Steady-state target tracking optimization and control method for mixed-integral controlled variables

Abstract

The present application relates to a steady-state target tracking optimization and control method for mixed integral type controlled variables, comprising steady-state optimization layer optimization problem decision variable and other variable construction; stability type controlled variable feasibility determination and soft constraint adjustment; integral type controlled variable feasibility determination and soft constraint adjustment; steady-state integral slope balance constraint condition construction; steady-state target tracking optimization and control. The present application is based on the form of the sum of multiple step manipulated variables to construct new decision variables, and introduces them into the feasibility determination and soft constraint adjustment stage of stability type controlled variables and integral type controlled variables, and the objective function of steady-state target tracking optimization, and constructs the constraint conditions of steady-state target tracking optimization according to the relaxed constraint factor and steady-state integral slope balance.

Description

Technical Field

[0001] This invention relates to the field of process control, and in particular to a steady-state target tracking optimization for mixed integral controlled variables. Background Technology

[0002] Double-layer Model Predictive Control (DLMPC) is an advanced process control method. The upper layer of DLMPC performs the calculation of local steady-state objectives and is called the steady-state optimization layer; the lower layer performs the task of tracking the steady-state objective and is called the dynamic control layer. The steady-state optimization layer mainly includes feasibility assessment and soft constraint adjustment, as well as steady-state objective calculation.

[0003] Integral-type controlled variables are typical examples of non-self-balancing processes. Once the system reaches steady state, the integral-type controlled variable continues to rise or fall with a step change in a manipulated variable. Feasibility assessments, soft constraint adjustments, and steady-state target calculations for stable controlled variables are not applicable to integral-type controlled variables.

[0004] For polymerization processes, it is difficult to simultaneously characterize the steady-state predicted values ​​of both stable and integral controlled variables at the steady-state optimization level. Furthermore, when the steady-state optimization problem is infeasible, a priority strategy for relaxing constraints on both stable and integral controlled variables is not considered. Dynamic control tracking of non-optimal steady-state objectives will increase energy consumption costs, among other issues. Summary of the Invention

[0005] The purpose of this invention is to propose an improved steady-state optimization layer for predictive control of complex systems with mixed integral controlled variables in a two-layer structure model, including feasibility determination of mixed integral controlled variables, soft constraint adjustment strategy, and steady-state target calculation.

[0006] The technical solution adopted by this invention to achieve the above objectives is: a steady-state target tracking optimization and control method for mixed integral controlled variables, comprising the following steps:

[0007] 1) Based on the polymerization process control system, the decision variables, manipulated variables, stable controlled variables and integral controlled variables of the steady-state optimization layer optimization problem are constructed; among them, the manipulated variables are used to represent the cooling water flow rate and feed flow rate of the polymerization process, the stable controlled variables are used to represent the reactant concentration and mass, and the integral controlled variables are used to represent the liquid level;

[0008] 2) Based on the decision variables, manipulated variables, and integral controlled variables, construct a feasibility determination and soft constraint adjustment optimization problem for integral controlled variables, and obtain the integral relaxation factor. and

[0009] 3) Based on the decision variable, manipulated variable, stable controlled variable, integral controlled variable, and integral relaxation factor. and The feasibility determination and soft constraint adjustment optimization problem of constructing stable controlled variables is presented, and the stable relaxation factor is obtained. and

[0010] 4) Based on the decision variables and integral-type controlled variables, construct the constraints for the steady-state integral slope equilibrium;

[0011] 5) Based on the decision variable, manipulated variable, stable controlled variable, and integral controlled variable, the integral relaxation factor... and Stable relaxation factor and In addition to the constraint of steady-state integral slope balance, an objective function is constructed, and the local steady-state optimal objective manipulated variable is obtained by solving it. and controlled variable

[0012] 6) Manipulate the optimal target variable and controlled variable As the tracking target of the dynamic control layer, the cooling water flow rate and feed flow rate are adjusted according to the optimal solution of the dynamic control layer.

[0013] Step 1) is as follows:

[0014] 1) For a polymerization reaction process control system with m manipulated variables, the decision variable Δu in the steady-state optimization problem is... ab (k) is a 2m×1 matrix:

[0015] Δu ab (k)

[0016] =[Δu 1,a (k)Δu 1,b (k)Δu 2,a (k)Δu 2,b (k)…Δu m,a (k)Δu m,b (k)] T

[0017] Where Δu i,a (k) and Δu i,b (k) are shown below:

[0018] Δu i,a (k)=MΔu i,ss (k)+(M-1)Δu i,ss (k+1)+…+Δu i,ss (k+M-1)

[0019] Δu i,b (k)=(M-1)Δu i,ss (k)+(M-2)Δu i,ss (k+1)+…+Δu i,ss (k+M-2)

[0020] Where i = 1, 2, ..., m, M is the control time domain of the dynamic control layer, k represents the time step; u represents the manipulated variable, Δu represents the increment of the manipulated variable, and the subscript ss represents the steady state;

[0021] 2) Based on the decision variable Δu ab (k) construct the future M-step manipulated variables, stable controlled variables, and integral controlled variables;

[0022] Construct the difference coefficient matrix C ab,diff As shown below:

[0023]

[0024] Constructing the manipulation variable u for the future M steps ss (k+M-1) is shown below:

[0025] u ss (k+M-1)=u ss (k-1)+C ab,diff Δu ab (k)

[0026] Based on the stable steady-state gain coefficient matrix Constructing stable controlled variables for the next M steps As shown below:

[0027]

[0028] Construct the integral-type controlled variable difference coefficient matrix S ab,N As shown below:

[0029]

[0030] Where n is the number of integral controlled variables, S ab,N It is the step response difference coefficient matrix, and the elements S in the matrix are... xy,N S represents the integral step response coefficient, the subscript xy indicates the mapping of the y-th manipulated variable to the x-th controlled variable, and N is the cutoff time of the polymerization reaction process control system model. xy,N This represents the step response coefficient of the y-th manipulated variable to the x-th controlled variable at time N;

[0031] Construct the integral controlled variable for the next M steps based on the difference in its steady-state initial values. As shown below:

[0032]

[0033] Step 2), as follows:

[0034] The feasibility determination and soft constraint adjustment optimization problem for constructing integral-type controlled variables is as follows:

[0035]

[0036] stΩ r Z r ≤Ψ r

[0037] Among them, W r These are the weighting coefficients of the integral relaxation factor. and These represent the upper and lower limits of the integral relaxation factor, respectively, and Z, the decision variable in the feasibility determination and soft-constraint adjustment optimization problem of the integral controlled variable. r As shown below:

[0038]

[0039] The coefficient Ω of the constraint condition for the integral controlled variable r and constraint boundary Ψ r They are shown below:

[0040]

[0041] Among them, u min and u max These are the lower and upper limits of the manipulated variable, respectively; and These are the upper and lower limits of the hard constraints for integral-type controlled variables, respectively. and These are the upper and lower limits of the hard constraints for the integral-type controlled variable, respectively; I n It is an n-dimensional identity matrix; e r For the dynamic prediction error of the integral controlled variable; u ss (k-1) represents the manipulation variables of the steady-state optimization layer;

[0042] Solve the feasibility determination and soft constraint adjustment optimization problem of integral controlled variables to obtain the integral relaxation factor. and

[0043] Step 3) is as follows:

[0044] Feasibility assessment and soft constraint adjustment optimization problem for constructing stable controlled variables:

[0045]

[0046] st Ω s Z s ≤Ψ s

[0047] Among them, W s These are the weighting coefficients of the stable relaxation factor. and These represent the upper and lower bounds of the stable relaxation factor, respectively, and Z, the decision variable in the feasibility determination of the stable controlled variable and the soft-constraint adjustment optimization problem. s As shown below:

[0048]

[0049] The coefficient Ω of the constraint condition for a stable controlled variable s and constraint boundary Ψ s As shown below

[0050]

[0051] in, and These are the upper and lower limits of the hard constraints for stable controlled variables, respectively. and These are the upper and lower limits of the hard constraints for stable controlled variables, respectively; e s The dynamic prediction error of the stable controlled variables; q is the number of stable controlled variables; I q It is a q-dimensional identity matrix;

[0052] Solve the feasibility determination and soft constraint adjustment optimization problem for stable controlled variables to obtain the stable relaxation factor. and

[0053] Step 4), as follows:

[0054] Integral controlled variable in the next M+1 steps and Equal to each other, we obtain the constraint condition for steady-state integral slope equilibrium:

[0055]

[0056] Where, ΔS ab As shown below:

[0057]

[0058] ΔS nmThis represents the difference step response coefficient of the m-th manipulated variable to the n-th integral controlled variable.

[0059] Step 5), as follows:

[0060]

[0061] stu min ≤u ss (k)≤u max

[0062]

[0063] u ss (k+M-1)=u ss (k-1)+C ab,diff Δu ab (k)

[0064]

[0065] Among them, y TT Indicates the target tracking value; Represents all controlled variables, including and Q ss and R ss These are the weight matrices for the controlled variable and the manipulated variable, respectively;

[0066] By solving the objective function, the locally steady-state optimal target manipulated variable given by the steady-state optimization layer is obtained. Controlled variables including stable controlled variables and integral controlled variables As the target tracked by the dynamic control layer, the dynamic control layer uses a model predictive control algorithm to solve for the feed flow rate and cooling water flow rate in the polymerization reaction process.

[0067] Steady-state target tracking optimization and control systems for mixed integral controlled variables include:

[0068] The variable construction module is used to construct decision variables, manipulated variables, stable controlled variables, and integral controlled variables for steady-state optimization layer optimization problems based on the polymerization reaction process control system. Among them, the manipulated variables are used to represent the cooling water flow rate and feed flow rate in the polymerization reaction process, the stable controlled variables are used to represent the reactant concentration and mass, and the integral controlled variables are used to represent the liquid level.

[0069] The feasibility assessment and soft constraint adjustment module is used to construct a feasibility assessment and soft constraint adjustment optimization problem for integral controlled variables based on decision variables, manipulated variables, and integral controlled variables, and to obtain integral relaxation factors. and Based on decision variables, manipulated variables, stable controlled variables, integral controlled variables, and integral relaxation factors. and The feasibility determination and soft constraint adjustment optimization problem of constructing stable controlled variables is presented, and the stable relaxation factor is obtained. and

[0070] The steady-state integral slope equilibrium constraint construction module is used to construct the steady-state integral slope equilibrium constraint conditions based on the decision variables and integral controlled variables.

[0071] The local steady-state optimal objective variable solution module is used to solve for decision variables, manipulated variables, stable controlled variables, integral controlled variables, and integral relaxation factors. and Stable relaxation factor and In addition to the constraint of steady-state integral slope balance, a steady-state target tracking optimization problem is constructed, and the local steady-state optimal target manipulation variable is obtained by solving it. and controlled variable

[0072] The dynamic control module is used to manipulate the optimal target variables. and controlled variable As the tracking target of the dynamic control layer, the cooling water flow rate and feed flow rate are adjusted according to the optimal solution of the dynamic control layer.

[0073] The present invention has the following beneficial effects and advantages:

[0074] For systems employing a two-layer model predictive control (MMC) structure that includes both stable and integral controlled variables, an improved steady-state optimization layer is proposed. A new decision variable is constructed based on the sum of multi-step manipulated variables and introduced into the feasibility assessment and soft constraint adjustment stages of both stable and integral controlled variables, as well as the objective function for steady-state target tracking optimization. Constraints for steady-state target tracking optimization are constructed based on the obtained relaxation factor and steady-state integral slope balance. Finally, the steady-state target tracking optimization problem is solved to obtain the steady-state optimization objective. The dynamic control layer tracks the steady-state optimization objective, employing a model predictive control algorithm to reduce costs and energy consumption. Attached Figure Description

[0075] Figure 1 A flowchart of the present invention. Detailed Implementation

[0076] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0077] like Figure 1As shown, the present invention mainly includes the following steps: constructing decision variables and other variables for the steady-state optimization layer optimization problem; determining the feasibility and adjusting soft constraints for stable controlled variables; determining the feasibility and adjusting soft constraints for integral controlled variables; constructing constraint conditions for steady-state integral slope balance; and performing steady-state target tracking optimization and dynamic control.

[0078] This invention is designed for complex systems with mixed integral controlled variables. The manipulated variables represent the cooling water flow rate and feed flow rate in the polymerization reaction process, the stable controlled variables represent the reactant concentration and mass, and the integral controlled variable represents the liquid level. In the following text, u will represent the manipulated variable, and Δu will represent the increment of the manipulated variable.

[0079] (1) Construction of decision variables and other variables for steady-state optimization problems, including the following steps:

[0080] (1.1) Determine the decision variables for the steady-state optimization problem: For a system with m manipulated variables, the steady-state optimization problem...

[0081] Decision variable Δu in state optimization problem ab (k) is a 2m×1 matrix, as shown below:

[0082] Δu ab (k)

[0083] =[Δu 1,a (k)Δu 1,b (k)Δu 2,a (k)Δu 2,b (k)…Δu m,a (k)Δu m,b (k)] T

[0084] Where Δu i,a (k) and Δu i,b (k) are shown below:

[0085] Δu i,a (k)=MΔu i,ss (k)+(M-1)Δu i,ss (k+1)+…+Δu i,ss (k+M-1)

[0086] Δu i,b (k)=(M-1)Δu i,ss (k)+(M-2)Δu i,ss (k+1)+…+Δu i,ss (k+M-2)

[0087] Where u = 1, 2, ..., m, M is the control time domain of the dynamic control layer, k represents the time step; u represents the manipulated variable, Δu represents the increment of the manipulated variable, and the subscript ss represents the steady state.

[0088] (1.2) Expression of other variables: According to the decision variable Δu described in step (1.1) ab (k) constructs the manipulated variable, the stable controlled variable, and the integral controlled variable.

[0089] Define the difference coefficient matrix C ab,diff As shown below:

[0090]

[0091] Constructing the manipulation variable u for the future M steps ss (k+M-1) is shown below:

[0092] u ss (k+M-1)=u ss (k-1)+C ab,diff Δu ab (k)

[0093] Based on the stable steady-state gain coefficient matrix (Obtained through identification) Construct stable controlled variables for the next M steps. As shown below:

[0094]

[0095] Construct the integral-type controlled variable difference coefficient matrix S ab,N As shown below:

[0096]

[0097] Where n is the number of integral controlled variables, and S ab,N It is the step response difference coefficient matrix, and the element S in the matrix is... xy,N S represents the integral step response coefficient, the subscript xy indicates the mapping of the y-th manipulated variable to the x-th controlled variable, and N is the cutoff time of the polymerization reaction process control system model. xy,N S represents the step response coefficient of the y-th manipulated variable to the x-th controlled variable at time N; ab,N The acquisition process is existing technology, and the step response difference coefficient matrix S in the patent "A Predictive Control Method for a Two-Layer Structure Model for Oil and Gas Pipeline Network Systems" (CN118795768A) can be referenced. N,diff .

[0098] Construct the integral controlled variable for the next M steps based on the difference in its steady-state initial values. As shown below:

[0099]

[0100] (2) Feasibility determination and soft constraint adjustment for integral controlled variables, the specific operations are as follows:

[0101] Based on the decision variables, manipulated variables, stable controlled variables, and integral controlled variables from step (1), the feasibility determination and soft constraint adjustment optimization problem for integral controlled variables is constructed as follows:

[0102]

[0103] stΩ r Z r ≤Ψ r

[0104] Among them, W r These are the weighting coefficients of the integral relaxation factor. and These represent the upper and lower limits of the integral relaxation factor, respectively, and Z, the decision variable in the feasibility determination and soft-constraint adjustment optimization problem of the integral controlled variable. r As shown below:

[0105]

[0106] The coefficient Ω of the constraint condition for the integral controlled variable r and constraint boundary Ψ r As shown below

[0107]

[0108] Among them, u min and u max These are the lower and upper limits of the manipulated variable, respectively; and These are the upper and lower limits of the hard constraints for integral-type controlled variables, respectively. and These are the upper and lower limits of the hard constraints for the integral-type controlled variable, respectively; I n e is the identity matrix; r This represents the dynamic prediction error of the integral-type controlled variable.

[0109] Solve the feasibility determination and soft constraint adjustment optimization problem of integral controlled variables to obtain the relaxation factor. and

[0110] (3) Feasibility assessment and soft constraint adjustment for stable controlled variables, the specific procedures are as follows:

[0111] Based on the decision variables, manipulated variables, stable controlled variables, and integral controlled variables from step (1), and the relaxation factors from step (2) and The feasibility assessment and soft constraint adjustment optimization problem for constructing stable controlled variables is as follows:

[0112]

[0113] st Ω s Z s ≤Ψ s

[0114] Among them, W s These are the weighting coefficients of the stable relaxation factor. and These represent the upper and lower bounds of the stable relaxation factor, respectively, and Z, the decision variable in the feasibility determination of the stable controlled variable and the soft-constraint adjustment optimization problem. s As shown below:

[0115]

[0116] The coefficient Ω of the constraint condition for a stable controlled variable s and constraint boundary Ψ s As shown below

[0117]

[0118] in, and These are the upper and lower limits of the hard constraints for stable controlled variables, respectively. and These are the upper and lower limits of the hard constraints for stable controlled variables, respectively; e s The dynamic prediction error of the stable controlled variables; q is the number of stable controlled variables; I q Given a q-dimensional identity matrix, solve the feasibility determination and soft-constraint adjustment optimization problem for stable controlled variables to obtain the relaxation factor. and

[0119] (4) Constructing the constraint conditions for steady-state integral slope balance, the specific operation is as follows:

[0120] The constraint condition for steady-state integral slope equilibrium is defined as the integral controlled variable in the M+1 future steps. and The following constraint condition for the steady-state integral slope balance is obtained when they are equal:

[0121]

[0122] Where ΔS ab As shown below:

[0123]

[0124] ΔS nm The differential step response coefficient of the m-th manipulated variable to the n-th integral controlled variable is obtained through prior identification.

[0125] (5) Steady-state target tracking optimization, the specific operation is as follows:

[0126] Based on steps (1) to (4), the constraints for the steady-state target tracking optimization problem are constructed, and the objective function is as follows:

[0127]

[0128] stu min ≤u ss (k)≤u max

[0129]

[0130] u ss (k+M-1)=u ss (k-1)+C ab,diff Δu ab (k)

[0131]

[0132] Among them, y TT Indicates the target tracking value; Represents all controlled variables, including and Q ss and R ss These are the weight matrices for the controlled variable and the manipulated variable, respectively.

[0133] By solving the objective function, the locally steady-state optimal target manipulated variable given by the steady-state optimization layer is obtained. and controlled variable As the target tracked by the dynamic control layer, the dynamic control layer uses model predictive control algorithms to solve for parameters such as feed flow rate and cooling water flow rate in the polymerization reaction process.

[0134] Although the embodiments of this invention have been disclosed above, they are not intended to limit the invention. Any person skilled in the art can make possible variations and modifications to the technical solutions of this invention using the disclosed methods and techniques without departing from the spirit and scope of this invention. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this invention, without departing from the content of the technical solutions of this invention, shall fall within the protection scope of this invention. The parts of this invention not described in detail are well-known to those skilled in the art.

Claims

1. A steady-state target tracking optimization and control method for mixed integral controlled variables, characterized in that, Includes the following steps: 1) Based on the polymerization process control system, the decision variables, manipulated variables, stable controlled variables and integral controlled variables of the steady-state optimization layer optimization problem are constructed; among them, the manipulated variables are used to represent the cooling water flow rate and feed flow rate of the polymerization process, the stable controlled variables are used to represent the reactant concentration and mass, and the integral controlled variables are used to represent the liquid level; 2) Based on the decision variables, manipulated variables, and integral controlled variables, construct a feasibility determination and soft constraint adjustment optimization problem for integral controlled variables, and obtain the integral relaxation factor. and 3) Based on the decision variable, manipulated variable, stable controlled variable, integral controlled variable, and integral relaxation factor. and The feasibility determination and soft constraint adjustment optimization problem of constructing stable controlled variables is presented, and the stable relaxation factor is obtained. and 4) Based on the decision variables and integral-type controlled variables, construct the constraints for the steady-state integral slope equilibrium; 5) Based on the decision variable, manipulated variable, stable controlled variable, and integral controlled variable, the stable relaxation factor... and In addition to the constraint of steady-state integral slope balance, an objective function is constructed, and the local steady-state optimal objective manipulated variable is obtained by solving it. and controlled variable 6) Manipulate the optimal target variable and controlled variable As the tracking target of the dynamic control layer, the cooling water flow rate and feed flow rate are adjusted according to the optimal solution of the dynamic control layer.

2. The steady-state target tracking optimization and control method for mixed integral controlled variables according to claim 1, characterized in that, Step 1) is as follows: 1) For a polymerization reaction process control system with m manipulated variables, the decision variable Δu in the steady-state optimization problem is... ab (k) is a 2m×1 matrix: Δu ab (k) =[Δu 1,a (k)Δu 1,b (k)Δu 2,a (k)Δu 2,b (k)…Δu m,a (k)Δu m,b (k)] T Where Δu i,a (k) and Δu i,b (k) are shown below: Thu i,a (k)=MΔu i,ss (k)+(M-1)Δu i,ss (k+1)+…+Δu i,ss (k+M-1) Δu i,b (k)=(M-1)Δu i,ss (k)+(M-2)Δu i,ss (k+1)+…+Δu i,ss (k+M-2) Where i = 1, 2, ..., m, M is the control time domain of the dynamic control layer, k represents the time step; u represents the manipulated variable, Δu represents the increment of the manipulated variable, and the subscript ss represents the steady state; 2) Based on the decision variable Δu ab (k) construct the future M-step manipulated variables, stable controlled variables, and integral controlled variables; Construct the difference coefficient matrix C ab,diff As shown below: Constructing the manipulation variable u for the future M steps ss (k+M-1) is shown below: u ss (k+M-1)=u ss (k-1)+C ab,diff Δu ab (k) Based on the stable steady-state gain coefficient matrix Constructing stable controlled variables for the next M steps As shown below: Construct the integral-type controlled variable difference coefficient matrix S ab,N As shown below: Where n is the number of integral controlled variables, S ab,N It is the step response difference coefficient matrix, and the elements S in the matrix are... xy,N S represents the integral step response coefficient, the subscript xy indicates the mapping of the y-th manipulated variable to the x-th controlled variable, and N is the cutoff time of the polymerization reaction process control system model. xy,N This represents the step response coefficient of the y-th manipulated variable to the x-th controlled variable at time N; Construct the integral controlled variable for the next M steps based on the difference in its steady-state initial values. As shown below:

3. The steady-state target tracking optimization and control method for mixed integral controlled variables according to claim 1, characterized in that, Step 2), as follows: The feasibility determination and soft constraint adjustment optimization problem for constructing integral-type controlled variables is as follows: s.t.Ω r Z r ≤Ψ r Among them, W r These are the weighting coefficients of the integral relaxation factor. and These represent the upper and lower limits of the integral relaxation factor, respectively, and Z, the decision variable in the feasibility determination and soft-constraint adjustment optimization problem of the integral controlled variable. r As shown below: The coefficient Ω of the constraint condition for the integral controlled variable r and constraint boundary Ψ r They are shown below: Among them, u min and u max These are the lower and upper limits of the manipulated variable, respectively; and These are the upper and lower limits of the hard constraints for integral-type controlled variables, respectively. and These are the upper and lower limits of the hard constraints for the integral-type controlled variable, respectively; I n It is an n-dimensional identity matrix; e r For the dynamic prediction error of the integral controlled variable; u ss (k-1) represents the manipulation variables of the steady-state optimization layer; Solve the feasibility determination and soft constraint adjustment optimization problem of integral controlled variables to obtain the integral relaxation factor. and 4. The steady-state target tracking optimization and control method for mixed integral controlled variables according to claim 1, characterized in that, Step 3) is as follows: Feasibility assessment and soft constraint adjustment optimization problem for constructing stable controlled variables: s.t.Ω s Z s ≤Ψ s Among them, W s These are the weighting coefficients of the stable relaxation factor. and These represent the upper and lower bounds of the stable relaxation factor, respectively, and Z, the decision variable in the feasibility determination of the stable controlled variable and the soft-constraint adjustment optimization problem. s As shown below: The coefficient Ω of the constraint condition for a stable controlled variable s and constraint boundary Ψ s As shown below in, and These are the upper and lower limits of the hard constraints for stable controlled variables, respectively. and These are the upper and lower limits of the hard constraints for stable controlled variables, respectively; e s The dynamic prediction error of the stable controlled variables; q is the number of stable controlled variables; I q It is a q-dimensional identity matrix; Solve the feasibility determination and soft constraint adjustment optimization problem for stable controlled variables to obtain the stable relaxation factor. and 5. The steady-state target tracking optimization and control method for mixed integral controlled variables according to claim 1, characterized in that, Step 4), as follows: Integral controlled variable in the next M+1 steps and Equal to each other, we obtain the constraint condition for steady-state integral slope equilibrium: Where, ΔS ab As shown below: ΔS nm This represents the gain of the step response coefficient of the m-th manipulated variable to the n-th integral controlled variable.

6. The steady-state target tracking optimization and control method for mixed integral controlled variables according to claim 1, characterized in that, Step 5), as follows: s.t.u min ≤u ss (k)≤u max u ss (k+M-1)=u ss (k-1)+C ab,diff Δu ab (k) Among them, y TT Indicates the target tracking value; Represents all controlled variables, including and Q ss and R ss These are the weight matrices for the controlled variable and the manipulated variable, respectively; By solving the objective function, the locally steady-state optimal target manipulated variable given by the steady-state optimization layer is obtained. Controlled variables including stable controlled variables and integral controlled variables As the target tracked by the dynamic control layer, the dynamic control layer uses a model predictive control algorithm to solve for the feed flow rate and cooling water flow rate in the polymerization reaction process.

7. A steady-state target tracking optimization and control system for mixed integral controlled variables, characterized in that, include: The variable construction module is used to construct decision variables, manipulated variables, stable controlled variables, and integral controlled variables for steady-state optimization layer optimization problems based on the polymerization reaction process control system. Among them, the manipulated variables are used to represent the cooling water flow rate and feed flow rate in the polymerization reaction process, the stable controlled variables are used to represent the reactant concentration and mass, and the integral controlled variables are used to represent the liquid level. The feasibility assessment and soft constraint adjustment module is used to construct a feasibility assessment and soft constraint adjustment optimization problem for integral controlled variables based on decision variables, manipulated variables, and integral controlled variables, and to obtain integral relaxation factors. and Based on decision variables, manipulated variables, stable controlled variables, integral controlled variables, and integral relaxation factors. and The feasibility determination and soft constraint adjustment optimization problem of constructing stable controlled variables is presented, and the stable relaxation factor is obtained. and The steady-state integral slope equilibrium constraint construction module is used to construct the steady-state integral slope equilibrium constraint conditions based on the decision variables and integral controlled variables. The local steady-state optimal objective variable solution module is used to solve for decision variables, manipulated variables, stable controlled variables, integral controlled variables, and integral relaxation factors. and Stable relaxation factor and In addition to the constraint of steady-state integral slope balance, a steady-state target tracking optimization problem is constructed, and the local steady-state optimal target manipulation variable is obtained by solving it. and controlled variable The dynamic control module is used to manipulate the optimal target variables. and controlled variable As the tracking target of the dynamic control layer, the cooling water flow rate and feed flow rate are adjusted according to the optimal solution of the dynamic control layer.

Images