is the material mass, C p is the specific heat capacity, Q in , Q out are the input heat flow and output heat flow, ΔH r is the heat release of the reaction, r(t) is the reaction rate; The cooling system equation, the formula is as follows: Q out Q(t) = U·F cooling ·(T(t) - T cool (t)), or, Q in Q(t) = U·F hoting ·(T hot (t) - T(t)), Where U is the heat transfer coefficient.

4. The method for predicting and controlling the temperature of the oxidation kettle according to claim 3, characterized in that, The reaction rate is expressed by the Arrhenius formula equation, the formula is as follows: where k is the reaction rate constant, C A (t) is the concentration of the reactant, E a is the activation energy of the reaction, and R is the molar gas constant.

5. The oxidation kettle temperature prediction and control method according to claim 3, characterized in that, Convert the differential equation model into a discretized spatial prediction model, and the formula of the spatial prediction model is as follows: T(t + 1) = A·T(t) + B·u(t), Where T(t + 1) is the temperature at the prediction time t + 1, T(t) is the temperature at the current time t, u(t) is the control input at the current time, and A and B are the state transition matrix parameters.

6. The oxidation kettle temperature prediction control method according to claim 5, wherein The generated oxidation kettle temperature prediction model, the formula is as follows: where, is the predicted temperature k steps into the future starting from the current time t; u(t + i) is the control input for the next k steps.

7. The temperature prediction and control method for the oxidation kettle according to claim 1, characterized in that Obtain the objective function of the model predictive controller based on the error between the predicted output and the reference trajectory, the formula is as follows: where J is the objective function, T(t + k|t) is the predicted temperature, T ref (t + k) is the reference temperature, u(t + k) is the control input, and λ is the penalty factor for the change in the control input.

8. The oxidation kettle temperature prediction control method according to claim 1, characterized in that In the rolling optimization process of model predictive control, use the improved artificial bee colony algorithm to roll update the optimal control input sequence until the objective function is minimized to obtain the specific steps of the optimal control input including: S81. Take the control input sequence as the search individual of the artificial bee colony algorithm and initialize the search space; S82. Generate new solutions through differential evolution mutation and update the individual positions by combining local random perturbations; S83. Calculate the cost function value of each individual, and the cost function includes the temperature tracking error and the control input change penalty term; S84. Iteratively optimize until the convergence condition is met, and output the optimal control input sequence.

9. The temperature prediction control method of the oxidation kettle according to claim 8, characterized in that, The differential evolution mutation strategy for improving the artificial bee colony algorithm includes: S91, for the individual x in the current population i Perform differential mutation operation to generate a mutant vector v i , and the formula is as follows: v i = x r1 + F·(x r2 - x r3 ) where x r1 , x r2 , x r3 are randomly selected distinct individuals, and F is a scaling factor; S92, for the mutant vector v i and the original individual x i perform a crossover operation to generate a new solution u i ; S93. Use the greedy selection strategy to retain the solution with better fitness and update the population.

Images