Temperature control method and system for reaction kettle for synthesizing isothiocyanate

Abstract

The invention belongs to the field of temperature control, and particularly relates to a reaction kettle temperature control method and system for synthesizing isothiocyanate. The method comprises the following steps: acquiring temperature, material concentration and coolant flow in a kettle in real time, calculating a reaction rate estimated value through a nonlinear mixed model, and calculating an active factor by combining catalyst input time and an index inactivation function; performing unequal weight combination on the three parameters to generate a scheduling variable, and calculating a current value and a second-order difference value of the scheduling variable; selecting a reference state space model according to the current value of the scheduling variable, and correcting the matrix by using a second-order difference value to obtain an LPV model; constructing a quadratic programming optimization problem, wherein an objective function comprises a temperature tracking error penalty term and a control input increment penalty term for adjusting the weight; when the temperature deviation exceeds the threshold value continuously, a control input value penalty term is added; and solving an optimization problem and outputting a first element of the control sequence to a temperature control execution mechanism. According to the invention, the temperature control precision and stability of the isothiocyanate synthesis reaction kettle are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of temperature control, and particularly relates to a reaction kettle temperature control method and system for isothiocyanate synthesis. BACKGROUND

[0002] As a key intermediate in the field of fine chemicals, isothiocyanate is synthesized in a closed reaction kettle, and has strong exothermic, nonlinear, and time-varying characteristics, that is, the reaction rate decreases with the decrease of material concentration, the catalyst presents exponential deactivation with running time, and the thermal inertia of the reaction kettle causes temperature response hysteresis. These characteristics make temperature a core process parameter directly affecting product selectivity, yield, and production safety: too low temperature will lead to incomplete reaction and increase of by-products, and too high temperature may cause local overheating, material overflow, and other safety risks, so high-precision and high-stability closed-loop control of the reaction kettle temperature is required.

[0003] The current industry mainstream adopts the traditional proportional-integral-derivative (PID) control strategy to realize temperature regulation, but its limitations cannot meet the stringent requirements of isothiocyanate high-quality synthesis, because the PID controller is a linear controller, and its control parameters remain unchanged after being set. This fixed-parameter linear control method often shows low control accuracy, slow response speed, easy overshoot and oscillation, and other problems when dealing with complex reaction processes with nonlinear behavior, ultimately affecting the product quality stability.

[0004] In order to overcome the shortcomings of traditional PID control, the industry has proposed advanced control strategies such as model predictive control (MPC). MPC predicts future outputs by establishing a system model, and generates control sequences combined with optimization algorithms, which is more suitable for complex industrial processes in theory. However, traditional MPC mostly uses a single linear model, which can only maintain good control effect near a specific working point when the model is established; when the reaction condition deviates from this working point, the deviation between the model and the actual system will significantly increase, which not only cannot realize precise temperature control, but also may cause further deterioration of the working condition due to control decision errors.

[0005] Linear parameter varying (LPV) model provides a solution to the above problems. It can approximate the nonlinear behavior of the system through a set of linear models that dynamically change with scheduling variables, and theoretically covers all operating conditions of isothiocyanate synthesis. However, in practical applications, LPV model still faces three major technical difficulties: first, the scheduling variable design is difficult, and the core parameters that can comprehensively reflect the reaction rate, catalyst activity, temperature and other key states need to be selected. If the variable selection is not appropriate, the system dynamic characteristics cannot be accurately represented; second, the model update response is slow. When the reaction condition changes suddenly, the existing LPV model is difficult to quickly correct the matrix parameters, resulting in control lag; third, the control weight balance is difficult. The control law weight needs to be adjusted dynamically according to the operating condition. In stable conditions, the response speed needs to be guaranteed first, and in severe conditions, the control stability needs to be guaranteed first. The traditional fixed weight strategy cannot balance both, which is easy to cause overshoot or slow response problems.

[0006] In summary, the existing temperature control technology cannot adapt to the complex characteristics of isothiocyanate synthesis reaction, and a temperature control scheme that can accurately represent the nonlinear system, quickly respond to condition changes, and dynamically balance the control performance is needed to meet the high requirements of industrial production on temperature control accuracy, stability and safety. SUMMARY

[0007] Therefore, the purpose of the present application is to provide a reaction kettle temperature control method and system for isothiocyanate synthesis to solve the technical problems of low temperature control accuracy, slow response and easy overshoot of the existing temperature control method, which cannot meet the high-quality synthesis requirements.

[0008] To solve the above problems, the technical scheme of a reaction kettle temperature control method for isothiocyanate synthesis provided by the present application is as follows: A reaction kettle temperature control method for isothiocyanate synthesis, comprising: Obtaining real-time operating parameters of the reaction kettle, including the temperature in the kettle, the concentration of the material and the flow rate of the coolant, and based on the real-time operating parameters, calculating the reaction rate estimate value using a preset nonlinear hybrid model, and calculating the catalyst activity factor according to the catalyst input time and the exponential deactivation function; Non-equal weight linear combination of the temperature in the kettle, the reaction rate estimate value and the catalyst activity factor is performed to form a one-dimensional scheduling variable, and the current value of the scheduling variable and its second-order difference value in the past preset time window are calculated; based on the current value of the scheduling variable, a reference state space model is selected from a preset local model library, and the second-order difference value is used to nonlinearly correct the matrix of the reference state space model to obtain the linear parameter varying model at the current time; constructing a quadratic programming optimization problem based on the linear parameter varying model, a target function of the quadratic programming optimization problem including a penalty term for temperature tracking error in a prediction horizon and a penalty term for control input increment, a weight coefficient of the penalty term for control input increment being adjusted according to an absolute value of the second-order difference value via a preset gain function; when a deviation of the reactor temperature from a set value continuously exceeds a preset threshold value for a predetermined number of times, adding a penalty term for a control input value in the target function, solving the quadratic programming optimization problem to obtain an optimal control sequence, and outputting a first element of the optimal control sequence as a control instruction to a temperature control actuator.

[0009] Further, the nonlinear hybrid model is composed of a mechanism part and a neural network compensation part, the mechanism part calculating a basic reaction rate using an Arrhenius equation, and inputting the reactor temperature, material concentration and coolant flow to the neural network compensation part to obtain a compensation value, and summing the compensation value and the basic reaction rate as the reaction rate estimation value.

[0010] Further, a calculation formula of the catalyst activity factor is: ; in the formula, is an initial activity of the catalyst, is a deactivation constant, is a running time after the catalyst is put in.

[0011] Further, a calculation formula of the one-dimensional scheduling variable is: ; in the formula, is the scheduling variable at a current time, is a normalized reactor temperature, is a weight of the normalized reactor temperature; is a normalized reaction rate estimation value, is a weight of the normalized reaction rate estimation value; is a normalized catalyst activity factor, is a weight of the normalized catalyst activity factor, , , is a preset constant and a sum of the three is 1.

[0012] Further, the weight of the normalized reactor temperature is greater than the weight of the normalized reaction rate estimation value , and the weight of the normalized reaction rate estimation value is greater than the weight of the normalized catalyst activity factor .

[0013] Further, the calculation formula of the second-order difference value of the scheduling variable in the past preset time window is: In the formula, is the second-order difference value of the scheduling variable at the current sampling time, is the scheduling variable at the previous time, is the scheduling variable at the time before the previous time.

[0014] Further, the calculation formula of the weight coefficient of the penalty term of the control input increment is: ; In the formula, is the basic weight value, is the gain adjustment coefficient, is the exponential growth coefficient, is the absolute value of the second-order difference value of the scheduling variable.

[0015] Further, when the deviation of the kettle temperature from the set value continuously exceeds the preset threshold value for a predetermined number of times, a penalty term for the control input value is added in the target function, including: When the absolute value of the deviation of the kettle temperature from the set value is greater than 2℃ for 5 control periods in succession, a penalty term is added in the target function Wherein, is the future control input sequence to be optimized, is a non-zero penalty weight matrix for the control input value.

[0016] Further, the kettle temperature is measured by a thermocouple installed on the reaction kettle, the material concentration is measured by an online analytical instrument installed on the reaction kettle, and the coolant flow is measured by an electromagnetic flowmeter installed on the reaction kettle.

[0017] The technical scheme of the reaction kettle temperature control system for isothiocyanate synthesis provided by the application is: A reaction kettle temperature control system for isothiocyanate synthesis, comprising a processor and a memory, and the memory stores a computer program, when the computer program is run by the processor, the reaction kettle temperature control method for isothiocyanate synthesis in any one of the above technical solutions is realized.

[0018] The beneficial effects of the application are: The present application can obtain the nonlinear characteristics of the reaction process under different working conditions by constructing a composite scheduling variable which fuses the reaction rate estimate value and the catalyst activity factor, and using the current value and the change trend information of the variable to determine the linear parameter variation model. Meanwhile, the adjustment of the control law is linked with the change degree of the scheduling variable, so that the control is more rapid when the operation is stable, and more smooth when the working condition changes sharply, which balances the rapidity and stability of the control, and suppresses the overshoot and oscillation. In the case that the temperature continuously deviates from the set value, the recovery ability in the face of large disturbance is enhanced by the additional punishment of the control input value, and the temperature control precision and stability of the isothiocyanate synthesis reaction kettle are improved. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 The step flow chart of the isothiocyanate synthesis reaction kettle temperature control method of the present application; Figure 2 The structure block diagram of the isothiocyanate synthesis reaction kettle temperature control system of the present application. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application.

[0021] The specific embodiments of the isothiocyanate synthesis reaction kettle temperature control method of the present application are as follows: As shown in the figure, the isothiocyanate synthesis reaction kettle temperature control method comprises the following steps: Figure 1 S1, obtaining the real-time operation parameters of the reaction kettle including the kettle temperature, material concentration and coolant flow, and based on the real-time operation parameters, calculating the reaction rate estimate value by using a preset nonlinear hybrid model, and calculating the catalyst activity factor according to the catalyst input time and the exponential deactivation function. Specifically, in this step, the kettle temperature measured by a distributed control DCS, the material concentration measured by an online analyzer and the coolant flow measured by an electromagnetic flowmeter are collected. The nonlinear hybrid model is composed of a mechanism part and a neural network compensation part. The mechanism part adopts the Arrhenius equation

[0022] to calculate the basic reaction rate, wherein, is the reaction rate constant under ideal conditions, is the energy of the reaction, i.e. the activation energy, is the ideal gas constant, ​The absolute temperature is used. The neural network compensation section uses the in-vessel temperature, material concentration, and coolant flow rate as inputs to correct model errors. In one embodiment, the neural network compensation section is located after the mechanism section. The neural network compensation section is preferably an MLP network. The in-vessel temperature, material concentration, and coolant flow rate are input into the MLP network to obtain compensation values. The sum of the compensation values ​​and the basic reaction rate is used as the reaction rate estimate.

[0023] In this embodiment, the formula for calculating the catalyst activity factor is: ;In the formula, This represents the initial activity of the catalyst. The deactivation constant is determined experimentally beforehand. This refers to the operating time after the catalyst is added. This represents the catalyst activity factor at the current sampling time.

[0024] S2, the estimated values ​​of the reactor temperature and reaction rate are linearly combined with the catalyst activity factor with non-equal weights to form a one-dimensional scheduling variable, and the current value of the scheduling variable and its second-order difference value within a preset time window are calculated; based on the current value of the scheduling variable, a reference state-space model is selected from a preset local model library, and the matrix of the reference state-space model is nonlinearly corrected using the second-order difference value to obtain the linear parameter change model at the current time.

[0025] In this step, a one-dimensional scheduling variable integrates multiple key factors affecting the polymerization reaction process into a single index, representing the current overall operating status. The calculation formula is as follows: ; In the formula, The normalized temperature inside the vessel. The weighting of the normalized in-bottle temperature; This is the normalized estimate of the reaction rate. The weights of the normalized reaction rate estimates; The normalized catalyst activity factor, The weights of the normalized catalyst activity factors, , , The three are preset positive constants, and their sum is 1. In one specific embodiment, Take 0.6, Take 0.3, Take 0.1.

[0026] Based on the current sampling time The moment before And the moment before Given the values ​​of the scheduling variables, calculate the second-order difference of the scheduling variables at the current time using the following formula: .

[0027] In one embodiment, the local model library includes local model A (reaction initiation / low temperature model), local model B (violent reaction / high temperature and high activity model), and local model C (reaction end / catalyst deactivation model). Based on historical operating data, the entire operating range of the scheduling variables is pre-divided into multiple sub-intervals, and a linear state-space model is identified near the center point of each sub-interval. The state-space model together constitutes the local model library. The reactor exhibits significant thermal inertia; the influence of control inputs, such as coolant flow rate, on the output reactor temperature is not instantaneous but rather achieved through a dynamic process that alters the internal state of the system, making it a strictly true system. During control operation, the current value of the scheduling variable is used... Based on the location, a reference state-space model is selected from the local model library using the nearest neighbor principle or interpolation method. }

[0028] The formula for nonlinear correction of a matrix is: ;In the formula, and This is a preset correction factor. This is the preset correction direction matrix. The matrix of the selected baseline model is nonlinearly corrected using the second-order difference values ​​of the scheduling variables to obtain the current linear parameter variation model (LPV model) that varies with the parameters. Since the scheduling variables consist of core parameters such as the reaction rate estimate, in-vessel temperature, and catalyst activity factor, their second-order difference values ​​best represent the acceleration or deceleration trend of the reaction process, which alters the system's inherent characteristics and stability. In the baseline state-space model, The matrix defines the system's inherent characteristics and stability, therefore only corrections are needed. The matrix can most effectively obtain the main nonlinear behavior of the system, and the correction mainly acts on... matrix.

[0029] S3, construct a quadratic programming optimization problem based on the linear parameter variation model. The objective function of the quadratic programming optimization problem includes a penalty term for the temperature tracking error in the prediction time domain and a penalty term for the control input increment. The weight coefficient of the penalty term for the control input increment is adjusted according to the absolute value of the second-order difference value via a preset gain function.

[0030] The quadratic programming problem in this step is an adaptive optimization problem that is constructed and solved in each control cycle. Specifically, the prediction time domain is set to Np, and the control time domain is set to Nc. Based on the LPV model at the current time step, the in-vessel temperature output sequence is predicted for the next Np time steps. , At the current moment Predicting the future Output of the internal temperature of the vessel at any given time.

[0031] In this embodiment, the mathematical expression of the objective function is: + ; In the formula, for Temperature setpoint at any time For the current moment The future that is determined The control increment applied at any given time, i.e., the change in coolant flow rate. This is a positive definite weighting matrix for temperature tracking error, used to penalize the deviation between the predicted temperature and the set value. The larger the value, the more the controller tends to make the temperature accurately track the set value. The positive definite weight matrix is ​​used to control the increment of the input and to penalize the change in the control action. Its value is dynamically adjusted according to the second difference value of the scheduling variable. The larger the value, the smoother and more conservative the control action.

[0032] In this step, the formula for calculating the weight coefficient of the penalty term controlling the input increment is: ; In the formula, Basic weight value, This is the gain adjustment coefficient. The exponential growth coefficient, This represents the absolute value of the second-order difference of the scheduling variable. The positive definite weight matrix mentioned above... It is a diagonal matrix in which all elements on the diagonal are equal to the weight coefficients of the calculated penalty term for the control input increment. .

[0033] Specifically, the second-order difference of the scheduling variable can be understood as the acceleration of the change in operating state. A larger absolute value of the second-order difference indicates that the process is undergoing a drastic and rapid transition, such as when the reaction enters a highly exothermic phase. In this case, the controller's actions need to be more gradual and conservative. Through the exponential gain function, even a small increase in the absolute value of the second-order difference of the scheduling variable can lead to a change in the weighting coefficient. Increase the amount of input, thereby increasing the penalty for changes in control input and making the control action more gentle.

[0034] For example, setting a base weight value =0.1, gain adjustment coefficient =5, exponential growth coefficient =50. During the steady-state phase of the process, the values ​​of the scheduling variables for three consecutive control cycles may be 0.550, 0.551, and 0.552. The calculated second-order difference value of the scheduling variables is 0, at which point the weighting coefficient... With a base weight value of 0.1, the controller responds relatively sensitively. However, if the process suddenly accelerates, and the scheduling variable values ​​become 0.55, 0.56, and 0.59, then their first-order differences are 0.01 and 0.03 respectively, and their second-order difference value is 0.02. At this point, the weighting coefficients... The weighting factor is approximately 8.69. Increasing the weighting factor from 0.1 to 8.69 will suppress drastic changes in the opening of the cooling water valve and ensure stability during the change process.

[0035] S4, when the deviation between the temperature inside the vessel and the set value exceeds the preset threshold for a predetermined number of consecutive times, a penalty term for the control input value is added to the objective function; the optimal control sequence is obtained by solving the quadratic programming optimization problem, and the first element of the optimal control sequence is output as a control command to the temperature control actuator.

[0036] In an optional embodiment, when the deviation between the temperature inside the vessel and the set value continuously exceeds a preset threshold for a predetermined number of times, a penalty term for the control input value is added to the objective function, including: When the absolute value of the deviation between the internal temperature and the set value is greater than 2℃ for five consecutive control cycles, a penalty term is added to the objective function J. ,in, The future control input sequence to be optimized. This is the weight matrix for non-zero penalties on control input values.

[0037] In polymerization reaction control, if a strong exothermic disturbance occurs, the reactor temperature may remain higher than the setpoint even when the cooling water valve is fully open. In this case, conventional model-predicted control objective functions primarily penalize changes in the control variable, causing the controller to continue requesting a control output that can no longer be increased, resulting in integral saturation. This problem can be mitigated by penalizing the control input value itself. For example, the setpoint temperature is 350℃, and the control cycle is 10s. Over five consecutive control cycles (50s), the detected reactor temperatures are 352.2℃, 352.5℃, 352.8℃, 352.6℃, and 352.4℃, respectively, with absolute deviations from the setpoint all greater than 2℃. At this point, an emergency mechanism is triggered. When optimizing its objective function J, the controller adds an additional penalty term, which is expressed through a non-zero weight matrix. For example, a diagonal matrix with diagonal elements of 0.5 penalizes the magnitude of future control input sequences. This forces the optimizer to seek a control input solution with a smaller magnitude while maintaining temperature tracking, that is, slightly reducing the cooling water valve opening from 100%, which helps the controller escape saturation and regain temperature control once the disturbance weakens.

[0038] The constructed quadratic programming problem and its constraints are input into an online quadratic programming solver, which can be used to solve the problem using methods such as the interior point method or the effective set method, to obtain the optimal control increment sequence in the future control time domain. , Based on the rolling optimization principle of model predictive control, only the first element of the optimal control increment sequence is taken. and compare it with the control quantity from the previous moment. Add them together to get the actual control quantity at the current moment. The actual control quantity The signal is converted into a standard current signal and sent to the electro-proportional valve that regulates the coolant flow, thereby achieving closed-loop control of the reactor temperature.

[0039] This invention first constructs a non-equal-weighted scheduling variable that integrates the in-vessel temperature, reaction rate estimate, and catalyst activity factor, and then combines this with a second-order difference value to correct the baseline state-space model to obtain an LPV model. This solves the problem of poor adaptability of the traditional MPC single linear model and overcomes the difficulties of inaccurate scheduling variable design and slow model update response in existing LPV models. It can accurately characterize the nonlinear and time-varying characteristics of the reaction under all operating conditions, laying the model foundation for high-precision temperature control. Second, in the quadratic programming optimization problem, this invention dynamically adjusts the weight of the control input increment penalty term according to the second-order difference value of the scheduling variable through a preset gain function. This achieves rapid response under stable operating conditions and stable control under severe operating conditions, effectively balancing control speed and stability, and avoiding the overshoot or lag problems easily caused by traditional fixed-weight strategies. Finally, for situations where the temperature continuously deviates from the setpoint, this invention adds a control input penalty term to the objective function, which can effectively alleviate integral saturation, enhance the system's recovery ability under strong disturbances, and further ensure temperature control stability. The method of this invention significantly improves the temperature control accuracy and stability of the isothiocyanate synthesis reactor, better meeting the stringent temperature parameter requirements of high-quality synthesis processes, while reducing the problems of increased by-products and safety risks caused by temperature fluctuations.

[0040] Specific embodiments of the temperature control system for the reactor used in isothiocyanate synthesis provided by the present invention: like Figure 2As shown, the temperature control system for the reactor used in isothiocyanate synthesis includes a processor and a memory. The memory stores a computer program, which, when executed by the processor, implements the temperature control method for the reactor used in isothiocyanate synthesis in the above embodiments.

[0041] The temperature control system for the reactor used in isothiocyanate synthesis also includes other components well known to those skilled in the art, such as communication buses and communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

[0042] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented by computer-readable / executable instructions stored or otherwise maintained on such a computer-readable medium.

[0043] While various embodiments of the invention have been shown and described in this specification, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention.

Claims

1. A method for temperature control in a reactor used for isothiocyanate synthesis, characterized in that, include: The real-time operating parameters of the reactor, including the internal temperature, material concentration, and coolant flow rate, are obtained. Based on the real-time operating parameters, the reaction rate estimate is calculated using a preset nonlinear mixing model. At the same time, the catalyst activity factor is calculated based on the catalyst addition time and the exponential deactivation function. The estimated values ​​of the in-vessel temperature and reaction rate are combined linearly with the catalyst activity factor in a non-equal weighted manner to form a one-dimensional scheduling variable, and the current value of the scheduling variable and its second-order difference value within a preset time window are calculated. Based on the current value of the scheduling variable, a reference state space model is selected from a preset local model library, and then the matrix of the reference state space model is nonlinearly corrected using the second-order difference value to obtain the linear parameter change model at the current time. A quadratic programming optimization problem is constructed based on the linear parameter variation model. The objective function of the quadratic programming optimization problem includes a penalty term for the temperature tracking error in the prediction time domain and a penalty term for the control input increment. The weight coefficient of the penalty term for the control input increment is adjusted according to the absolute value of the second-order difference value via a preset gain function. When the deviation between the temperature inside the vessel and the set value exceeds the preset threshold for a predetermined number of consecutive times, a penalty term for the control input value is added to the objective function; the optimal control sequence is obtained by solving the quadratic programming optimization problem, and the first element of the optimal control sequence is output as a control command to the temperature control actuator.

2. The method for temperature control of a reactor for isothiocyanate synthesis according to claim 1, characterized in that, The nonlinear hybrid model consists of a mechanistic part and a neural network compensation part. The mechanistic part uses the Arrhenius equation to calculate the basic reaction rate. The reactor temperature, material concentration and coolant flow rate are input into the neural network compensation part to obtain the compensation value. The sum of the compensation value and the basic reaction rate is used as the estimated reaction rate value.

3. The method for temperature control of a reactor for isothiocyanate synthesis according to claim 1, characterized in that, The formula for calculating the catalyst activity factor is as follows: ;In the formula, This represents the initial activity of the catalyst. The deactivation constant is This refers to the operating time after the catalyst is added.

4. The method for temperature control of a reactor for isothiocyanate synthesis according to claim 3, characterized in that, The formula for calculating the one-dimensional scheduling variable is as follows: ； In the formula, The scheduling variable at the current moment. The normalized temperature inside the vessel. The weighting of the normalized in-bottle temperature; This is the normalized estimate of the reaction rate. The weights for the normalized reaction rate estimates; The normalized catalyst activity factor, The weights of the normalized catalyst activity factors, , , The three numbers are preset positive constants and their sum is 1.

5. The method for temperature control of a reactor for isothiocyanate synthesis according to claim 4, characterized in that, Weight of normalized in-bottle temperature Weights greater than the normalized reaction rate estimate And the weights of the normalized reaction rate estimates The weight of the catalyst activity factor greater than the normalized value .

6. The method for temperature control of a reactor for isothiocyanate synthesis according to claim 4, characterized in that, The formula for calculating the second-order difference value of the scheduling variable within the past preset time window is as follows: ;In the formula, The second difference value of the scheduling variable at the current sampling time. The scheduling variable is from the previous time step. This refers to the scheduling variable from the previous time step.

7. The method for temperature control of a reactor for isothiocyanate synthesis according to claim 6, characterized in that, The formula for calculating the weight coefficient of the penalty term for the control input increment is as follows: ； In the formula, Basic weight value, This is the gain adjustment coefficient. The exponential growth coefficient, It is the absolute value of the second difference of the scheduling variable.

8. The method for temperature control of a reactor for isothiocyanate synthesis according to claim 1, characterized in that, When the deviation between the temperature inside the vessel and the set value exceeds a preset threshold a predetermined number of times, a penalty term for the control input value is added to the objective function, including: When the absolute value of the deviation between the internal temperature and the set value exceeds 2℃ for five consecutive control cycles, a penalty term is added to the objective function. ,in, The future control input sequence to be optimized. This is the weight matrix for non-zero penalties on control input values.

9. The method for temperature control of a reactor for isothiocyanate synthesis according to claim 1, characterized in that, The temperature inside the reactor is measured by a primary thermocouple installed on the reactor, the material concentration is measured by an online analytical instrument installed on the reactor, and the coolant flow rate is measured by an electromagnetic flow meter installed on the reactor.

10. A temperature control system for a reactor used in the synthesis of isothiocyanates, characterized in that, It includes a processor and a memory, the memory storing a computer program that, when run by the processor, implements a method for temperature control of a reactor for isothiocyanate synthesis as described in any one of claims 1-9.

Images