Method and System for Predicting Microbial Culture Concentration

Abstract

The present invention relates to a system and method for predicting a microbial culture production amount using a metabolic model equation that controls a metabolic model, the system and method comprising: calculating a predicted concentration of a substrate from the metabolic model equation; calculating an actual concentration of the substrate from spectral data of a culture solution; and comparing the predicted concentration of the substrate with the actual concentration of the substrate to update the parameters of the metabolic model equation, thereby improving accuracy in predicting a microbial culture production amount using the metabolic model equation.

Description

[Technical field]

[0001] The present invention relates to a method and system for predicting the culture concentration of a microorganism in real time during the process of culturing the microorganism. [Background technology]

[0002] The recent global trend of emphasizing environmentally friendly production has led to an increase in the production of bioplastics, which are made from recyclable raw materials rather than monomers based on fossil fuels such as petroleum and natural gas.

[0003] Bioplastics have the advantage that they emit less carbon dioxide during the production process than petroleum-based plastics, and depending on the material and product, they can decompose naturally within a few years, eliminating the waste problem.

[0004] To produce bioplastics, the microorganisms must be grown in an optimal environment, the number of microorganisms must be increased through cultivation, and the bioplastic raw materials must be produced through cellular metabolism. In this process, it is essential to create an optimal growth environment for the microorganisms to grow and metabolize substances.

[0005] Various conditions must be considered depending on the type of microorganism, such as the concentration of the substrate used as microbial food, oxygen concentration, feed injection method, temperature, pH, etc., and traditionally, these conditions have been mainly discovered through experiments. However, the inefficiency of adjusting these process variables through actual experiments, which requires a lot of time and cost, has been raised as an issue.

[0006] Under these circumstances, in order to measure the concentration of microorganisms during the culture process, there are methods in which a metabolic model equation is used to estimate the parameters of the metabolic model equation through culture experiments, or an arbitrary initial value is set and the concentration of microorganisms during the process is predicted from the metabolic model equation.

[0007] However, such conventional techniques have the problem that accurate estimation cannot be performed due to changes in conditions during the process, and there is currently a demand for a method to improve such problems.

[0008] The following patent documents describe related prior art: [Prior art documents] [Patent documents]

[0009] [Patent Document 1] Republic of Korea Patent Application No. 10-2020-7013036 [Patent Document 2] Republic of Korea Patent Application No. 10-2020-7004943 Summary of the Invention [Problem to be solved by the invention]

[0010] The present invention has been devised to solve the above-mentioned problems, and has an object to provide a method and system for predicting microbial concentration information in real time during a microbial culture process. [Means for solving the problem]

[0011] In order to solve the above-mentioned problems, the present invention provides a system for predicting a microbial culture production amount, comprising: an incubator for culturing microorganisms in a culture solution; a spectrum acquisition unit for acquiring a real-time spectrum of the culture solution; a memory device storing a concentration actual measurement model which is an artificial intelligence model trained to calculate an actual measured substrate concentration from the spectrum data; and a metabolic model equation which specifies the metabolic activity of the microorganism; and a control unit for calculating a predicted value of a microbial production amount from the concentration actual measurement model and the metabolic model equation, wherein the control unit receives spectral data transferred from the spectrum acquisition unit, inputs the spectral data into the concentration actual measurement model to calculate an actual measured substrate concentration, calculates a predicted substrate concentration from the metabolic model equation, updates parameters of the metabolic model equation from data on the actual measured substrate concentration and the predicted substrate concentration, and calculates a predicted value of a microbial production amount from the metabolic model equation with updated parameters.

[0012] The present invention also provides a microbial culture concentration prediction device comprising a spectroscopic measurement unit that acquires spectral data of the culture solution in a microbial culture vessel, and a microbial production amount prediction unit that inputs the spectral data into a concentration actual measurement model to calculate an actual substrate concentration in the culture solution, and updates parameters of a microbial metabolic model using the actual substrate concentration to predict the production concentration of the microorganism. In this case, the microbial production amount prediction unit comprises a substrate concentration actual measurement unit that calculates an actual substrate concentration from the spectral data, a metabolic model unit that calculates a predicted substrate concentration from a metabolic model equation that defines the metabolic activity of the microorganism, and a parameter update unit that updates parameters of the metabolic model equation from the actual substrate concentration and the predicted substrate concentration, and the metabolic model unit may comprise a microbial production amount prediction unit that applies parameters updated in the parameter update unit to the metabolic model equation and calculates a predicted value of the microbial culture concentration from the metabolic model equation to which the updated parameters have been applied.

[0013] Furthermore, the substrate concentration measurement unit includes a concentration measurement model that is machine-learned to calculate the substrate concentration from the spectrum data, and the real-time spectrum data is input to the concentration measurement model to calculate the actual substrate concentration. In this case, the metabolic model formula includes a concentration calculation model formula for predicting the concentration of E. coli cells in the culture solution and a substrate concentration calculation model formula for calculating the concentration of glucose, which is a substrate.

[0014] Furthermore, the present invention provides a method for predicting the amount of microbial production in a culture process, the method including: a spectral data acquisition step for acquiring a real-time spectrum of the culture solution in a microbial culture vessel; an actual substrate concentration calculation step for inputting the real-time spectral data into a previously learned concentration measurement model to calculate the actual substrate concentration in the culture solution; a predicted substrate concentration calculation step for calculating a predicted substrate concentration from the microbial metabolic model equation; a parameter updating step for comparing the actual substrate concentration calculated in the actual substrate concentration calculation unit with the predicted substrate concentration calculated in the metabolic model unit, and updating parameters of the metabolic model equation so that the predicted substrate concentration matches the actual substrate concentration; and a microbial production amount prediction step for predicting the amount of microbial production from the metabolic model equation to which the updated parameters have been applied.

[0015] In this case, the predicted substrate concentration calculation step is characterized in that parameters of a metabolic model formula are estimated from concentration data of each component in the culture solution that has already been obtained, and the predicted substrate concentration is calculated from the metabolic model formula by applying the estimated parameters, and the metabolic model formula is characterized in that it includes a concentration calculation model formula for predicting the concentration of E. coli cells in the culture solution, and a substrate concentration calculation model formula for calculating the concentration of glucose as a substrate. Effect of the Invention

[0016] The present invention has the effect of improving the accuracy of calculation of the culture concentration of a microorganism by updating the parameters of the metabolic model formula using the real-time substrate concentration when calculating the culture concentration of a microorganism according to the metabolic model formula. [Brief description of the drawings]

[0017] [Figure 1] FIG. 1 is a block diagram of a microbial concentration prediction system according to the present invention. [Diagram 2] FIG. 2 is a diagram illustrating each step of the prediction of a microbial concentration according to the present invention. [Diagram 3] This is a metabolic model showing the metabolic relationships of E. coli cultivation and production. [Figure 4] FIG. 1 is a diagram showing predicted substrate concentration values ​​and actually measured substrate concentration values ​​in an embodiment of the present invention. [Diagram 5] FIG. 13 is a diagram showing predicted concentration values ​​of each component according to a metabolic model equation with updated parameters according to the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0018] Hereinafter, the embodiments of the present invention will be described in detail with reference to the accompanying drawings so that a person having ordinary skill in the art to which the present invention pertains can easily carry out the present invention. However, the present invention can be embodied in various different forms, and is not limited to the embodiments described herein. In the drawings, in order to clearly explain the present invention, parts that are not related to the description are omitted, and similar parts are denoted by similar reference numerals throughout the specification.

[0019] In the present invention, learning or training means training a computer algorithm that has a predetermined output value for an input value in the technical field of artificial intelligence, or a neural network in which such an algorithm is realized in the form of a neural network, and refers to one of the processes of performing a neural network algorithm by a computer calculation device. Also, in the present invention, a predetermined "generation of a model" means training the neural network before learning or a computer algorithm that realizes it to generate a computer algorithm that has completed learning so as to generate a desired output value for an input value.

[0020] The present invention relates to a method and system for predicting microbial culture concentration in an incubator.

[0021] More specifically, the present invention relates to a prediction of the culture concentration of a microorganism whose metabolic pathway has been identified through genetic manipulation, and includes a spectrometer that supplies a substrate used as food for the microorganism to a culture vessel and obtains a real-time spectrum of the substrate in the culture solution. The present invention also includes a computer algorithm that predicts the substrate concentration and the microorganism concentration using a metabolic model formula based on the identified metabolic pathway, and includes a previously trained machine learning algorithm that can calculate the actual concentration of the substrate by learning the real-time spectral data of the spectrometer. The present invention includes a method and system that compares the calculated actual concentration of the substrate with a predicted value of the substrate concentration using the metabolic model formula, updates parameters of the metabolic model formula using an optimization algorithm, and applies the updated parameters to the metabolic model formula to predict the microorganism concentration.

[0022] According to the present invention, when predicting the concentration of a microorganism based on a metabolic model equation, the parameter setting values ​​are fixed to initial setting values, the substrate concentration is measured in real time using spectral data, and the parameters of the metabolic model equation are updated based on this, thereby providing a more accurate concentration prediction technique based on the metabolic model equation.

[0023] The present invention will now be described in detail with reference to the accompanying drawings.

[0024] 1. Prediction system for microbial culture concentration according to the present invention The system for predicting a microbial culture concentration according to the present invention will be described with reference to FIG.

[0025] The prediction system for microbial culture concentration of the present invention includes a microbial culture vessel 10, a spectrometer 20 that acquires real-time spectral data of the culture solution in the microbial culture vessel, a substrate concentration actual measurement unit 31 that inputs the real-time spectral data into a concentration prediction model that has already been learned to calculate the actual substrate concentration in the culture solution, a metabolic model unit 32 that is equipped with the microbial metabolic model and calculates the substrate concentration according to the metabolic model, and a parameter calculation unit 33 that compares the actual substrate concentration calculated in the actual substrate concentration calculation unit with the predicted substrate concentration calculated in the metabolic model unit, and calculates and updates parameters of the metabolic model of the metabolic model unit so that the predicted substrate concentration matches the actual substrate concentration, and a microbial production amount prediction unit 30 that applies the updated parameters to predict the production amount of the microorganisms from the metabolic model with the updated parameters.

[0026] In an embodiment of the present invention, the microorganism may refer to Escherichia coli.

[0027] 1.1.Cultivation Department 10 The microbial culture unit of the present invention comprises, in addition to each of the components of a normal incubator, a dissolved oxygen sensor 11 that measures the concentration of dissolved oxygen in the culture chamber where the culture is carried out, a feed amount measurement unit 12 that measures the feed amount of glucose, and a volume measurement unit 13 that measures the volume of the culture solution.

[0028] (1) Dissolved oxygen sensor 11 The dissolved oxygen sensor measures the concentration of dissolved oxygen in the culture chamber and transmits the result to the control unit.

[0029] (2) Feed amount measuring unit 12 The feed amount of glucose supplied to the culture chamber is transmitted to the control unit.

[0030] (3) Volume measurement section 13 The volume of the culture medium in the culture chamber is measured and transmitted to the control unit.

[0031] 1.2.Spectrometry section 20 This component is for acquiring spectral data of the culture solution in the incubator, and the spectral data of the culture solution is acquired using a known spectrometer.

[0032] 1.3. Microbial production amount prediction unit 30 The microbial production amount prediction unit calculates parameters of a metabolic model formula that defines a microbial metabolic model from the measured substrate concentration calculated from the spectral data, and predicts the microbial concentration from the microbial metabolic model formula by applying the calculated parameters, thereby predicting the microbial production amount. The microbial production amount prediction unit outputs the production amount as the predicted microbial concentration, and uses this to adjust and output the process control variables of the incubator.

[0033] The microbial production amount prediction unit 30 may physically comprise a memory device for storing the concentration measurement model, metabolic model equation, and optimization algorithm, and a control unit processor for performing calculations such as calculating the actual measured substrate concentration, calculating the microbial concentration, and calculating / updating parameters from these.

[0034] (1) Substrate concentration measurement unit 31 This is a component that measures the actual substrate concentration from the spectrum data of the culture solution. The substrate concentration measurement unit is equipped with a concentration measurement model that has been machine-learned to calculate the substrate concentration from the spectrum data of the culture solution, and calculates the actual substrate concentration from this.

[0035] The substrate concentration measurement unit is equipped with a known prediction model such as a partial least squares (PLS) model that calculates the substrate concentration from the spectrum data of the culture solution as a concentration measurement model, and calculates the substrate concentration from the input spectrum data. In the present invention, the substrate concentration value calculated from the spectrum data using the concentration measurement model is defined as the measured substrate concentration.

[0036] (2) Metabolic Model Section 32 The microbial production amount prediction unit 30 includes a metabolic model unit 32. The metabolic model unit is equipped with a computer-executable algorithm that calculates the concentration of a substrate according to a metabolic model formula that defines the metabolic activity of microbial culture, as a metabolic model formula. The substrate concentration calculation model formula as the metabolic model formula loaded in the metabolic model unit includes predetermined parameters and a material balance equation expressed by a differential equation, and the concentration of each component of the culture solution calculated in the metabolic model unit is defined as a predicted concentration in the present invention.

[0037] The metabolic model section outputs two types of output values.

[0038] First, the first step estimates the parameters of the metabolic model equation from the concentration values ​​of each component obtained through the culture experiment, and outputs a predicted substrate concentration value calculated by applying the estimated parameters to the substrate concentration calculation model equation. The predicted substrate concentration value is compared with the actually measured substrate concentration in the parameter update unit described later and is applied to update the parameters.

[0039] The second is to receive updated parameters from the parameter update unit described later and apply them to the microbial cell concentration model equation of the metabolic model equation, and calculate and output a predicted value of the microbial concentration using the microbial cell concentration model equation having the updated parameters as the amount of microbial production.

[0040] (3) Parameter update unit 33 The parameter update unit updates each parameter of the metabolic model formula using the calculated actual substrate concentration and predicted substrate concentration, and transfers the updated parameters to the metabolic model unit. The parameters of the metabolic model formula are updated so that the actual substrate concentration value and the predicted substrate concentration value match as closely as possible, and the details of the method will be described later in the section on the method of predicting the microbial culture concentration of the present invention.

[0041] 2. Method for predicting microbial culture concentration in culture medium according to the present invention The method for predicting the microbial culture concentration in a culture solution according to the present invention is composed of steps performed by each component of the above-mentioned system for predicting the microbial culture concentration. Each step will be described with reference to FIG.

[0042] (1) Real-time spectrum data acquisition step (S10) This is a step in which the spectroscopic measurement unit 20 acquires real-time spectral data from the culture solution in the culture unit.

[0043] (2) Calculation step of measured substrate concentration (S20) This is a step in which the acquired real-time spectrum data is input to the substrate concentration measurement unit 31 to measure the substrate concentration. The real-time spectrum data is input to the already learned concentration measurement model of the substrate concentration measurement unit 31, and the actual substrate concentration value is output.

[0044] (3) Predicted substrate concentration calculation step (S30) The predicted substrate concentration is calculated according to the substrate concentration metabolic model equation installed in the metabolic model section. The predicted substrate concentration is calculated by substituting the concentration data of each component in the culture solution already obtained through the culture experiment described below into the metabolic model equation consisting of a differential equation and a rate equation, and estimating the parameters of the metabolic model equation using a known optimization technique.

[0045] Known optimization algorithms include least squares method & weighted least squares method, gradient descent search method, Newton method (Newton-Raphson method), Gauss-Newton method, Levenberg-Marquardt method, genetic algorithm, ant colony optimization (ACO), simulated annealing (SA), memetic algorithm (MA), evolutionary computing method, evolutionary strategy method, evolutionary programming, etc., and the present invention should not be construed as being limited to application to any particular algorithm.

[0046] After that, the estimated parameters are substituted into the metabolic model equation, and the predicted substrate concentration is calculated from the substrate concentration metabolic model equation.

[0047] (4) Parameter update step (S40) This is a procedure for updating the parameter values ​​of the metabolic model formula in the metabolic model section 32 using the actually measured substrate concentration value and the calculated predicted substrate concentration value.

[0048] The parameter update is a step of comparing the time-series actual measured substrate concentration values ​​with the predicted substrate concentration values, and updating each parameter value of the metabolic model equation estimated in the previous predicted substrate concentration calculation step by applying a known optimization algorithm such as a genetic algorithm so that the predicted substrate concentration according to the metabolic model and the actual measured substrate concentration optimally match.

[0049] (4) Microbial production amount prediction step (S50) This is a procedure for outputting a predicted value of the microbial production amount by calculating the microbial concentration from the microbial cell concentration metabolic model equation with updated parameter values.

[0050] <Example> An embodiment of the present invention will now be described.

[0051] (1) Acquisition of spectral data and calculation of actual concentration First, the spectral data acquired in a time series from the spectrum data obtained from the incubator is input to an actual substrate concentration calculation model that has undergone machine learning using a known PLS model or the concentration value of the substrate in the culture medium obtained through a culture experiment and the spectral data as learning data, and the actual substrate concentration in a time series is calculated.

[0052] The culture experiment was carried out under the following culture experimental conditions.

[0053] <Culture experimental conditions for parameter estimation and training of the model for calculating measured substrate concentrations> Strain: E. Coli W3110 Initial dry cell weight (DCW): 0.1887g / L Initial substrate concentration: 20g / L Culture temperature: 35℃, Incubator rotation speed: 500-900 rpm Culture solution pH: 6.95 (22.2% NH 4 OH (ammonia water: pH adjusting solution) Feed rate: Continuous feed at 50ml / h after 10hr (no feed before 10hr) Feed solution: glucose 700g / L, MgSO 4 (Magnesium sulfate) 15g / L, trace metal solution 10ml / L

[0054] (2) Calculation of metabolic model parameters and predicted substrate concentrations <Metabolic model formula applied to the present invention> In an embodiment of the present invention, an E. coli metabolic model formula is loaded into the metabolic model section, and the metabolic model formula includes an E. coli cell concentration calculation model formula (Formula 1) in the culture medium, a substrate concentration calculation model formula (Formula 2) for calculating the concentration of the substrate glucose, a model formula for calculating the concentration of the produced acetate (Formula 3), and a dissolved oxygen concentration calculation model formula (Formula 4) for calculating the amount of dissolved oxygen, and is composed of a balance equation and a rate equation for calculating the concentrations of E. coli, glucose, and acetate, and the amount of dissolved oxygen.

[0055] The E. coli metabolic model equation applied in the examples of the present invention is based on the macroscopic kinetic model ( FIG. 3( a )) and acetate cycling system model ( FIG. 3( b )) related to E. coli cultivation proposed in Modeling overflow metabolism in Escherichia coli by acetate cycling, Biochemical Engineering Journal 125 (2017) 23-30, Emmanuel Anane, et al.

[0056]

number

[0057]

number

[0058]

number

[0059]

number

[0060] (X is the concentration of E. coli cells, S is the concentration of the substrate glucose, A is the concentration of acetate, F is the feed, V is the volume, μ is the growth rate constant, q sox is the rate of glucose uptake through oxidative metabolism, q sof is the rate of glucose uptake through excess metabolism, q sA is the rate of glucose uptake through acetate metabolism, q m is the cell retention constant, Y em is the yield excluding cell retention, Y xsof is the yield of the cell / glucose over-metabolism pathway, Y xa is the production yield of acetate / cell, q smax is the maximum glucose uptake rate constant, K ia , K s are the inhibition constants of glucose uptake by acetate, glucose uptake by acetate, and q sox is the rate of glucose uptake through oxidative metabolism, q sof is the rate of glucose uptake through excess metabolism, P Amax is the maximum acetate formation rate constant, K O is the affinity constant for oxygen consumption, K ap is the saturation constant for intracellular acetate production (monod type), q A is the consumption rate of acetate, p A is the rate of acetate formation, q sA is the rate of glucose uptake through acetate metabolism, q sof is the rate of glucose uptake through excess metabolism, Y as is the yield of acetate produced through excess metabolism (acetate / glucose), q Amax is the maximum acetate uptake rate constant, K is is the inhibition constant of acetate uptake by glucose, K sa is the affinity constant for acetate consumption, DOT is the oxygen saturation (%) = DO a / DO * 、 DO a is the concentration of dissolved oxygen (mg / L), DO *is the saturated dissolved oxygen concentration (mg / L) of the culture medium at the given process conditions, K La is the oxygen transfer coefficient, q O is the rate of oxygen consumption, H is the Henry's law constant, Y os is the glucose / oxygen consumption ratio (yield), Y oa is the acetate / oxygen production yield, q sox is the rate of glucose uptake through oxidative metabolism, q m indicates the cell retention constant.)

[0061] As another embodiment, in the above formula, in formula 3,

number

number

[0062] <Calculation of metabolic model parameters> The data obtained through the culture experiment was substituted into the E. coli metabolic model equation, and the genetic algorithm, one of the known optimization techniques, was applied to estimate the parameter values. The estimated parameter values ​​are shown in the table below.

[0063] [Table 1]

[0064] <Calculation of predicted substrate concentration> The estimated parameters were applied to Equation 2 above to calculate the predicted substrate concentration.

[0065] (3) Updating parameters The parameters of the metabolic model formula were updated by comparing the measured substrate concentration obtained from the spectrum data of the culture solution with the predicted substrate concentration calculated from the metabolic model formula. Among known optimization algorithms, a genetic algorithm was applied to update the parameters of the metabolic model formulas of Formulas 1 to 4 in the metabolic model section so that the difference between the measured substrate concentration and the substrate concentration according to the substrate metabolic model formula of Formula 2 was minimized.

[0066] FIG. 4 is a graph showing predicted substrate concentration values ​​Estimated_S using parameters estimated from culture experiment data and actually measured substrate concentration values ​​Measured_S obtained from spectrum data.

[0067] (4) Calculating the predicted concentration of E. coli cells After updating the parameter values, the E. coli cell concentration was predicted from the cell concentration metabolic model equation with updated parameters. Figure 5 is a graph showing the predicted substrate concentration Predicted_S, predicted acetate concentration Predicted_A, and predicted E. coli cell concentration Predicted_X predicted from the metabolic model equation with updated parameters.

[0068] The names of the various parts used in the description of the present invention and in the drawings are given below. [Explanation of symbols]

[0069] 10...Culture department 20…Spectrometry section 30...Production amount prediction section 31...Substrate concentration measurement section 32…Metabolic model section 33…Parameter update section

Claims

1. A spectroscopic measurement unit that acquires spectral data of the culture solution in the microorganism incubator; a microbial production amount prediction unit that inputs the spectrum data into a concentration measurement model to calculate an actual substrate concentration in the culture solution, updates parameters of a microbial metabolic model using the actual substrate concentration, and predicts the production concentration of the microorganism; A device for predicting microbial culture concentration comprising:

2. The microbial production amount prediction unit is a substrate concentration measuring unit for calculating an actual substrate concentration from the spectrum data; A metabolic model unit that calculates a predicted substrate concentration from a metabolic model equation that defines the metabolic activity of the microorganism; a parameter update unit that updates parameters of the metabolic model formula based on the actual substrate concentration and the predicted substrate concentration; The apparatus for predicting a microbial culture concentration according to claim 1 , comprising:

3. The metabolic model part comprises: The prediction device for microbial culture concentration according to claim 2, further comprising a microbial production amount prediction unit that applies the parameters updated in the parameter update unit to the metabolic model formula and calculates a predicted value of the microbial culture concentration from the metabolic model formula to which the updated parameters have been applied.

4. The substrate concentration measurement unit includes: A concentration measurement model machine-learned to calculate the concentration of the substrate from the spectral data; 3. The apparatus for predicting a microbial culture concentration according to claim 2, wherein the real-time spectrum data is input to the concentration measurement model to calculate an actual measured substrate concentration.

5. The metabolic model formula is:

3. The apparatus for predicting a microbial culture concentration according to claim 2, comprising a concentration calculation model equation for predicting a concentration of E. coli cells in a culture solution, and a substrate concentration calculation model equation for calculating a concentration of glucose as a substrate.

6. A spectral data acquisition step of acquiring a real-time spectrum of the culture medium of the microorganism incubator; an actual substrate concentration calculation step of inputting the real-time spectrum data into a previously trained concentration measurement model to calculate an actual substrate concentration in the culture solution; A predicted substrate concentration calculation step of calculating a predicted substrate concentration of the substrate from the microbial metabolism model formula; a parameter updating step of comparing the actual substrate concentration calculated in the actual substrate concentration calculation unit with the predicted substrate concentration calculated in the metabolic model unit, and updating parameters of the metabolic model formula so that the predicted substrate concentration coincides with the actual substrate concentration; A microbial production amount prediction step of predicting the production amount of a microbial organism from the metabolic model formula to which the updated parameters are applied; A method for predicting the amount of microorganism produced in a culture process, comprising:

7. The predicted substrate concentration calculation step includes: The parameters of the metabolic model are estimated from the concentration data of each component in the culture medium that has already been obtained.

7. The method for predicting the amount of microbial production in a culture process according to claim 6, characterized in that the estimated parameters are applied to calculate a predicted substrate concentration from the metabolic model equation.

8. The metabolic model formula is:

8. The method for predicting the amount of microbial production in a culture process according to claim 7, comprising: a concentration calculation model equation for predicting the concentration of E. coli cells in a culture solution; and a substrate concentration calculation model equation for calculating the concentration of glucose as a substrate.

9. An incubator for culturing microorganisms in a culture solution; A spectrum acquisition unit for acquiring a real-time spectrum of the culture solution; A concentration measurement model, which is an artificial intelligence model trained to calculate an actual substrate concentration from spectral data, and a memory device storing a metabolic model formula that defines the metabolic activity of the microorganism; A control unit that calculates a predicted value of the amount of microbial production from the concentration measurement model and a metabolic model formula; It is equipped with The control unit is A system for predicting a microbial culture production amount, characterized in that spectral data is transferred from the spectrum acquisition unit and input into the concentration measurement model to calculate an actual substrate concentration, a predicted substrate concentration is calculated from the metabolic model equation, parameters of the metabolic model equation are updated from data on the actual substrate concentration and the predicted substrate concentration, and a predicted value of the microbial production amount is calculated from the metabolic model equation with updated parameters.

Images