Reverse regulation and control method for food and medicine production processing

Abstract

Provided is a reverse regulation and control method for food and medicine production processing. The method comprises the following steps: S1, on the basis of product requirements for food and medicine production, determining an expected output trajectory, the expected output trajectory being a production processing target, and, on the basis of a biochemical reaction process, constructing a process dynamics model (S1); S2, according to the production processing target, performing an inverse operation on the process dynamics model, and solving a corresponding production processing regulation and control curve, the production processing regulation and control curve being production operation processing (S2); and S3, on the basis of a given food and medicine production target, performing a control operation using the production operation processing, so as to complete the food and medicine production target (S3). The described reverse regulation and control method for production processing can more efficiently and quickly obtain regulation and control processing of a production process, and provide solutions for rapid design of regulation and control processing for different products, solving the problem of dynamic regulation and control processing design.

Description

A food and medicine production process reverse control method Technical Field

[0001] The present invention relates to the technical field of dynamic regulation of biochemical reaction processes under given production process targets, and in particular to a reverse regulation method for food and medicine production processes. Background Art

[0002] With global economic development and social progress, small-batch, high-profit, and high-variety production methods, exemplified by industrial biochemical technologies, are gaining popularity to meet people's ever-growing needs. Food and pharmaceutical production typically involves biochemical changes or reactions, and product targets are determined by the control of key process variables such as temperature, flow rate, and concentration. Once the process is established, systems such as field digital controllers and industrial distributed control systems (DCS) can implement real-time control according to the process curve to achieve the final production requirements or product goals.

[0003] The design of production control processes is closely related to product quality, output, and production time. For the food and pharmaceutical industries, control process design is the foundation of production. The increasing diversity of product types places higher demands on the rapid design of control processes, making them a key component of intelligent manufacturing. Generally, control processes rely primarily on repeated experiments by process engineers, but these experimental methods require significant manpower, material resources, and time, making them particularly difficult to adapt to the diverse, small-batch, and multi-target production demands of modern intelligent manufacturing. Furthermore, existing experimental design methods struggle to capture dynamic production operation processes.

[0004] Summary of the Invention

[0005] The present invention provides a reverse control method for food and pharmaceutical production processes. A process kinetics model is constructed based on a biochemical reaction process, and an inverse operation is performed on the process kinetics model. Under the condition of ensuring the stability of process kinetic changes, the production operation process under a given production target is obtained. This overcomes the problem that traditional experimental methods require a large amount of manpower, material resources and time. The control process of the production process can be obtained more effectively and quickly, providing a solution for the rapid design of control processes for different products and solving the problem of dynamic control process design.

[0006] In order to solve the above technical problems, an embodiment of the present invention provides a method for reverse control of food and medicine production processes, which includes the following steps:

[0007] Step S1: Determine the expected output trajectory based on the product requirements of food and pharmaceutical production, which is the production process target, and construct a process kinetic model based on the biochemical reaction process;

[0008] Step S2: performing an inverse operation on the process kinetics model according to the production process target to obtain a corresponding production process control curve, where the production process control curve is a production operation process;

[0009] Step S3: Based on the given production target of food or medicine, the production operation process is used to perform control operations to achieve the production target of food or medicine.

[0010] In one embodiment of the present invention, in step S1, the process dynamics model is written as a process state equation as follows:

[0011] Where x∈R n For the biochemical process state, is the state trajectory, u∈R m is the process input, y∈R m is the process output, f(·), g1(·), ..., g m (·), h1(·), ..., h m (x) are all nonlinear smooth functions;

[0012] For the convenience of representation, the above process state equation is written in vector form: y=h(x)

[0013] Among them, u=[u1,...,u m ] T ,y=[y1,...,y m ] T ,x=[x1,...,x n ] T ,g(x)=[g1(x),...,g m (x)],h(x)=[h1(x),...,h m (x)] T , [·] T T represents the transpose of a vector or matrix;

[0014] Let the output expected trajectory be y d =[y d1 ,...,y dm ], according to the process state equation in vector form above, the relationship between the production process control curve and the state trajectory is as follows: y d =h(x d )

[0015] Among them, u d is the production process control curve, u d =[u d1 ,...,udm ];x d is the state trajectory, x d =[x d1 ,...,x dn ].

[0016] In one embodiment of the present invention, the output y of the process kinetics model is i Find the derivative:

[0017] First output y i Find the first-order derivative:

[0018] Among them, express Represents the function h i (x) along the derivative of the vector field f(x), similarly Represents the function h i (x) along the vector field g j The derivative of (x) is when the i-th output y i The relative order of is greater than 1, then At this time, we can get:

[0019] Then output y i Find the k-th derivative:

[0020] Among them, r i is the i-th output y i The relative order of

[0021] Finally, for the i-th output y i Find r i Derivative, then

[0022] in,

[0023] Use the vector [r1,r2,...,r m ],r=r1+r2+...r m ≤n represents the relative order of m outputs, for each output y i Corresponding to r i The secondary derivative is as follows:

[0024] For simplicity, the above formula is written in vector form as follows: (r) =B(x)+A(x)u

[0025] in,

[0026] Let A(x) be a non-singular invertible matrix at the point x = x°, and obtain the inverse model of the process dynamics model: u = (A(x°)) -1 (y (r) -B(x)

[0027] Among them, u is the process input, y is the process output, x∈R n is the state of the biochemical process, x° is the state of the biochemical process x∈R n The local steady-state value of (A(x°)) -1 represents the inverse matrix of A(x) at the local steady state.

[0028] In one embodiment of the present invention, the biochemical process state x∈R n It includes external dynamics ξ and internal dynamics η, and the external dynamics are defined as the desired output trajectory:

[0029] Among them, y1, y2, ..., y r is the process output of the process dynamics model, is the derivative of each order of the process output of the process kinetics model;

[0030] Define the internal dynamics as: η=[η1,η2,...,η n-r ] T .

[0031] In one embodiment of the present invention, according to the inverse model of the process dynamics model, in the coordinate system (ξ, η), let y = y d , that is, (ξ,η)=(ξ d ,η d ), u=u d , the inverse model between internal and external dynamics is obtained as follows:

[0032] Among them, a(ξ d ,η d ) is a non-singular reversible matrix, ξ d is the external dynamics, η d is the internal dynamics, u d It is the process control curve.

[0033] In one embodiment of the present invention, in step S2, the specific steps of solving the production process control curve are as follows:

[0034] Step S21: Knowing that the desired output trajectory is the external dynamics, obtain the internal dynamics;

[0035] Step S22: Based on the internal dynamics obtained above, a production process control curve is obtained according to an inverse model between the internal and external dynamics;

[0036] Among them, in step S21, the specific steps of obtaining internal dynamics are as follows:

[0037] Step S211: Construct nonlinear internal dynamic equations:

[0038] in, ξ0 is the initial steady-state value of the external dynamics, η0 represents the initial steady-state value of the internal dynamics;

[0039] Step S212: Convert the above nonlinear internal dynamic equation into an integral equation:

[0040] Q consists of a stable part and an unstable part:

[0041] Among them, Q s represents the stable part, all its eigenvalues ​​are in the left half plane of the matrix Q; Q u represents the unstable part, all of whose eigenvalues ​​are in the right half plane of the matrix Q;

[0042] Define the bounded state transfer matrix φ(t) as:

[0043] Then the above nonlinear internal dynamic equation is equivalent to the following integral equation, using η d (t),ξ d (t) represents η d ,ξ d , select the integration time domain as [t,t+t p ], the approximate integral equation is as follows:

[0044] Step S213: Use the Picard iterative method to solve the above approximate integral equation:

[0045] Where m≥1, m is the number of iterations, when m→∞, η m (t)→η d (t);

[0046] Step S214: Use discretization method to obtain internal dynamics in vector form:

[0047] The internal dynamics η(t) are expressed in the time domain t∈[t0,t0+t p ] is evenly divided into n segments, and the length of each segment is Inside, η(t) is approximately a constant, and the internal dynamics in vector form after discretization are as follows:

[0048] in, Denotes η(t) at time t∈[t0,t0+t p ]The discretized vector within;

[0049] When the number of Picard iterations m increases to the point where the unstable internal dynamics and stabilize internal dynamics When converged, the internal dynamics in vector form are obtained for:

[0050] Where M is the reversible internal dynamic decomposition matrix;

[0051] In step S22, the internal dynamics in the form of the above vector is used According to the inverse model between internal and external dynamics, the calculation formula of the production process control curve is obtained as follows:

[0052] Among them, n t = round up ((t-t0)n / (t p -t0)), is a non-singular invertible matrix.

[0053] In one embodiment of the present invention, in step S213, the approximate integral equation is divided into unstable internal dynamic η u (t) and stable internal dynamics η s (t), whose eigenvalues ​​are λ1 and -λ2 respectively, then:

[0054] Among them, η d =M -1 [η u η s ] T , M is the reversible internal dynamic decomposition matrix, λ1,λ2>0.

[0055] In one embodiment of the present invention, the process kinetics model is a material balance, heat balance and energy balance relationship equation between materials.

[0056] In one embodiment of the present invention, the production process objectives include: requirements for product production with concentration and temperature as output variables during the production process.

[0057] In one embodiment of the present invention, the production operation process includes: a production operation process with feed flow rate and jacket temperature as operating variables.

[0058] As can be seen from the above technical solutions, the inverse control method for food and pharmaceutical production processes described herein utilizes inverse design methods based on known biochemical process kinetic models to solve the production control process corresponding to the production output target, providing a rapid method for determining the biochemical control process. This method not only avoids the significant labor, material, and time consumption associated with experimental methods for determining production operation processes, but is also applicable to determining production operation processes based on dynamically changing production requirements, significantly improving the efficiency of determining biochemical control processes and enterprise production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] FIG1 is a flow chart of the reverse control method for food and medicine production processes according to the present invention;

[0060] FIG2 is a diagram showing an example of a biochemical process reactor used in the present invention;

[0061] FIG3 is a process control curve of feed flow rate and jacket temperature of an application example of a biochemical reactor of the present invention;

[0062] FIG4 is a diagram comparing the production requirements and actual production results of material B in an application example of a biochemical reactor of the present invention. DETAILED DESCRIPTION

[0063] The present invention will be further described below with reference to the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it. However, the embodiments are not intended to limit the present invention.

[0064] 1 , the food and medicine production process reverse control method of the present invention comprises the following steps:

[0065] Step S1: Determine the expected output trajectory based on the product requirements of food and pharmaceutical production, which is the production process target, and construct a process kinetic model based on the biochemical reaction process;

[0066] Step S2: performing an inverse operation on the process kinetics model according to the production process target to obtain a corresponding production process control curve, where the production process control curve is a production operation process;

[0067] Step S3: Based on the given production target of food or medicine, the production operation process is used to perform control operations to achieve the production target of food or medicine.

[0068] The inverse control method for food and pharmaceutical production processes provided by the present invention constructs a process kinetics model based on the biochemical reaction process, performs inverse operations on the process kinetics model, and obtains the production operation process under a given production target while ensuring the stability of the process kinetic changes. This overcomes the problem that traditional experimental methods require a large amount of manpower, material resources and time, and can more effectively and quickly obtain the control process of the production process, providing a solution for the rapid design of control processes for different products; and solves the problem of dynamic control process design.

[0069] In this embodiment, in step S1, the process dynamics model is written as the process state equation as follows:

[0070] Where x∈R n For the biochemical process state, is the state trajectory, u∈R m is the process input, y∈R m is the process output, f(·), g1(·), ..., g m (·), h1(·), ..., h m (x) are all nonlinear smooth functions;

[0071] For the convenience of representation, the above process state equation is written in a more compact vector form:

[0072] In the formula: u=[u1,...,u m ] T ,y=[y1,...,y m ] T ,x=[x1,...,x n ] T ,g(x)=[g1(x),...,g m (x)],h(x)=[h1(x),...,h m (x)] T ;

[0073] in,[·] T T represents the transpose of a vector or matrix;

[0074] Let the output expected trajectory be y d =[y d1 ,...,y dm ], according to the process state equation in vector form above, the relationship between the production process control curve and the state trajectory is as follows:

[0075] Among them, u d is the production process control curve, u d =[ud1 ,...,u dm ];x d is the state trajectory, x d =[x d1 ,...,x dn ].

[0076] Furthermore, for the process kinetics model, the output y of the process i is i Find the derivative:

[0077] First, output y for the i-th process i Find the first-order derivative:

[0078] For the convenience of description, the above formula uses express and Represents the function h i (x) along the derivative of the vector field f(x), similarly Represents the function h i (x) along the vector field g j The derivative of (x) is when the i-th output y i The relative order of is greater than 1, then At this time, we can get:

[0079] Then for the i-th output y i Find the second-order derivative:

[0080] in ) is expressed as a function L f h i (x) along the derivative of the vector field f(x), similarly Expressed as function L f h i (x) along the vector field g j (x). If the i-th output y i The relative order of is greater than 2, at this time:

[0081] Continue to output y for the i-th i Find the k-th derivative:

[0082] Among them, r i is the i-th output y i The relative order of

[0083] Finally, for the i-th output y i Find r i Derivative, then

[0084] in,

[0085] Use the vector [r1,r2,...,r m ],r=r1+r2+...r m ≤n represents the relative order of m outputs, for each output y i Corresponding to r i The secondary derivative is as follows:

[0086] For simplicity, the above formula is written in vector form as follows: (r) =B(x)+A(x)u

[0087] in,

[0088] Let A(x) be a non-singular invertible matrix at the point x = x°, and obtain the inverse model of the process dynamics model: u = (A(x°)) -1 (y (r) -B(x)

[0089] Among them, u is the process input, y is the process output, x∈R n is the state of the biochemical process, x° is the state of the biochemical process x∈R n The local steady-state value of (A(x°)) -1 represents the inverse matrix of A(x) at the local steady state.

[0090] Specifically, the biochemical process state x∈R n It consists of two parts: one is the external dynamics ξ that reflects the input-output relationship, and the other is the internal dynamics η that cannot be observed from the output-output relationship. The external dynamics are defined as the expected output trajectory:

[0091] Among them, r is the number of external dynamic corresponding, y1,y2,...,y r is the process output of the process dynamics model, is the derivative of each order of the process output of the process kinetics model;

[0092] Define the internal dynamics as: η=[η1,η2,...,η n-r ] T ,

[0093] Among them, nr is the number of internal dynamic correspondences.

[0094] In step S2, solving the production process control curve includes the following steps:

[0095] Step S21: Knowing that the desired output trajectory is the external dynamics, obtain the internal dynamics;

[0096] Step S22: Based on the internal dynamics obtained above, a production process control curve is obtained according to an inverse model between the internal and external dynamics;

[0097] Among them, in step S21, the specific steps of obtaining internal dynamics are as follows:

[0098] Step S211: Construct nonlinear internal dynamic equations:

[0099] in, ξ0 is the initial steady-state value of the external dynamics, η0 represents the initial steady-state value of the internal dynamics;

[0100] Step S212: Convert the above nonlinear internal dynamic equation into an integral equation:

[0101] Q consists of a stable part and an unstable part:

[0102] Among them, Q s represents the stable part, all its eigenvalues ​​are in the left half plane of the matrix Q; Q u represents the unstable part, all of whose eigenvalues ​​are in the right half plane of the matrix Q;

[0103] Define the bounded state transfer matrix φ(t) as:

[0104] Then the above nonlinear internal dynamic equation is equivalent to the following integral equation, using η d (t),ξ d (t) represents η d ,ξ d , select the integration time domain as [t,t+t p ], the approximate integral equation is as follows:

[0105] Step S213: Use the Picard iterative method to solve the above approximate integral equation:

[0106] Where m≥1, m is the number of iterations, when m→∞, η m (t)→η d (t);

[0107] Step S214: Use discretization method to obtain internal dynamics in vector form:

[0108] The internal dynamics η(t) are expressed in the time domain t∈[t0,t0+t p] is evenly divided into n segments, and the length of each segment is Inside, η(t) is approximately a constant, and the internal dynamics in vector form after discretization are as follows:

[0109] in, Denotes η(t) at time t∈[t0,t0+t p ]The discretized vector within;

[0110] When the number of Picard iterations m increases to the point where the unstable internal dynamics and stabilize internal dynamics When converged, the internal dynamics in vector form are obtained for:

[0111] Where M is the reversible internal dynamic decomposition matrix;

[0112] In step S22, the internal dynamics in the form of the above vector is used According to the inverse model between internal and external dynamics, the calculation formula of the production process control curve is obtained as follows:

[0113] Among them, n t = round up ((t-t0)n / (t p -t0)), is a non-singular invertible matrix.

[0114] Specifically, in step S213, the approximate integral equation is divided into unstable internal dynamic η u (t) and stable internal dynamics η s (t), whose eigenvalues ​​are λ1 and -λ2 respectively, then:

[0115] Among them, η d =M -1 [η u η s ] T , M is the reversible internal dynamic decomposition matrix, λ1,λ2>0.

[0116] In this embodiment, the process kinetic model is the material balance, heat balance and energy balance relationship equations between materials; the production process objectives include: the requirements for product production with concentration and temperature as output variables in the production process; the production operation process includes: the production operation process with feed flow rate and reaction equipment jacket temperature as operating variables.

[0117] The application principle of the reverse control method for food and pharmaceutical production described in this embodiment is explained below in conjunction with actual application scenarios. A specific example of preparing cyclopentene using a continuous stirred tank reactor (CSTR) is given. Cyclopentene is commonly used in the pharmaceutical field.

[0118] As shown in Figure 2, CSTR is a reactor widely used in biochemical reactions and is a typical, highly nonlinear chemical reactor in the process industry. The chemical production process of cyclopentene using CSTR has typical nonlinear and non-minimum phase characteristics. The chemical reaction formula is as follows:

[0119] The above reaction is to generate the main product cyclopentene (referred to as material B) and the by-product dicyclopentadiene (referred to as material D) from cyclopentadiene (referred to as material A), and the by-product cyclopentanone (referred to as material C) is generated by further reaction of material B.

[0120] The specific steps of the reverse control method for cyclopentene production process include:

[0121] Step 1: Construct a kinetic model of the biochemical reaction process

[0122] For the above biochemical reaction process, a biochemical reaction process kinetic model is constructed, and the process kinetic model is written to satisfy the molar balance equation and energy balance equation of material A and material B as follows:

[0123] Among them C A and C B is the molar concentration of materials A and B, T is the reactor temperature, F is the feed flow rate, T C is the temperature of the cooling jacket. C P and ρ are the specific heat capacity and density of the reaction mixture, respectively, ΔH i is the heat of reaction.

[0124] The reaction rate coefficient is related to the reaction temperature and process parameters. The specific system parameters are shown in Table 1. The specific calculation formula of the reaction rate coefficient is as follows: k i (T) = k i0 exp(E i / RT)

[0125] Table 1

[0126] According to the needs of the production product, the molar concentration C of material B B The desired output trajectory is required for the reactor temperature T, while the control variables are the feed flow rate and the equipment jacket temperature. CAs input, the molar concentration of material B is C B and the reactor temperature T as output, the reactor steady-state equilibrium point is and T e =407.15K, which corresponds to and

[0127] Step 2: Determine production process goals

[0128] Now assume that the production process goal of a specific production process is to require the molar concentration of material B to change according to the following expected trajectory with production time t while controlling the reactor temperature to be constant:

[0129] Where 0<t≤t f , t f =0.9 is the terminal time of regulation.

[0130] Step 3: Use the model inverse algorithm to obtain the biochemical process control curve

[0131] The specific solution steps include:

[0132] Step 1: Rewrite the above biochemical reactor model into the standard state equation form:

[0133] The model input u1=F,u2=T C , state variable x1=C B ,x2=T,x3=C A , the model outputs y1=x1, y2=x2, and the inverse operation of the model is the known y d For the desired output trajectory, we need to find the input trajectory u d and state trajectory x d , that is, it is known that: y 2d =0

[0134] Step 2: Derivative the outputs y1 and y2 in the state equation of step 1 with respect to time t for r1 and r2 times respectively until the input u = [u1, u2] is obtained. T and state variables x = [x1, x2, x3] T For this example, r1=r2=1 is the relative order of the process system, that is, the inverse model is:

[0135] Step 3: Transform the state variable x in step 1 into the form of external dynamics ξ and internal dynamics η. For this example, external dynamics ξ=[ξ1,ξ2] T =[x1,x2] T , internal dynamics η=[η1]=[x3]. From the internal and external dynamic expressions, we can see that this example does not need to perform internal and external dynamic decomposition through coordinate transformation. It is originally a dynamic system in the standard internal and external dynamic form.

[0136] Step 4: In the new coordinate system of the state variables in step 3, that is, [ξ1,ξ2,η1]=[x1,x2,x3], rewrite the inverse model in step 2 into the form of internal and external dynamic expressions:

[0137] Step 5: Due to the desired external dynamics ξ d = = [y 1d ,y 2d ] T Known, so the inverse model u d The calculation is to solve the internal state reference trajectory, that is, the bounded solution η1 of the internal state. The internal dynamics can be solved using the nonlinear internal dynamic equation. For the nonlinear internal dynamic equation, u in step 4 can be substituted into the state equation under the new coordinate system (ξ,η) to obtain:

[0138] in Substitute the steady-state initial value into the above formula and rewrite the nonlinear internal dynamic equation into the following form:

[0139] The eigenvalue of Q is in the right half plane, that is, η1 is an unstable internal dynamics, and the solution of the nonlinear internal dynamics equation needs to be solved by reverse integration. That is, the solution of the above internal dynamics equation is equivalent to the following integral equation:

[0140] Where ξ1, ξ2 and η1 should be ξ1(τ), ξ2(τ) and η1(τ), which are abbreviated here for convenience and belong to the integrand. In the actual application process, only a period of time [0,t p ] output trajectory information within the preview time t p =0.9, the longer it is, the smaller the tracking error is. That is, the above integral equation can be approximated as:

[0141] Step 6: The nonlinear integral equation in step 5 can be solved by the Picard iteration method, namely:

[0142] Where m is the number of iterations. When m→∞, η m (t)→η d (t).

[0143] Step 7: In the actual application process, as the number of iterative integrations m increases in step 6, the result gradually converges to the actual value η d (t), but η m The analytical expression of (t) will become complex and even difficult to describe, resulting in an increasingly large amount of calculation until the application problem becomes uncomputable. In such a case, the internal dynamic η(t) needs to be discretized in the computer program.

[0144] Assume that the internal dynamic η(t) is uniformly decomposed into n segments in the future, that is, into n+1 discrete numerical vectors, with each segment length t p / n time, η(t) is approximately considered to be a constant value, rather than the original function that continuously changes with time. The discretized internal dynamic vector form is expressed as follows:

[0145] in Representation function A vector with a length of n+1=900 after discretization within time t∈[0,0.9].

[0146] Since the internal dynamics η(t) solved in step 6 is an unstable internal dynamics, the unstable internal dynamics vector value is taken The recursive formula solves each iteration process The numerical vector solution of , for this implementation case, the specific solution formula is as follows:

[0147] because is the numerical vector obtained by the m-1th Picard iteration. When solving the mth integral iteration, the integral solution results in each recursive formula are in numerical form. This process avoids solving complex analytical solutions and greatly reduces the amount of calculation.

[0148] Leveraging known internal dynamics To solve the inverse model u d , according to the formula in step 4:

[0149] where n t = rounded (1000t / (0.9)), u d1 is the process control curve of feed flow rate, u d2 It is the process control curve of the equipment jacket temperature, and the process control curve is also the production operation process.

[0150] Step 4: Use production operation process to control operations to achieve production goals

[0151] Since the molar concentration of material B is C B and reactor temperature T are the desired output trajectory, that is, the production process target. The two inverse models u obtained in step 7 above are d1 and u d2 They are the production operation processes for feed flow rate and equipment jacket temperature, which are used to control operations to achieve target production requirements.

[0152] In order to verify the effect of the reverse control method for the production process implemented in this application, Figure 3 is the process control curve of the feed flow rate and the reactor jacket temperature obtained according to the production target, and Figure 4 is a comparison of the actual production and production demand results of material B obtained according to the process control curve of the feed flow rate in Figure 3. It can be seen that the "stable reverse method" curve basically fits the "CB reference trajectory" curve. The "stable reverse method" curve represents the actual production result of material B completed by the reverse control method in this embodiment, and the "CB reference trajectory" curve represents the production demand of material B, which means that the actual production result of material B completed by the reverse control method in this embodiment can meet the production demand of material B, further highlighting the effectiveness of the reverse control method for the production process provided by this embodiment.

[0153] It can be seen from this that the reverse control method for food and pharmaceutical production processes provided by the present invention constructs a process kinetics model based on the biochemical reaction process, performs inverse operations on the process kinetics model, and obtains the production operation process under a given production target while ensuring the stability of the process kinetics changes. This overcomes the problem that traditional experimental methods require a large amount of manpower, material resources and time. It can not only obtain the control process of the production process more effectively and quickly, providing a solution for the rapid design of the control process of different products, but also solves the problem of dynamic control process design.

[0154] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0155] The present application is described with reference to the flow chart and / or block diagram of the method, device (system), and computer program product according to the embodiment of the present application. It should be understood that each flow process and / or box in the flow chart and / or block diagram and the combination of the flow process and / or box in the flow chart and / or block diagram can be realized by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processing machine or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for realizing the function specified in one flow chart flow or multiple flows and / or one box or multiple boxes of the block diagram.

[0156] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0157] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0158] Obviously, the above embodiments are merely examples for clarity of explanation and are not intended to limit the implementation methods. Those skilled in the art will appreciate that other variations or modifications can be made based on the above description. It is not necessary and impossible to enumerate all implementation methods here. Obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A method for reverse control of food and medicine production process, characterized in that: The following steps are involved: Step S1: Determine the expected output trajectory according to the product needs of food and medicine production, the expected output trajectory is the production process target, and build a process kinetic model according to the biochemical reaction process; Step S2: According to the production process target, the process kinetics model is inversely calculated to obtain a corresponding production process control curve, where the production process control curve is a production operation process; Step S3: Based on the given production target of food or medicine, the production operation process is used to perform control operations to achieve the production target of food or medicine.

2. The food and medicine production process reverse control method according to claim 1, characterized in that: In step S1, the process kinetic model is written as the process state equation as follows: Where x∈R n The state of the biochemical process, is the state trajectory, u∈R m is the process input, y∈R m is the process output, f(·), g1(·), …, g m (·), h1(·), …, h m (x) are all nonlinear smooth functions; For the convenience of representation, the above process state equation is written in vector form: y=h(x) Among them, u=[u1,…,u m ] T ,y=[y1,…,y m ] T ,x=[x1,…,x n ] T ,g(x)=[g1(x),…,g m (x)],h(x)=[h1(x),…,h m (x)] T , [·] T T represents the transpose of a vector or matrix; Let the output expected trajectory be y d =[y d1 ,…,y dm ], according to the process state equation in vector form, the relationship between the production process control curve and the state trajectory is as follows: y d =h(x d ) Among them, u d is the production process control curve, u d =[u d1 ,…,u dm ]; x d is the state trajectory, x d =[x d1 ,…,x dn ].

3. The inverse design method of the food and medicine production process control curve according to claim 2, characterized in that: For the process dynamics model, the output y of the i-th process is i Find the derivative: First, output y i Find the first-order derivative: Among them, 1≤i≤m, use express Represents the function h i (x) along the derivative of the vector field f(x), similarly Represents the function h i (x) along the vector field g j The derivative of (x) is i If the relative order of is greater than 1, then At this time, we can get: Then output y i Find the k-th order derivative: Among them, r i is the i-th output y i The relative order of Finally, for the i-th output y i Request r i The derivative is in, Use the vector [r1,r2,…,r m ],r=r1+r2+…r m ≤n represents the relative order of m outputs. For each output y i Corresponding to r i The secondary derivative is as follows: For simplicity, the above formula is written in vector form as follows: y (r) =B(x)+A(x)u in, Let A(x) be a non-singular reversible matrix at the point x = x°, and obtain the inverse model of the process dynamics model: u=(A(x°)) -1 (y (r) -B(x)) Among them, u is the process input, y is the process output, x∈R n is the state of the biochemical process, x° is the state of the biochemical process x∈R n The local steady-state value of (A(x°)) -1 represents the inverse matrix of A(x) at the local steady state.

4. The food and medicine production process reverse control method according to claim 2, characterized in that: The biochemical process state x∈R n It includes external dynamics ξ and internal dynamics η, and the external dynamics are defined as the desired output trajectory: Among them, y1,y2,...,y r is the process output of the process kinetics model, are the derivatives of the process output of the process kinetics model; Define the internal dynamics as: n=[n1,n2,…,n n-r ] T 。 5. The food and medicine production process reverse control method according to claim 3, characterized in that: According to the inverse model of the process kinetic model, in the coordinate system (ξ,η), let y = y d , that is, (ξ,η)=(ξ d ,η d ), u=u d , the inverse model between internal and external dynamics is obtained as follows: Among them, a(ξ d ,η d ) is a non-singular reversible matrix, ξ d is the external dynamics, η d is the internal dynamics, u d It is the process control curve.

6. The method for reverse control of food and medicine production process according to claim 5, characterized in that: In step S2, the specific steps for solving the production process control curve are as follows: Step S21: Knowing that the expected output trajectory is the external dynamics, obtain the internal dynamics; Step S22: According to the internal dynamics obtained above, according to the inverse model between the internal and external dynamics, a production process control curve is obtained; Among them, in step S21, the specific steps of obtaining the internal dynamics are as follows: Step S211: Construct nonlinear internal dynamic equations: in, ξ0 is the initial steady-state value of the external dynamics, η0 is the initial steady-state value of the internal dynamics; Step S212: Convert the above nonlinear internal dynamic equation into an integral equation: Q consists of a stable part and an unstable part: Among them, Q s represents the stable part, all its eigenvalues ​​are in the left half plane of the matrix Q; Q u represents the unstable part, all its eigenvalues ​​are in the right half plane of the matrix Q; Define the bounded state transfer matrix φ(t) as: Then the above nonlinear internal dynamic equation is equivalent to the following integral equation, using η d (t), ξ d (t) represents η d , d , select the integration time domain as [t,t+t p ], the approximate integral equation is as follows: Step S213: Use the Picard iteration method to solve the above approximate integral equation: Where m≥1, m is the number of iterations, when m→∞, η m (t)→η d (t); Step S214: Using a discretization method to obtain the internal dynamics in vector form: The internal dynamics η(t) are expressed in the time domain t∈[t0,t0+t p ] is evenly divided into n segments, and the length of each segment is Inside, η(t) is approximately a constant, and the internal dynamics in vector form after discretization are as follows: in, represents η(t) at time t∈[t0,t0+t p ]The discretized vector in ; When the number of Picard iterations m increases to an unstable internal dynamic and stabilize internal dynamics When converged, the internal dynamics in vector form is obtained for: Where M is the reversible internal dynamic decomposition matrix; In step S22, the internal dynamics in the form of the above vector is used According to the inverse model between internal and external dynamics, the calculation formula of the production process control curve is obtained as follows: Among them, n t =Round((t-t0)n / (t p -t0)), is a non-singular invertible matrix.

7. The food and medicine production process reverse control method according to claim 6, characterized in that: In step S213, the approximate integral equation is divided into unstable internal dynamic η u (t) and stable internal dynamics η s (t), whose eigenvalues ​​are λ1 and -λ2 respectively, then: Among them, η d =M -1 [η u η s ] T , M is the reversible internal dynamic decomposition matrix, λ1,λ2>0.

8. The food or medicine production process reverse control method according to claim 1 or 2, characterized in that: The process kinetic model is a material balance, heat balance and energy balance relationship equation between materials.

9. The food and medicine production process reverse control method according to claim 1, characterized in that: The production process objectives include: requirements for the production of products with concentration and temperature as output variables during the production process.

10. The method for reverse control of food and medicine production process according to claim 1, characterized in that: The production operation process includes: a production operation process with feed flow rate and equipment jacket temperature as operation variables.

Images