Method and system for evaluating optimal performance of control effect in fermentation process of 1, 3-propylene glycol

Abstract

The invention provides a 1, 3-propylene glycol fermentation process control effect optimal performance evaluation method and system, and relates to the technical field of fermentation process regulation, the method comprises the following steps: constructing a matrix space model of a 1, 3-propylene glycol fermentation process, and obtaining an output variable measurement value; predicting a state variable and designing an estimator based on the model information, and reconstructing the state variable by using the estimator in combination with the output measured value and the predicted value to obtain a state variable estimated value; designing an objective function by combining model information according to the state variable estimation value, and determining control input of the fermentation system by minimizing the function; and estimating the deviation between the current control input and the optimal control input through a partial derivative by utilizing model information, and constructing an index for evaluating the optimum control effect of the fermentation process based on the deviation of the current control input and an objective function. According to the method, real-time evaluation is achieved, the response speed is increased, specific uncertainty information is not needed, universality is high, and it is guaranteed that the fermentation process is efficient, stable and reliable.

Description

Technical Field

[0001] This invention relates to the field of fermentation process control technology, and in particular to a method and system for evaluating the optimal performance of 1,3-propanediol fermentation process control. Background Technology

[0002] 1,3-Propanediol, as an important chemical raw material, is widely used in the synthesis of polymer materials, antifreeze, and various pharmaceuticals. Its production methods include chemical synthesis and microbial fermentation. Currently, industrial production mainly uses chemical synthesis, but this method requires high temperature, high pressure, and catalyst conditions, resulting in harsh operating environments and high costs, which restricts the further development of 1,3-propanediol industrial production. Bio-fermentation mainly uses intestinal bacteria (such as Klebsiella pneumoniae) to produce 1,3-propanediol using glycerol as a substrate; this route is still in the laboratory research stage. Because the 1,3-propanediol fermentation process involves a large number of system state variables reflecting key characteristics of the process operation, these variables determine the implementation of relevant control and monitoring strategies, significantly impacting production safety and economy. However, due to its complex mechanism and strong nonlinear characteristics, the vector dimension of the output variable measurements obtained through sensors often does not match the dimension of the system state variables required to regulate the fermentation process. Therefore, it is necessary to use a state estimator to reconstruct the system state variables based on the available measured values ​​of the output variables and model information, while minimizing the interference of various random signals, thereby obtaining accurate and reliable estimates of the system state variables. Based on the obtained system state variable estimates, the control inputs for the fermenter are then calculated, resulting in a smooth and stable material output.

[0003] While the aforementioned process control methods can improve production efficiency and reduce economic costs, the complexity of the fermentation process, the numerous influencing factors, and the strong coupling between these factors mean that controllers designed using these methods often fail to meet expectations in real-world environments. Therefore, it is crucial to determine whether the designed control method is suitable for the current environment or operating conditions; whether the method performs well in different fermentation processes or under different real-world conditions for the same fermentation process; and whether there is room for improvement in the current controller. This information is essential for truly achieving effective control of the fermentation process.

[0004] Therefore, evaluating the control performance of the 1,3-propanediol fermentation process is of great significance. How to evaluate the control performance of the current fermentation process under actual working conditions is an urgent problem to be solved. Summary of the Invention

[0005] Therefore, embodiments of the present invention provide a method and system for evaluating the optimal control performance of the 1,3-propanediol fermentation process, which solves the problems of insufficient control performance evaluation and unsatisfactory control effects in the prior art.

[0006] To address the aforementioned problems, embodiments of the present invention provide a method for evaluating the optimal control performance of a 1,3-propanediol fermentation process, the method comprising:

[0007] Step S1: Establish a matrix space model of the 1,3-propanediol fermentation process and obtain the measured values ​​of the output variables of the fermentation system;

[0008] Step S2: Based on the information from the matrix space model, calculate the predicted values ​​of the state variables of the fermentation system, design a state variable estimator, and use the state variable estimator to reconstruct the measured values ​​of the output variables and the predicted values ​​of the state variables to obtain the estimated values ​​of the state variables of the fermentation system.

[0009] Step S3: Based on the estimated state variables of the fermentation process and the information from the matrix space model, design the corresponding objective function for control effect, and obtain the control input of the fermentation system by minimizing the objective function;

[0010] Step S4: Based on the information from the matrix space model, approximate the deviation between the current control input and the optimal control input value using partial derivatives;

[0011] Step S5: Based on the deviation between the current control input and the optimal control input value, and in conjunction with the objective function, construct an evaluation index for the optimal performance of the fermentation process control.

[0012] Preferably, the matrix space model of the 1,3-propanediol fermentation process is as follows:

[0013] x t+1 =A t x t +B t u t +w t ;

[0014] y t =C t x t +D t u t +v t ;

[0015] In the formula, t is the sampling time; A t B is the system matrix; t The system input matrix; C t D is the system measurement matrix; t For direct transmission of the matrix; x t For system state variables; ut For system control input; y t The system output variable is obtained directly from the sensor; w t As the process noise that primarily affects the system's state variables, v t The observation noise is the main factor affecting the measured values ​​of the output variable.

[0016] Preferably, the structure of the state variable estimator is represented as follows:

[0017]

[0018] in

[0019]

[0020] In the formula, t is the sampling time; These are estimates of the system state variables; For predicted values ​​of system state variables; L t C is the estimated gain of the state variable estimator; t For the system measurement matrix; y t The system output variable is obtained directly from the sensor measurement. The estimated values ​​of the system state variables from the previous time step; u t-1 The system control input from the previous moment; A t-1 B is the system matrix from the previous time step; t-1 This is the system input matrix from the previous time step.

[0021] Preferably, the objective function for the control effect is:

[0022]

[0023] in

[0024]

[0025]

[0026] In the formula, Let be the objective function for the control effect of the ideal system at time t, which is an estimate of the system's state variables. and system control input u t The function is: Q and R are two weight matrices, which together determine whether to focus more on the accuracy of the system state variables or the magnitude of the system control input in the control effect; K is the control gain.

[0027] Preferably, the deviation between the current control input and the optimal control input value is:

[0028]

[0029] In the formula, Δu t The deviation between the actual system control input and the ideal system control input; I is the identity matrix; Let be the objective function for the control effect of the system at time t, which is an estimate of the system state variables. and system control input u t The function; L t C is the estimated gain of the state variable estimator; t B is the system measurement matrix; t Input matrix for the system; Δ t It is uncertain.

[0030] Preferably, the method for constructing an evaluation index for the optimal control effect of the fermentation process based on the deviation between the current control input and the optimal control input value, combined with the objective function, specifically includes:

[0031] First, quantify Δ t The impact on control performance is the difference between the objective function of the ideal system and the objective function of the actual system along Δ. t Calculate the partial derivative in the direction:

[0032]

[0033] In the formula, LD t The difference between the objective functions of the ideal system and the actual system; For the objective function of the actual system, These are estimates of the actual system state variables. For actual system control input; Δ t It is uncertain;

[0034] Then solve for the control input of the actual system. Along Δ t A method for obtaining evaluation indicators of optimal performance in controlling the fermentation process by directional derivation.

[0035] Preferably, the evaluation index for the optimal performance of the fermentation process control is expressed as:

[0036]

[0037] In the formula, As evaluation indicators; This represents the cumulative value of the objective function of the actual system. Let be the objective function for the control effect at time t, which is an estimate of the system state variables. and system control input u t The function; I is the identity matrix; L tC is the estimated gain of the state variable estimator; t B is the system measurement matrix; t The input matrix is ​​used for the system.

[0038] This invention also provides a system for evaluating the optimal control effect of a 1,3-propanediol fermentation process. This system is used to implement the aforementioned method for evaluating the optimal control effect of a 1,3-propanediol fermentation process, and specifically includes:

[0039] The fermentation process matrix space model establishment module is used to establish a matrix space model of the 1,3-propanediol fermentation process and obtain the measured values ​​of the output variables of the fermentation system.

[0040] The state variable estimation module is used to calculate the predicted values ​​of the state variables of the fermentation system based on the information of the matrix space model, design the state variable estimator, and reconstruct the measured values ​​of the output variables and the predicted values ​​of the state variables using the state variable estimator to obtain the estimated values ​​of the state variables of the fermentation system.

[0041] The control input optimization module is used to design a corresponding control effect objective function based on the estimated state variables of the fermentation process and the information of the matrix space model. The control input of the fermentation system is obtained by minimizing the objective function.

[0042] The control input deviation calculation module is used to approximate the deviation between the current control input and the optimal control input value by using partial derivatives based on information from the matrix space model.

[0043] The control effect evaluation module is used to construct an evaluation index for the optimal control effect of the fermentation process based on the deviation between the current control input and the optimal control input value, combined with the objective function.

[0044] This invention also provides an electronic device, which includes a processor, a memory, and a bus system. The processor and the memory are connected through the bus system. The memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to achieve the above-described method for evaluating the optimal control effect of the 1,3-propanediol fermentation process.

[0045] This invention also provides a computer storage medium storing a computer software product, the computer software product including several instructions for causing a computer device to execute the above-described method for evaluating the optimal control effect of the 1,3-propanediol fermentation process.

[0046] As can be seen from the above technical solutions, this invention application has the following beneficial effects:

[0047] This invention provides a method and system for optimal performance evaluation of the control effect in a 1,3-propanediol fermentation process. First, by designing a matrix space model and a state variable estimator, the state variables of the fermentation system can be accurately estimated, ensuring the precision of the control input. Second, the performance indicators are based on minimizing the objective function, enabling quantitative evaluation of the current controller's performance and facilitating intuitive judgment of its quality. Third, this scheme achieves real-time evaluation, reducing time delays and improving the response speed of the production process. Finally, it requires no specific information about uncertainties and possesses universality, applicable to different environments and fermentation processes, demonstrating its strong adaptability and practicality. These advantages collectively ensure the high efficiency, stability, and reliability of the fermentation process. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the embodiments will be briefly described below. Referring to the accompanying drawings will provide a clearer understanding of the features and advantages of the present invention. The drawings are illustrative and should not be construed as limiting the present invention in any way. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort. Wherein:

[0049] Figure 1 This is a flowchart of a method for evaluating the optimal control effect of a 1,3-propanediol fermentation process, as provided in the embodiments.

[0050] Figure 2 The performance index diagram is shown in the example of the optimal performance evaluation method for controlling the 1,3-propanediol fermentation process.

[0051] Figure 3 This is a block diagram of a system for evaluating the optimal control effect of a 1,3-propanediol fermentation process, as provided in the embodiments. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] Example 1

[0054] like Figure 1 As shown in the embodiments of the present invention, a method for evaluating the optimal control effect of 1,3-propanediol fermentation process is proposed. The method includes:

[0055] Step S1: Establish a matrix space model of the 1,3-propanediol fermentation process and obtain the measured values ​​of the output variables of the fermentation system;

[0056] Step S2: Based on the information from the matrix space model, calculate the predicted values ​​of the state variables of the fermentation system, design a state variable estimator, and use the state variable estimator to reconstruct the measured values ​​of the output variables and the predicted values ​​of the state variables to obtain the estimated values ​​of the state variables of the fermentation system.

[0057] Step S3: Based on the estimated state variables of the fermentation process and the information from the matrix space model, design the corresponding objective function for control effect, and obtain the control input of the fermentation system by minimizing the objective function;

[0058] Step S4: Based on the information from the matrix space model, approximate the deviation between the current control input and the optimal control input value using partial derivatives;

[0059] Step S5: Based on the deviation between the current control input and the optimal control input value, and in conjunction with the objective function, construct an evaluation index for the optimal performance of the fermentation process control.

[0060] As can be seen from the above technical solution, this invention proposes a method and system for evaluating the optimal control performance of a 1,3-propanediol fermentation process. This method can react in real-time and sensitively to the deviation of the current control performance of the 1,3-propanediol fermentation process from the optimal control performance under different working environments, determining whether the current controller is working normally and how far it is from the theoretical optimum. This method is applicable to different fermentation objects and different working environments, making the obtained control performance evaluation results universally applicable and still practical when applied to fermentation process control systems for different objects. Because the performance index proposed in this invention provides a real-time quantitative evaluation of the current control effect based on the theoretically optimal control input value in the actual working environment, it can be directly applied as a compensation term to the input variable, thereby directly improving the control performance of the original system under the influence of unknown disturbances in the actual working environment. Furthermore, since the performance index described in this invention does not require information such as the type and source of the disturbance, this feature ensures its practicality and convenience in different working environments, eliminating the need for additional identification or estimation methods to obtain specific information about unknown disturbances in the actual working environment.

[0061] In this embodiment, in step S1, the ideal matrix space model of the 1,3-propanediol fermentation process is established as follows:

[0062] x t+1 =A t x t +B t u t +w t ;

[0063] y t =C t x t +D t u t +v t ;

[0064] In the formula, t is the sampling time; A t B is the system matrix; t The system input matrix; C t D is the system measurement matrix; t For direct transmission of the matrix; x t For system state variables; u t For system control input; y t The system output variable is obtained directly from the sensor; w t As the process noise that primarily affects the system's state variables, v t The observation noise is the main factor affecting the measured values ​​of the output variable.

[0065] In this embodiment,

[0066]

[0067]

[0068] Therefore, a matrix space model of the 1,3-propanediol fermentation process was established, and sensors were used to obtain the output variable measurement values ​​of the response, namely the carbon dioxide emissions produced by microbial metabolic activities.

[0069] It should be noted that this embodiment uses a bio-fermentation process of 1,3-propanediol to illustrate the present invention. However, the fermentation process applicable to the present invention is not limited to the 1,3-propanediol fermentation process; different fermentation processes can be adapted simply by changing the matrix space model information.

[0070] Furthermore, the established fermentation process model is not limited to a matrix space model. Since a matrix space model can more easily represent the relationship between control inputs and system output variables during the design process, this embodiment primarily uses a matrix space model to represent the system model. Other models can also be applied to this invention, including but not limited to transfer function forms, impulse response function forms, and output sequence forms. To facilitate the evaluation of the control performance using the method described in this invention, it is recommended to first convert other forms of system models into a matrix space model.

[0071] In this embodiment, in step S2, the predicted values ​​of the state variables of the fermentation system are calculated based on the information of the matrix space model, a state variable estimator is designed, and the measured values ​​of the output variables and the predicted values ​​of the state variables are reconstructed using the state variable estimator to obtain the estimated values ​​of the state variables of the fermentation system.

[0072] Specifically, the system state variable estimator uses Kalman filtering to obtain estimates of the system state variables. However, different methods can be chosen to obtain these estimates for different fermentation objects and working environments. These include, but are not limited to, using extended Kalman filtering for systems with strong nonlinearity, using finite impulse response filtering for systems requiring only unbiasedness and desiring to reduce the number of computational iterations, and using robust filtering for systems with higher robustness requirements. However, as long as the representation of the estimated value satisfies the following form, it is applicable to this invention:

[0073]

[0074] In the formula, These are estimates of the system state variables; The predicted values ​​of the system state variables can be calculated using the following formula:

[0075]

[0076] In the formula, u is the estimated value of the system state variables at the previous time step. t-1 A is the system control input from the previous moment. t-1 Let B be the system matrix from the previous time step. t-1 Let be the system input matrix at the previous time step. Since the system in this embodiment is a slowly time-varying system, the system matrix and input matrix at the previous time step can be approximately considered to be the same as those at the current time step, i.e.:

[0077] A t-1 =A t ;

[0078] B t-1 =B t .

[0079] L t In this embodiment, L is the estimated gain of the state variable estimator. t The Kalman filter is used to obtain the result, that is:

[0080]

[0081] y tThis is the system output variable, and its physical meaning changes depending on the fermentation object. In this embodiment, it specifically refers to the carbon dioxide emissions during the fermentation process, which can be collected by a carbon dioxide sensor placed at the gas collection port of the fermenter.

[0082] In this embodiment, in step S3, based on the estimated values ​​of the state variables of the fermentation process and the information of the matrix space model, a corresponding control effect objective function is designed, and the control input of the fermentation system is obtained by minimizing the objective function.

[0083] Specifically, the objective function for controlling the effect takes the form of a quadratic form, namely:

[0084]

[0085] In the formula, Let be the objective function for the control effect of the ideal system at time t, which is an estimate of the system's state variables. and control input u t The objective function is Q. Q and R are two weight matrices that together determine whether to prioritize the accuracy of the system state variables or the magnitude of the system control input in the control effect. In different fermentation processes, depending on the desired control effect, other forms of objective functions can also be used, including but not limited to higher-order norms and nonlinear functions.

[0086] In this embodiment, R = 10.

[0087] Furthermore, control input u t Feedback control is used, that is:

[0088]

[0089] In the formula, K is the control gain. In this embodiment, it is solved by minimizing the objective function, and its specific expression is as follows:

[0090]

[0091] In the formula, For the objective function mentioned above, the equation can be transformed into an algebraic Riccati form, and a substitution method can be used to solve it. However, since infinite time forms are generally considered in practice, the equation can be solved directly from the derived algebraic Riccati equation, i.e.:

[0092] K = [-2.8795 -1.1764 -0.7381] T .

[0093] In this embodiment, in step S4, the deviation between the current control input and the optimal control input value is approximated by partial derivatives based on the information of the matrix space model.

[0094] Specifically, the system matrix space model, i.e. the actual model (system), affected by unknown disturbances and noise, is expressed as follows:

[0095]

[0096]

[0097] In the formula, These are estimates of the actual system state variables. The estimated values ​​of the actual system state variables at the next moment. For actual system control input, For the actual system control input variables. A t B t C t and D t These are the system matrix, system input matrix, system measurement matrix, and direct transfer matrix, respectively. The specific values ​​of these four matrices are consistent with those described above. t and v t The specific data for process noise and observation noise are consistent with those mentioned earlier. Δ t Evidence of compensation for the impact of disturbances and noise on the system output variables, F t Δ t This represents the compensation matrix for the disturbances, i.e., noise, experienced by the system's state variables. Since there is an explicit relationship between the two, it is represented by a single vector Δ. t and a transformation matrix F t This is referred to as uncertainty in the following text.

[0098] Due to the influence of unknown interference and noise, the estimates given by the state variable estimator mentioned above are for the ideal system, not for the actual system. The deviation between these estimates and the best estimates for the actual system's state variables is shown below:

[0099]

[0100] In the formula, Let I be an identity matrix and L be the estimated values ​​of the actual system state variables. t The estimation gain of the state variable estimator, which is calculated above, is Δ. t Due to uncertainty, C t With B t These are the system observation matrix and the system input matrix, which remain consistent with those described earlier. Δu tThe deviation between the actual system control input and the ideal system control input, i.e., the required approximation, is defined as follows:

[0101]

[0102] In the formula, u t For (ideal) system control input, This is the actual system control input. Due to the lack of information about uncertainties, this input cannot be explicitly solved. Its definition is as follows:

[0103]

[0104] In the formula, K represents the estimated state variables of the actual system. δ The control gain of the actual system is calculated in this embodiment by minimizing the objective function, and its specific expression is as follows:

[0105]

[0106] In the formula, Let be the objective function for the control effect at time t, which is an estimate of the system state variables. and system control input u t The function (objective function of the ideal system), This is the objective function of the actual system.

[0107] Furthermore, due to It is the control input that minimizes the objective function of the actual system, so it should satisfy the first principle, that is, the objective function is minimal to a given extent. The partial derivatives are 0:

[0108]

[0109] In the formula, It is the sign of the partial derivative. This represents finding the partial derivative of the numerator along the denominator.

[0110] Further Taylor expansion of the above equation:

[0111]

[0112] In the formula, 0(||Δu k ||) refers to a higher-order infinitesimal, which is ignored because its size is too small compared to the previous terms, and according to Δu k Reorganize the above formula:

[0113]

[0114] At the same time, u tBecause u can be approximated as the control input of an ideal model (system) when the influence of unknown disturbances and noise is small. t and Equivalent. According to first principles, the objective function of the ideal model is equivalent to u. t The derivative should also be 0, that is:

[0115]

[0116] Substituting the above equation into the reorganized expression, and omitting the product of the second derivative terms, it can be further simplified to:

[0117]

[0118] When the influence of unknown interference and noise is small, u can be approximated. t and Equivalently, the second half of the above equation is approximated:

[0119]

[0120] In the formula, and All are continuous, Δu t The latter part can be approximated using the Lagrange mean value theorem, and we can obtain:

[0121]

[0122] Will as well as Substituting the expression into the above formula, we get:

[0123]

[0124] Substituting the above equation into Δu t The expression yields:

[0125]

[0126] Thus, the deviation between the optimal value of the actual model (system) control input and the ideal model (system) control input is approximated by the measured value of the output variable.

[0127] In this embodiment, in step S5, an evaluation index for the optimal control effect of the fermentation process is constructed based on the deviation between the current control input and the optimal control input value, combined with the objective function.

[0128] Specifically, the control performance evaluation index is constructed by quantifying Δ t The impact on control performance is the difference between the objective function of the ideal model (system) and the objective function of the actual model (system) along Δ. tCalculate the partial derivative in the direction of the objective function. Define the difference between the objective function of the ideal model and the actual model as LD. t :

[0129]

[0130] As mentioned above, the form of quantized control performance is the objective function, and the control input is also obtained by minimizing the objective function. Therefore, the aforementioned LD... t This can be understood as the deviation between the control effect of the ideal model's control input on the actual model and the control effect of the actual model's control input on the ideal model; that is, the difference between the objective function of the control input substituted into the actual model and the objective function substituted into the ideal model. Fundamentally, this is due to the influence of unknown disturbances and noise, which cause the theoretical optimal value of the control input of the original ideal model to no longer be the theoretical optimal value of the control input of the current, deviated actual model. The state estimation device and the controller designed earlier cannot perceive or synchronously change the control strategy according to the influence of unknown disturbances and noise, thus leading to a decline in the system's control performance.

[0131] To further investigate the impact of unknown disturbances and noise on the control performance, LD t Along Δ t Differentiating the direction, we get:

[0132]

[0133] because It is the value obtained by substituting the control input of the ideal model into its corresponding objective function, and therefore is not affected by unknown disturbances or noise. Therefore, it is effective for Δ... t The derivative is 0, which further simplifies the above equation to:

[0134]

[0135] Using the chain rule, the above equation can be expressed as:

[0136]

[0137] In this embodiment, the actual model control input is along Δ t Solving for the directional partial derivative requires reviewing Δu again. t The defining formula, namely:

[0138]

[0139] Simultaneously move both sides of the equation along Δ t Differentiating the direction, we get:

[0140]

[0141] Due to ut As the control input for the ideal model, its relationship with Δ t It is irrelevant. Therefore, it follows Δ t Since the differential of direction is 0, the above equation can be further simplified to:

[0142]

[0143] Δu t Substituting the expression, we get:

[0144]

[0145] Therefore, the evaluation index for the optimal performance of the fermentation process control is:

[0146]

[0147] In the formula, As evaluation indicators; This represents the cumulative value of the objective function of the actual system; all other terms are as described above. Substituting the remaining terms obtained above into this expression yields:

[0148]

[0149] The items are as follows:

[0150]

[0151]

[0152]

[0153] In the formula, Q t R represents the relative importance of control accuracy to control performance. t This represents the relative importance of control consumption compared to control performance.

[0154] Substituting the specific numerical values ​​of the model in the example yields:

[0155]

[0156] Ψ t =20.7394;

[0157] Θ t =[0.0036 0.0216 0.2045] T .

[0158] The control performance of the fermentation process at time t is represented by a row number and the system state variable x. t Consistent. And each row corresponds to x.t The control performance exhibited by the system state variables in each row at that moment.

[0159] It is worth noting that, The value evaluates the difference between the control performance of a controller designed based on an ideal model of the 1,3-propanediol fermentation process and the control performance of a controller theoretically designed based on an actual model of the 1,3-propanediol fermentation process. Specifically, when When the value increases, it indicates a significant difference between the currently designed controller and the controller designed based on the actual model; that is, the current control performance is far from the optimal value of the actual model's control performance. Because... An increase in the value indicates that the system is significantly affected by unknown disturbances and noise. In this case, it is advisable to redesign the controller or use the index designed in this invention to compensate for the current control input.

[0160] when When the value decreases, it indicates that the currently designed controller is not significantly different from the controller designed based on the actual model. This means the control performance is gradually recovering, which can also be interpreted as the impact of unknown disturbances and noise on the system gradually weakening. At this point, the original controller can be maintained, or the indicators designed in this paper can still be used to compensate for the control performance, thereby reaching a new equilibrium state more quickly and achieving better control results. Thus, the performance evaluation indicator design is complete, and its evaluation results are as follows: Figure 2 As shown.

[0161] Example 2

[0162] like Figure 3 As shown, this invention provides a performance evaluation system for the optimal control effect of 1,3-propanediol fermentation process. This system is used to implement the performance evaluation method for the optimal control effect of the 1,3-propanediol fermentation process described in Example 1 above, and specifically includes:

[0163] The fermentation process matrix space model establishment module 100 is used to establish a matrix space model of the 1,3-propanediol fermentation process and obtain the measured values ​​of the output variables of the fermentation system.

[0164] The state variable estimation module 200 is used to calculate the predicted values ​​of the state variables of the fermentation system based on the information of the matrix space model, design the state variable estimator, and reconstruct the measured values ​​of the output variables and the predicted values ​​of the state variables using the state variable estimator to obtain the estimated values ​​of the state variables of the fermentation system.

[0165] The control input optimization module 300 is used to design a corresponding control effect objective function based on the estimated state variables of the fermentation process and the information of the matrix space model. The control input of the fermentation system is obtained by minimizing the objective function.

[0166] The control input deviation calculation module 400 is used to approximate the deviation between the current control input and the optimal control input value by using partial derivatives based on information from the matrix space model.

[0167] The control effect evaluation module 500 is used to construct an evaluation index for the optimal control effect of the fermentation process based on the deviation between the current control input and the optimal control input value, combined with the objective function.

[0168] This embodiment provides a system for evaluating the optimal control effect of a 1,3-propanediol fermentation process, used to implement the aforementioned method for evaluating the optimal control effect of a 1,3-propanediol fermentation process. Therefore, the specific implementation of this system can be found in the embodiments section of the method for evaluating the optimal control effect of a 1,3-propanediol fermentation process. For example, the fermentation process matrix space model establishment module 100, the state variable estimation module 200, the control input optimization module 300, the control input deviation calculation module 400, and the control effect evaluation module 500 are respectively used to implement steps S1, S2, S3, S4, and S5 in the aforementioned method for evaluating the optimal control effect of a 1,3-propanediol fermentation process. Therefore, the specific implementation can be referred to the descriptions of the corresponding embodiments. To avoid redundancy, further details are omitted here.

[0169] Example 3

[0170] This invention also provides an electronic device, which includes a processor, a memory, and a bus system. The processor and the memory are connected through the bus system. The memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to achieve the above-described method for evaluating the optimal control effect of the 1,3-propanediol fermentation process.

[0171] Example 4

[0172] This invention also provides a computer storage medium storing a computer software product, the computer software product including several instructions for causing a computer device to execute the above-described method for evaluating the optimal control effect of the 1,3-propanediol fermentation process.

[0173] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0174] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0175] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0176] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for evaluating the optimal performance of 1,3-propanediol fermentation process control, characterized in that, include: Step S1: Establish a matrix space model of the 1,3-propanediol fermentation process and obtain the measured values ​​of the output variables of the fermentation system; Step S2: Based on the information from the matrix space model, calculate the predicted values ​​of the state variables of the fermentation system, design a state variable estimator, and use the state variable estimator to reconstruct the measured values ​​of the output variables and the predicted values ​​of the state variables to obtain the estimated values ​​of the state variables of the fermentation system. Step S3: Based on the estimated state variables of the fermentation process and the information from the matrix space model, design the corresponding objective function for control effect, and obtain the control input of the fermentation system by minimizing the objective function; Step S4: Based on the information from the matrix space model, approximate the deviation between the current control input and the optimal control input value using partial derivatives; Step S5: Based on the deviation between the current control input and the optimal control input value, and in conjunction with the objective function, construct an evaluation index for the optimal performance of the fermentation process control.

2. The method for evaluating the optimal control effect of the 1,3-propanediol fermentation process according to claim 1, characterized in that, The matrix space model of the 1,3-propanediol fermentation process is as follows: x t+1 =A t x t +B t u t +w t ; y t =C t x t +D t u t +v t ; In the formula, t is the sampling time; A t B is the system matrix; t C is the system input matrix; t D is the system measurement matrix; t For direct transmission of the matrix; x t For system state variables; u t For system control input; y t The system output variable is obtained directly from the sensor; w t As the process noise that primarily affects the system's state variables, v t The observation noise is the main factor affecting the measured values ​​of the output variable.

3. The method for evaluating the optimal control effect of the 1,3-propanediol fermentation process according to claim 1, characterized in that, The structure of the state variable estimator is represented as follows: in In the formula, t is the sampling time; These are estimates of the system state variables; For predicted values ​​of system state variables; L t C is the estimated gain of the state variable estimator; t For the system measurement matrix; y t The system output variable is obtained directly from the sensor measurement. The estimated values ​​of the system state variables from the previous time step; u t-1 The system control input from the previous moment; A t-1 B is the system matrix from the previous time step; t-1 This is the system input matrix from the previous time step.

4. The method for evaluating the optimal control effect of the 1,3-propanediol fermentation process according to claim 1, characterized in that, The objective function for the control effect is: in In the formula, Let be the objective function for the control effect of the ideal system at time t, which is an estimate of the system's state variables. and system control input u t The function is: Q and R are two weight matrices, which together determine whether to focus more on the accuracy of the system state variables or the magnitude of the system control input in the control effect; K is the control gain.

5. The method for evaluating the optimal control effect of the 1,3-propanediol fermentation process according to claim 1, characterized in that, The deviation between the current control input and the optimal control input value is: In the formula, Δu t The deviation between the actual system control input and the ideal system control input; I is the identity matrix; Let be the objective function for the control effect of the system at time t, which is an estimate of the system state variables. and system control input u t The function; L t C is the estimated gain of the state variable estimator; t B is the system measurement matrix; t Input matrix for the system; Δ t It is uncertain; It is the sign of the partial derivative; This represents finding the partial derivative of the numerator along the denominator.

6. The method for evaluating the optimal control effect of the 1,3-propanediol fermentation process according to claim 1, characterized in that, The method for constructing an evaluation index for the optimal control effect of the fermentation process based on the deviation between the current control input and the optimal control input value, combined with the objective function, specifically includes: First, quantify Δ t The impact on control performance is the difference between the objective function of the ideal system and the objective function of the actual system along Δ. t Calculate the partial derivative in the direction: In the formula, LD t The difference between the objective functions of the ideal system and the actual system; For the objective function of the actual system, These are estimates of the actual system state variables. For actual system control input; Δ t It is uncertain; It is the sign of the partial derivative; This represents finding the partial derivative of the numerator along the denominator. Then solve for the control input of the actual system. Along Δ t A method for obtaining the evaluation index of the optimal performance of fermentation process control using directional partial derivative I.

7. The method for evaluating the optimal control effect of the 1,3-propanediol fermentation process according to claim 6, characterized in that, The evaluation index for the optimal performance of the fermentation process control is expressed as follows: In the formula, As evaluation indicators; This represents the cumulative value of the objective function of the actual system. Let be the objective function for the control effect at time t, which is an estimate of the system state variables. and system control input u t The function; I is the identity matrix; L t C is the estimated gain of the state variable estimator; t B is the system measurement matrix; t The input matrix is ​​used for the system.

8. A performance evaluation system for optimal control of 1,3-propanediol fermentation process, characterized in that, The system is used to implement the optimal performance evaluation method for controlling the 1,3-propanediol fermentation process as described in any one of claims 1 to 7, specifically including: The fermentation process matrix space model establishment module is used to establish a matrix space model of the 1,3-propanediol fermentation process and obtain the measured values ​​of the output variables of the fermentation system. The state variable estimation module is used to calculate the predicted values ​​of the state variables of the fermentation system based on the information of the matrix space model, design the state variable estimator, and reconstruct the measured values ​​of the output variables and the predicted values ​​of the state variables using the state variable estimator to obtain the estimated values ​​of the state variables of the fermentation system. The control input optimization module is used to design a corresponding control effect objective function based on the estimated state variables of the fermentation process and the information of the matrix space model. The control input of the fermentation system is obtained by minimizing the objective function. The control input deviation calculation module is used to approximate the deviation between the current control input and the optimal control input value by using partial derivatives based on information from the matrix space model. The control effect evaluation module is used to construct an evaluation index for the optimal control effect of the fermentation process based on the deviation between the current control input and the optimal control input value, combined with the objective function.

9. An electronic device, characterized in that, The electronic device includes a processor, a memory, and a bus system. The processor and the memory are connected through the bus system. The memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to achieve the optimal performance evaluation method for controlling the 1,3-propanediol fermentation process as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The computer storage medium stores a computer software product, which includes several instructions for causing a computer device to execute the optimal performance evaluation method for the 1,3-propanediol fermentation process control as described in any one of claims 1 to 7.

Images