Traditional Chinese medicine multi-strain synergistic fermentation control system

Abstract

The application discloses a traditional Chinese medicine multi-strain synergistic fermentation control system, and belongs to the technical field of fermentation control. The system comprises a tank body, a motor, a stirring main shaft, multi-layer stirring paddles, a temperature sensor and an electrode sensor, and is provided with a controller. The controller is provided with a metabolic intensity fusion calculation unit, a substrate consumption rate and a temperature weight are used to obtain a metabolite transient concentration weighted model; a synergistic entropy calculation unit is used to calculate a synergistic index based on pH, dissolved oxygen and the weighted model, and the synergistic index represents the synergistic degree of the bacterial flora; a coupling regulation factor calculation unit is used to calculate a coupling regulation factor by combining the rotation speed, the temperature and the synergistic index, so as to represent the support ability of the mass transfer condition to the synergistic fermentation. The system compares the factor with a target interval, automatically adjusts the rotation speed or the temperature, and iteratively optimizes until convergence. The application realizes real-time quantification of the multi-strain synergistic state and mass transfer, metabolic coupling regulation, and improves the controllability of the fermentation process and the product quality.

Description

Technical Field

[0001] This invention relates to the field of fermentation control technology, and in particular to a multi-strain synergistic fermentation control system for traditional Chinese medicine. Background Technology

[0002] Multi-strain co-fermentation of traditional Chinese medicine (TCM) refers to the use of two or more microbial strains (such as yeast, lactic acid bacteria, and Bacillus) in a closed fermentation vessel to work synergistically on the TCM substrate. Through metabolic synergy and interaction among the strains, the biotransformation and enhancement of the active ingredients in the TCM are achieved. Compared with single-strain fermentation, multi-strain co-fermentation can more fully decompose the macromolecular components of TCM, transform toxic substances, and produce a richer variety of secondary metabolites, thus showing broad application prospects in the modern production of TCM.

[0003] However, existing multi-strain synergistic fermentation control systems for traditional Chinese medicine have the following problems during the fermentation process: First, existing fermentation control systems use independent PID control with a single parameter. However, the essence of multi-strain fermentation of traditional Chinese medicine is that multiple strains share the same environment, resulting in complex metabolic competition and synergistic relationships. Operators judge the state of the microbial community by periodically sampling and offline detecting the biomass and metabolite concentrations of the strains, which leads to a delay of several hours in the detection results, missing the optimal control window, and thus making it difficult to perceive the microbial community relationship parameters. Second, in the multi-strain fermentation process, an intermediate product of one strain may become a key substrate for another strain, and the accumulation of a certain intermediate product may indicate an imbalance in metabolic pathways. However, if existing fermentation control systems focus on the final product indicators and macroscopic environmental parameters while ignoring the online monitoring of intermediate metabolites, they only see the local manifestations of fermentation and cannot see the metabolic coupling relationships occurring within the microbial community. Therefore, it is difficult to use the information on inter-strain interactions carried by intermediate products for control decisions. In addition, in existing technologies, the stirring speed and fermentation temperature are managed by two independent control loops. However, temperature not only directly affects the enzyme activity of the microbial community, but also indirectly affects the mass transfer efficiency by changing the viscosity of the fermentation broth; stirring not only affects the mixing uniformity and gas-liquid mass transfer, but also changes the apparent viscosity through shearing; and the effects of temperature and stirring on viscosity have an additive effect. Therefore, there is a lack of a comprehensive regulatory index that can simultaneously integrate information on microbial community status, mixing efficiency, and mass transfer resistance.

[0004] It is particularly important to note that existing publicly available multi-source data fusion monitoring methods based on metabolic entropy and catastrophic re-ignition simulated annealing algorithms (such as CN202610003967.8), while improving anomaly diagnosis sensitivity through the quantification of metabolic entropy uncertainty and the catastrophic re-ignition mechanism, are essentially passive anomaly diagnosis frameworks—identifying and matching anomalies after they occur. These methods rely on pre-set standard trajectory vectors and anomaly databases, and the temperature parameters of their simulated annealing algorithm must be positively correlated with metabolic entropy. The overall framework focuses on detecting whether anomalies have occurred and matching anomaly types. In contrast, this invention does not start with anomaly diagnosis but focuses on the active quantification of microbial community synergy and the active regulation of mass transfer-metabolism coupling in the context of multi-strain synergistic fermentation of traditional Chinese medicine. The two differ fundamentally in their technical objectives, indicator systems, and control logic. Summary of the Invention

[0005] The technical problem to be solved by the present invention is that the existing technology has the disadvantages of lacking quantitative means for the mutual restraint relationship between multiple microbial species, ignoring intermediate metabolites in control decision-making, and lacking coupled decision-making. To this end, we propose a multi-strain synergistic fermentation control system for traditional Chinese medicine.

[0006] In a first aspect, the present invention provides a multi-strain synergistic fermentation control system for traditional Chinese medicine, comprising a tank body, a motor fixedly installed on the top of the tank body, a stirring shaft fixedly connected to the output end of the motor, a central turbine stirring paddle, a staggered stirring paddle and a bottom turbine stirring paddle fixedly installed on the stirring shaft, a temperature sensor installed in the middle of the tank body, and an electrode sensor installed on the top of the tank body.

[0007] The multi-strain synergistic fermentation control system for traditional Chinese medicine also includes a controller, which is configured as follows:

[0008] The metabolic intensity fusion calculation unit is used to obtain the real-time concentration of various substrates in the main body of the tank, and combine it with the real-time temperature of the temperature sensor to obtain the metabolite transient concentration weighted model. The metabolite transient concentration weighted model characterizes the overall activity level of substrate consumption.

[0009] The synergistic entropy calculation unit is used to obtain the real-time pH and real-time dissolved oxygen of the electrode sensor, and combine them with the weighted modulus of the transient concentration of the metabolites to obtain a synergistic index, which characterizes the degree of synergy among multiple bacterial species.

[0010] The coupling control factor calculation unit is used to obtain the real-time speed of the motor and the real-time temperature of the temperature sensor, and combine them with the synergy index to obtain the coupling control factor, which characterizes the support capability of the current mass transfer conditions for synergistic fermentation.

[0011] The coupling regulatory factor is used to compare with a predetermined target interval for closed-loop iterative regulation; the synergistic index is used to quantify the degree of balance of metabolic activities of multiple bacterial species rather than to quantify process uncertainty; and the controller aims at active regulation rather than abnormal diagnosis.

[0012] Preferably, the metabolic intensity fusion calculation unit includes:

[0013] The substrate consumption rate calculation module is used to obtain the ratio of the concentration difference of each substrate at adjacent sampling times to the sampling time interval, and thus obtain the consumption rate of each substrate.

[0014] The weight mapping module is used to obtain the activity weight coefficient of each substrate based on the real-time temperature of the temperature sensor through a pre-calibrated temperature weight lookup table function;

[0015] The weighted fusion module is used to multiply the consumption rate of each substrate by the corresponding activity weight coefficient to obtain the weighted consumption rate of each substrate.

[0016] The module for calculating the modulus length is used to square the weighted consumption rates of each substrate, sum them up, and then take the square root of the summation to obtain the weighted modulus of the transient concentration of the metabolite.

[0017] Preferably, the temperature weighted lookup table function is calibrated through an offline shake-flask experiment:

[0018] The consumption rate of each substrate at different temperatures was measured, and the consumption rate at each temperature was divided by the maximum consumption rate of the substrate at all temperatures to obtain the activity weighting coefficient corresponding to each temperature.

[0019] When the real-time temperature of the temperature sensor is between two calibration points, the active weight coefficient is obtained by linear interpolation.

[0020] Preferably, the collaborative entropy calculation unit includes:

[0021] The microbial community activity inversion calculation module is used to obtain the specific metabolic rate of each microbial species based on the real-time pH and real-time dissolved oxygen of the electrode sensor through a pre-calibrated binary lookup table model of specific metabolic rate.

[0022] The metabolic flux weighting module is used to multiply the specific metabolic rate of each bacterial species by the weighting modulus of the transient concentration of the metabolite to obtain the actual metabolic flux of each bacterial species.

[0023] The percentage probability calculation module is used to divide the actual metabolic flux of each bacterial species by the sum of the actual metabolic fluxes of all bacterial species to obtain the metabolic percentage probability of each bacterial species.

[0024] The collaborative computing module is used to multiply the metabolic proportion probability of each bacterial species by its natural logarithm, sum all the multiplications, and then take the negative value to obtain the collaborative index.

[0025] Preferably, the specific metabolic rate binary lookup table model is calibrated through offline shake-flask experiments:

[0026] Under different combinations of pH and dissolved oxygen, the specific metabolic rate of each bacterial species was measured, and a two-dimensional response surface lookup table was established with pH and dissolved oxygen as inputs and specific metabolic rate as output.

[0027] When the real-time pH and real-time dissolved oxygen of the electrode sensor are between the calibration points, the specific metabolic rate of each bacterial species is obtained by bilinear interpolation.

[0028] Preferably, the coupling regulation factor calculation unit includes:

[0029] The mixing time calculation module is used to obtain the effective volume of the main body of the tank and the blade diameter of the central turbine agitator. Combined with the real-time stirring speed of the motor, the homogenization time constant of the whole tank is calculated according to the mixing time correlation formula. The mixing time correlation formula is expressed as the effective volume divided by the product of the stirring speed and the cube of the blade diameter.

[0030] The temperature viscosity correction module is used to obtain a temperature viscosity correction factor based on the fermentation broth temperature of the temperature sensor, combined with a pre-calibrated ideal gas constant, activation energy parameter and fermentation broth reference temperature, through an Arrhenius-type temperature correction formula. The temperature viscosity correction factor characterizes the multiplier effect of temperature on the viscosity of the fermentation broth.

[0031] The shear viscosity correction module is used to obtain a shear viscosity correction factor based on the stirring speed of the motor, combined with a pre-calibrated reference speed and shear thinning index, through a power-law shear correction formula. The shear viscosity correction factor characterizes the ratio of the influence of stirring shear on the apparent viscosity of the fermentation broth.

[0032] The apparent viscosity fusion calculation module is used to multiply the pre-calibrated basic viscosity coefficient by the temperature viscosity correction factor and the shear viscosity correction factor to obtain the dynamic apparent viscosity.

[0033] The coupling control factor calculation module is used to multiply the synergy index by the whole tank homogenization time constant and then divide by the dynamic apparent viscosity to obtain the coupling control factor.

[0034] Preferably, a manhole is provided on the main body of the tank, and an elliptical pull-out cover is installed on the manhole. A lifting lug is fixedly installed on the outer wall of the main body of the tank. A semi-pipe sleeve is arranged around the outer wall of the main body of the tank. A through valve is fixedly connected to both ends of the semi-pipe sleeve. An upper spray ring pipe is arranged inside the upper part of the main body of the tank. The upper spray ring pipe is fixed by an upper ring fixing bracket. A lower spray ring pipe is arranged inside the lower part of the main body of the tank. The lower spray ring pipe is fixed by a fixing bracket. A sampling valve is provided on the main body of the tank. A discharge valve is provided at the bottom of the main body of the tank.

[0035] The technical effects and advantages of this invention are as follows:

[0036] In this invention, the concentrations of various traditional Chinese medicine substrates collected by a near-infrared spectral probe installed at a manhole or sampling valve are extracted as consumption rates through time-difference processing. These rates are then weighted by temperature-dependent weighting coefficients and fused using the L2 norm to obtain a weighted modulus of transient metabolite concentrations. The pH and dissolved oxygen measured by electrode sensors are then inverted into specific metabolic rates for each bacterial species using a binary response surface model. This is multiplied by the weighted modulus of transient metabolite concentrations to obtain the actual metabolic flux. After normalization to a probability distribution, a synergy index is calculated using the information entropy formula. This synergy index transforms the fuzzy concept of balanced metabolism among bacterial species into a precise, calculable, comparable, and threshold-set indicator. When the synergy index approaches its maximum value, bacterial synergy is good; when it approaches zero, a single bacterial species severely suppresses other species. The control system does not need to wait for offline detection; it can obtain the synergistic state of the bacterial community in real time within each sampling cycle and make regulatory decisions.

[0037] In this invention, the dynamic information of intermediate metabolites is indirectly incorporated into the closed-loop control framework through multi-level feature extraction of substrate consumption rates. The substrate consumption rate naturally implies the dynamics of intermediate metabolite generation and consumption: after substrate consumption, it is first converted into various intermediate metabolites, and then further into final products. In the calculation of the synergy index, the changing trend of the actual metabolic flux of each bacterial species actually reflects the balance between the accumulation and consumption of intermediate metabolites: when the flux of a certain bacterial species continuously increases, intermediate products may accumulate; when the flux suddenly decreases, it may be due to the accumulation of intermediate products reaching the toxicity threshold, triggering feedback inhibition. Furthermore, the square operation of the consumption rate amplifies the signal of metabolically active substrates, allowing for the sensitive capture of shifts in the metabolic center of gravity. This enables the control system to detect early signals of abnormal accumulation of intermediate metabolites and intervene before the microbial imbalance evolves into an irreversible state.

[0038] In this invention, the stirring speed fed back from the motor, combined with the tank volume and impeller diameter, is calculated as the whole-tank homogenization time constant through a mixing time correlation formula. Temperature is transformed into a temperature-viscosity correction factor using an Arrhenius equation, and rotation speed is transformed into a shear-viscosity correction factor using a power-law model. These two factors are then integrated with the base viscosity coefficient to form the dynamic apparent viscosity. Multiplying the synergy index by the whole-tank homogenization time constant and dividing by the dynamic apparent viscosity yields a coupling control factor that unifies the two previously independent decisions of adjusting rotation speed and temperature into a single comprehensive index. A large coupling control factor indicates insufficient mass transfer conditions, requiring improvement in stirring or temperature; a small factor indicates microbial imbalance, requiring priority adjustment of temperature to restore microbial balance or reduction of shear strength.

[0039] Furthermore, by comparing the coupling control factor with the target range determined by offline optimization, the system can automatically select the optimal combination of control actions and use closed-loop iterative search to make the control factor converge to the optimal range of speed and temperature combination. Attached Figure Description

[0040] Figure 1 This is a front view of the overall structure of the multi-strain synergistic fermentation device for traditional Chinese medicine of the present invention;

[0041] Figure 2 For the present invention Figure 1 Enlarged schematic diagram of the structure at point A;

[0042] Figure 3 This is a side cross-sectional view of the multi-strain synergistic fermentation device for traditional Chinese medicine of the present invention;

[0043] Figure 4 For the present invention Figure 3 Enlarged schematic diagram of the structure at point A1;

[0044] Figure 5 This is a bottom cross-sectional view of the multi-strain synergistic fermentation device for traditional Chinese medicine of the present invention;

[0045] Figure 6 This is a schematic diagram of the entire process of multi-strain synergistic fermentation of traditional Chinese medicine in this invention.

[0046] In the diagram: 1-Tank body, 2-Motor, 3-Manhole, 4-Oval pull-out cover, 5-Lifting lug, 6-Half-pipe jacket, 7-Through-body valve, 8-Main mixing shaft, 9-Middle turbine agitator, 10-Staggered agitator, 11-Bottom turbine agitator, 12-Upper spray ring pipe, 13-Upper ring fixing frame, 14-Lower spray ring pipe, 16-Fixing frame, 17-Sampling valve, 18-Temperature sensor, 19-Electrode sensor, 20-Discharge valve. Detailed Implementation

[0047] The present invention will now be described in further detail with reference to the accompanying drawings and preferred embodiments.

[0048] The control system is designed for applications primarily in the multi-strain synergistic fermentation stage of modern Chinese medicine production. It addresses the dynamic control challenges posed by the complex metabolic competition and synergistic relationships among various microorganisms, such as yeast, lactic acid bacteria, and Bacillus, acting on the Chinese medicine substrate. These challenges are difficult to address effectively using traditional single-parameter PID control. Furthermore, the fermentation substrates of Chinese medicine exhibit complex characteristics, including multiple components (dozens of active ingredients such as polysaccharides, glycosides, flavonoids, alkaloids, and volatile oils) and multiple targets (different components act on different metabolic pathways). The extracts also contain large amounts of viscous components such as polysaccharides, tannins, and pectin, causing the fermentation broth to exhibit significant shear-thinning non-Newtonian fluid properties. This rheological behavior differs fundamentally from that of conventional biological fermentation broths, making it difficult to directly apply anomaly detection frameworks based on single scalar indicators (such as metabolic entropy) in general fermentation monitoring methods. This invention constructs a three-layer progressive index system consisting of a metabolite transient concentration weighted model, a synergy index, and coupled regulatory factors. This system enables real-time quantitative assessment and closed-loop control of the metabolic balance and mass transfer matching degree of microbial communities during the synergistic fermentation of traditional Chinese medicine, highlighting the special process requirements in the field of traditional Chinese medicine fermentation.

[0049] In an exemplary embodiment, in this fermentation control system, key parameters during the fermentation process are collected in real time by a near-infrared spectral probe, a temperature sensor 18, and an electrode sensor 19. The controller's built-in algorithm performs comprehensive analysis and outputs control commands.

[0050] like Figures 1 to 6 As shown, this embodiment provides a multi-strain synergistic fermentation control system for traditional Chinese medicine, including a tank body 1. A motor 2 is fixedly installed on the top of the tank body 1, and a stirring shaft 8 is fixedly connected to the output end of the motor 2. A central turbine-type stirring paddle 9, a staggered stirring paddle 10, and a bottom turbine-type stirring paddle 11 are fixedly installed on the stirring shaft 8. A temperature sensor 18 is installed in the middle of the tank body 1, and an electrode sensor 19 is installed on the top of the tank body 1. A manhole 3 is provided on the tank body 1, and an elliptical puller is installed on the manhole 3. The tank body 1 has a cover 4 and a lifting lug 5 fixedly installed on its outer wall. The outer wall of the tank body 1 is surrounded by a semi-pipe sleeve 6, and the two ends of the semi-pipe sleeve 6 are respectively fixedly connected to a through valve 7. The upper spray ring pipe 12 is installed inside the upper part of the tank body 1 and is fixed by an upper ring fixing bracket 13. The lower spray ring pipe 14 is installed inside the lower part of the tank body 1 and is fixed by a fixing bracket 16. A sampling valve 17 is installed on the tank body 1 and a discharge valve 20 is installed at the bottom of the tank body 1.

[0051] The fermentation control system also includes a controller, which is configured as follows:

[0052] The metabolic intensity fusion calculation unit is used to obtain the real-time concentration of various substrates in the main body of the tank 1, and combined with the real-time temperature of the temperature sensor 18, to obtain the metabolite transient concentration weighted model. The metabolite transient concentration weighted model characterizes the overall activity level of substrate consumption.

[0053] The synergistic entropy calculation unit is used to obtain the real-time pH and real-time dissolved oxygen of the electrode sensor 19, and combine them with the weighted model of metabolite transient concentration to obtain the synergistic index, which characterizes the degree of synergy among multiple bacterial species.

[0054] The coupling control factor calculation unit is used to obtain the real-time speed of motor 2 and the real-time temperature of temperature sensor 18, and combine them with the synergy index to obtain the coupling control factor. The coupling control factor characterizes the support capability of the current mass transfer conditions for synergistic fermentation.

[0055] Understandably, the process of using a fermentation device will be explained using the example of multi-strain synergistic fermentation of a certain traditional Chinese medicine compound (containing multiple herbs such as astragalus, codonopsis, and poria cocos):

[0056] First, preparation before fermentation: The pulverized Chinese medicinal material matrix is ​​put into the tank body 1 through the manhole 3 and sealed through the oval pull-out cover 4 on the manhole 3. Multi-strain bacterial solution (such as a mixed bacterial solution of Bacillus subtilis, lactic acid bacteria and yeast) is put into the tank through the manhole 3 or a special bacterial inlet. The motor 2 is started and the stirring shaft 8 drives the middle turbine stirring paddle 9, the staggered stirring paddle 10 and the bottom turbine stirring paddle 11 to rotate, so as to achieve uniform mixing of materials and bacteria.

[0057] Secondly, the through valve 7 controls the flow rate and temperature of the heat exchange medium (hot water or cooling water) in the half-pipe jacket 6 to achieve temperature control inside the tank. The temperature sensor 18 monitors the temperature of the fermentation liquid in real time and feeds it back to the controller.

[0058] During fermentation, fermentation broth samples can be periodically extracted via sampling valve 17 for offline calibration and verification. Electrode sensor 19 detects pH and dissolved oxygen in real time, temperature sensor 18 detects fermentation broth temperature in real time, and near-infrared spectroscopy probe (which can be installed at manhole 3 or sampling valve 17) acquires substrate concentration data in real time. All data is transmitted to the controller for analysis and control.

[0059] After fermentation, the fermentation products are discharged through the discharge valve 20. The upper spray ring pipe 12 and the lower spray ring pipe 14 are connected to an external cleaning water source. After being fixed by the upper ring fixing bracket 13 and the fixing bracket 16, the interior of the tank body 1 can be sprayed and cleaned in all directions. The large-diameter design of the manhole 3 makes it easy for operators to enter the tank for inspection and deep cleaning.

[0060] The aforementioned fermentation control system enables real-time quantification of the synergistic state of multiple microbial species and coupled regulation of mass transfer and metabolism, effectively improving the controllability of the fermentation process and the quality of the products.

[0061] The following description, in conjunction with the accompanying drawings, details the specific implementation process of each step in the fermentation process, from data transmission to the controller for analysis and control.

[0062] Step S1: Determine the weighted mode of metabolite transient concentration based on substrate consumption rate and temperature dependence weights.

[0063] In the synergistic fermentation of traditional Chinese medicine (TCM) by multiple microorganisms, the main body 1 of the tank simultaneously contains various TCM substrates (such as polysaccharides, glycosides, flavonoids, etc.) and multiple microbial species. Different microbial species have different metabolic capacities for the same substrate, and the enzyme activity of the same microbial species also varies at different temperatures. In order to quantify the overall activity of the entire microbial community in consuming the substrate in real time, rather than the consumption of a single substrate, this invention constructs a metabolic intensity fusion calculation unit.

[0064] Understandably, the consumption rate of substrates is the most direct reflection of the metabolic activity of the microbial community. A faster consumption rate indicates that the corresponding metabolic pathway is active. However, the consumption rates of different substrates vary due to substrate molecule size, structure, and species preference, making direct comparison or simple summation difficult. For example, the consumption of 1 gram of polysaccharide and the consumption of 1 gram of flavonoid have completely different implications for the metabolic state of the microbial community. Meanwhile, temperature is a key factor affecting the rate of enzymatic reactions; changes in temperature will redistribute the metabolic priorities of different species for different substrates. Therefore, it is necessary to construct a comprehensive index that can integrate information from multiple substrate components and reflect the role of temperature regulation.

[0065] In this embodiment, a near-infrared spectral probe installed at the manhole 3 or sampling valve 17 is used to acquire in real time the concentration data S of n kinds of traditional Chinese medicine substrate components inside the tank body 1. 1(t) ,S 2(t) ,S 3(t) ,...,S n(t) Near-infrared spectroscopy technology enables simultaneous, rapid, and non-destructive detection of multiple components in complex systems. Its probe can be directly installed on the tank, enabling real-time online monitoring and avoiding the time lag and contamination risks associated with offline sampling and testing.

[0066] Obtain the concentration difference ΔS of each substrate at adjacent sampling times. i(t) (ΔS) i(t) =S i(t) -ΔS i(t-Δt) The ratio of the sample rate to the sampling time interval Δt yields the real-time consumption rate of each substrate: .

[0067] Here, i represents the i-th substrate. The consumption rate is a vector component that includes the direction (increase or decrease) and magnitude of metabolic activity. A positive and large consumption rate indicates that the substrate is being consumed rapidly, and the corresponding bacterial species is metabolically active.

[0068] Meanwhile, based on the real-time temperature T of the temperature sensor 18, the activity weight coefficients of each substrate are obtained through a pre-calibrated temperature weight lookup table function. The temperature sensor 18 is installed in the middle of the tank, which is closest to the actual temperature of the fermentation broth and can effectively represent the thermodynamic environment of the microbial community.

[0069] The temperature weight lookup table function was calibrated through offline shake-flask experiments: For each substrate and strain combination, fermentation experiments were conducted in laboratory shake flasks with a series of different temperature gradients. The consumption rate of each substrate at different temperatures was measured, and the consumption rate at each temperature was divided by the maximum consumption rate of that substrate at all temperatures to obtain the activity weight coefficient w corresponding to each temperature. i The coefficient w i The value between 0 and 1 reflects the degree to which the current temperature promotes or inhibits the consumption process of a specific substrate. When the real-time temperature of temperature sensor 18 is between two calibration points, the activity weighting coefficient, i.e., w, is obtained using linear interpolation. i =f i (T t ).

[0070] Thus, two types of information have been obtained: first, the substrate consumption rate, which directly reflects metabolic dynamics; and second, the activity weighting coefficient w, which reflects the effect of temperature regulation. i .

[0071] To integrate the consumption rates of multiple substrate components into a single indicator that can characterize the overall metabolic intensity, a fusion strategy is needed that can amplify significant metabolic activity while avoiding the cancellation of positive and negative values. Simple summation will cancel each other out due to the different signs of the consumption rates, making it difficult to accurately reflect the overall metabolic activity. Therefore, this embodiment uses a weighted L2 norm (Euclidean norm) for fusion.

[0072] Specifically, the consumption rate of each substrate is multiplied by its corresponding activity weighting coefficient to obtain the weighted consumption rate of each substrate. The physical meaning of this product is the temperature-corrected substrate consumption rate; the more suitable the temperature, the greater the weight and the more significant its contribution.

[0073] Then, the weighted consumption rates of each substrate are squared and summed, and the sum is then squared to obtain the weighted model of transient metabolite concentrations.

[0074] ;

[0075] In the formula: MTCW is the transient concentration weighted mode of metabolites, a comprehensive index characterizing the current activity level of substrate consumption in the tank, M t This is a multi-component substrate concentration vector, reflecting the real-time concentrations of n kinds of traditional Chinese medicine substrates (polysaccharides, glycosides, flavonoids, etc.) in the container, i.e. That is, an n-dimensional column vector consisting of the consumption rates of n substrates, |⋅| w This is a weighted L2 norm operator, a way of defining a vector norm, where the subscript w indicates that the norm has weighting coefficients, n is the total composition of the Chinese herbal medicine substrate, and w i Let ΔS be the activity weight coefficient of the i-th substrate in the current microbial community metabolic network. i(t) Δt represents the real-time consumption rate of the i-th substrate, reflecting the change in the concentration of the i-th substrate per unit time, and Δt is the time interval between two adjacent samplings.

[0076] Thus, a core indicator capable of comprehensively characterizing the total kinetic energy of substrate consumption within the tank was obtained. The square operation amplifies the signal of metabolically active substrates, allowing for the sensitive capture of shifts in the metabolic center of gravity; the square root operation ensures dimensional consistency. When multiple substrates are consumed rapidly simultaneously, the MTCW value increases significantly; conversely, when metabolism stagnates or proceeds in different directions, the MTCW value decreases. This indicator compresses complex multi-component metabolic information into an intuitive scalar, providing crucial input for subsequent assessment of the synergistic state of the microbial community.

[0077] Step S2: Determine the synergy index based on the weighted model of pH, dissolved oxygen and transient metabolite concentrations.

[0078] After obtaining the MTCW (Metabolism-to-Weight) value, which characterizes the overall metabolic intensity, it is necessary to further identify whether this metabolic activity is accomplished by the synergistic action of multiple microbial species or is dominated by one or two dominant species. In the synergistic fermentation of traditional Chinese medicine, the ideal fermentation state is one where the metabolic activities of each species are relatively balanced, forming a synergistic effect of mutual promotion. However, existing pH and dissolved oxygen sensors can only provide macroscopic environmental parameters and cannot directly provide information on the metabolic activity of each microbial species.

[0079] It is understandable that pH and dissolved oxygen concentration are comprehensive reflections of microbial metabolic activity in the environment. Different microbial species have different sensitivities and tolerance ranges to pH and dissolved oxygen, and their metabolic activity is clearly correlated with these two parameters. By establishing this correlation model in advance through offline experiments, the real-time metabolic activity of each microbial species can be inverted and extrapolated using the real-time measured pH and dissolved oxygen values ​​during fermentation.

[0080] In this embodiment, electrode sensor 19 is installed on the top of the tank to measure the pH value and dissolved oxygen concentration (DO) of the fermentation broth in real time. These two parameters seem conventional, but through an offline pre-calibrated binary lookup table model of specific metabolic rate, they can be mapped to the specific metabolic rate q of each bacterial species. j(t) .

[0081] The specific metabolic rate binary lookup table model was calibrated through offline shake-flask or small-scale fermenter experiments: under different combinations of pH and dissolved oxygen, the specific metabolic rate (i.e., the rate of substrate consumption per unit cell concentration) of the j-th bacterial species was measured, and a two-dimensional response surface lookup table was established with pH and dissolved oxygen as inputs and specific metabolic rate as output. When the real-time pH and real-time dissolved oxygen of electrode sensor 19 were between the calibration points, the specific metabolic rate of each bacterial species was obtained using bilinear interpolation. The core value of this model lies in the fact that, during actual operation, the pH measured by electrode sensor 19... t and D Ot Substitute h j q is obtained through bilinear interpolation. j(t) ), to pH and D O Mapped to specific metabolic rates for each bacterial species:

[0082] q j(t) =h j (pH t D Ot );

[0083] q j(t) The meaning is: the j-th bacterial community at the current pH and D O Under certain conditions, the rate at which substrate is consumed per unit bacterial cell concentration. This is a deeper characteristic than pH itself; it translates environmental conditions into the bacterial community's willingness to engage.

[0084] The specific metabolic rate q of each bacterial species j(t) Multiplying this by the metabolite transient concentration weighted modulus (MTCW) yields the actual metabolic flux of each bacterial species. The physical meaning of this product is: the actual metabolic intensity contributed by the j-th bacterial species under the current overall metabolic activity level. If overall metabolism is stagnant (small MTCW), even if a certain bacterial species is theoretically very active (q...), the actual metabolic flux will be lost. j(t) If the overall metabolism is active, the actual contribution of all microorganisms will be suppressed; conversely, if the overall metabolism is active, the actual contribution of all microorganisms will be amplified.

[0085] To quantify the uniformity of metabolic activity among different bacterial species, we borrow the concept of entropy from information theory. Entropy can be used to measure the degree of disorder or uniformity in the distribution of a system. Here, we normalize the actual metabolic flux of each bacterial species to obtain the metabolic proportion probability of each species: .

[0086] Wherein, FACE is a synergy index, characterizing the uniformity of metabolic activity among multiple bacterial species, i.e., the activity coupling entropy, m is the total number of bacterial species, and the denominator is... This represents the sum of actual metabolic fluxes of all bacterial species, where j is the species index variable, j=1 represents the first species, j=2 represents the second species, ..., j=m represents the last species, and q represents the third species. j(t) q represents the specific metabolic rate of the j-th bacterial species, reflecting the rate at which the j-th bacterial species consumes substrate per unit bacterial cell concentration. It is obtained by looking up a table using pH and dissolved oxygen values ​​measured by electrode sensor 19 (top-mounted). k is also a bacterial species index variable (used for summing in the denominator), an index variable of the exact same type as j, only with a different letter to avoid confusion. k(t) Let be the specific metabolic rate of the k-th bacterial group, and be related to q. j(t) They are the same type of quantity; k is just another dummy variable used for summation.

[0087] Multiply the metabolic proportion probability of each bacterial species by its natural logarithm, sum all the multiplications, and then take the negative value to obtain the synergy index FACE:

[0088] ;

[0089] Thus, a core indicator capable of quantifying the synergistic state of the microbial community in real time was obtained. The higher the FACE value, the more balanced the metabolic activities of the various microbial species and the higher the degree of synergy; conversely, a FACE value close to 0 indicates that a single microbial species has severely suppressed other microbial species, resulting in an imbalance in the microbial community structure.

[0090] This indicator transforms the fuzzy concept of microbial community synergy into a precise, calculable, comparable, and threshold-settable indicator. This allows the control system to obtain real-time information about the microbial community's state within each sampling cycle and make corresponding regulatory decisions without waiting for offline detection results. Furthermore, changes in substrate consumption rates naturally imply the dynamics of intermediate metabolite generation and consumption. Through MTCW transmission, this indicator indirectly incorporates the dynamic information of intermediate metabolites into the closed-loop control framework. A sustained increase or sudden decrease in the flux of a particular microbial species indicates potential changes in metabolic pathways, thus providing the possibility for early intervention.

[0091] Step S3: Determine the coupling control factor based on synergy indicators, mixing efficiency, and dynamic apparent viscosity.

[0092] After obtaining the synergistic index FACE, which characterizes what the microbial community needs, the system needs to determine what the current physical environment (mainly mass transfer conditions) can provide. Mass transfer conditions are mainly determined by two operating parameters: stirring and temperature. Stirring affects mixing uniformity and shear rate, while temperature affects diffusion coefficient and fermentation broth viscosity. There is a coupling effect between the two; for example, decreasing temperature increases viscosity, thereby reducing mass transfer efficiency, while increasing stirring speed can enhance mixing, but may also generate excessive shear force that damages the microbial cells.

[0093] Understandably, in order to make optimal regulatory decisions, it is necessary to construct a comprehensive index that can simultaneously integrate the needs of the microbial community (synergistic FACE) and the mass transfer supply capacity (mixing time, apparent viscosity). This index should be able to reflect whether the existing mass transfer conditions are matched, excessive, or insufficient under the current synergistic state of the microbial community, thereby guiding whether to accelerate, increase, decrease, or slow down the process.

[0094] In this embodiment, the coupling control factor calculation unit first calculates the whole tank homogenization time constant τ. mix(n) The effective volume V of the main body 1 and the blade diameter D of the central turbine agitator 9 are obtained. Combined with the real-time stirring speed N of the motor 2, the homogenization time constant of the entire tank is calculated according to the mixing time correlation formula. In one embodiment, the mixing time correlation formula is expressed as:

[0095] ;

[0096] Where V is the effective volume, reflecting the total volume of the fermentation liquid in the fermenter; N is the stirring speed, reflecting the current actual output speed of motor 2; and D is the impeller diameter, reflecting the impeller diameter of the central turbine-type stirring paddle 9. The physical meaning of this correlation is: the larger the effective volume V, the longer the mixing time required; the faster the stirring speed N and the larger the impeller diameter D, the higher the convection circulation intensity, and the shorter the mixing time. τ mix(n) It is the inverse indicator of macroscopic mixing efficiency.

[0097] Next, the dynamic apparent viscosity μ is calculated. app(N,T) Traditional Chinese medicine fermentation broth is usually a non-Newtonian fluid, and its viscosity depends not only on temperature and concentration but also on shear rate.

[0098] First, based on the real-time fermentation broth temperature T from temperature sensor 18, combined with the pre-calibrated ideal gas constant R... g Activation energy parameter E a and fermentation broth reference temperature T ref The temperature viscosity correction factor is obtained using the Arrhenius-type temperature correction formula: This correction factor characterizes the magnitude of the effect of temperature on the viscosity of the fermentation broth. As temperature increases, molecular motion intensifies, interlayer friction decreases, viscosity decreases, and the value of this correction factor decreases.

[0099] Simultaneously, based on the stirring speed N of motor 2, combined with the pre-calibrated reference speed N0 and shear thinning index β, the shear viscosity correction factor is obtained through a power-law shear correction formula: This correction factor characterizes the magnitude of the effect of stirring shear on the apparent viscosity of the fermentation broth. For shear-thinned fluids (β<0), the higher the rotational speed, the greater the shear rate, the lower the apparent viscosity, and the smaller the value of this correction factor.

[0100] Finally, the pre-calibrated baseline viscosity coefficient μ0 (reflecting the intrinsic viscosity of the fermentation broth at a shear-free, reference temperature) is multiplied by the temperature viscosity correction factor and the shear viscosity correction factor to obtain the dynamic apparent viscosity: This dynamic apparent viscosity is a core indicator of mass transfer resistance. The higher the viscosity, the smaller the molecular diffusion coefficient, and the greater the mass transfer resistance.

[0101] Thus, we have obtained FACE, representing the demand of the microbial community, and τ, representing the mixed supply. mix(n) And μ, representing the mass transfer resistance app(N,T) To construct the coupling regulator STCF, these three elements need to be fused. The fusion logic is as follows:

[0102] The higher the synergy of the microbial community (the larger the FACE), the more sensitive the system is to mass transfer conditions, and the more significant the improvement in mass transfer. Therefore, STCF should be positively correlated with FACE.

[0103] The longer the mixing time (τ) mix(n) The larger the value (τ), the worse the mass transfer capacity provided by the current stirring, which is the bottleneck. Therefore, STCF should be compared with τ. mix(n) They are positively correlated.

[0104] The greater the mass transfer resistance (μ) app(N,T) The larger the value (μ), the more difficult molecular diffusion becomes, which is less conducive to co-fermentation. Therefore, STCF should be combined with μ app(N,T) It is negatively correlated.

[0105] Based on this, the synergy index FACE and the whole tank homogenization time constant τ are used to... mix(n) After multiplication, divide by the dynamic apparent viscosity μ app(N,T) The coupling regulation factor STCF is obtained:

[0106] ;

[0107] In the formula:

[0108] STCF is a coupling regulatory factor that characterizes the degree to which mass transfer and mixing under current stirring and temperature conditions promote or inhibit co-fermentation. mix(n) E is a function of mixing time, reflecting the time required to complete one homogeneous mixing of the entire tank at a stirring speed N. μ0 is the basic viscosity coefficient, reflecting the intrinsic viscosity of the fermentation broth under shear-free conditions. a R is the activation energy parameter, reflecting the sensitivity of the fermentation broth viscosity to temperature. g Let T be the ideal gas constant, T be the temperature of the fermentation broth, and T be the temperature measured by temperature sensor 18. ref is the reference temperature of the fermentation broth, N0 is the reference rotation speed, and β is the shear thinning index, an empirical index of the shear thinning behavior of the fermentation broth.

[0109] Thus, a core indicator has been obtained that can comprehensively characterize the current mass transfer conditions' support for co-fermentation. This indicator unifies the two previously independent variables of stirring speed and temperature into a single framework. A higher STCF value indicates that the current mixing efficiency is severely insufficient relative to the needs of the microbial community and the resistance to mass transfer, and the system urgently needs intervention to improve mass transfer (such as increasing the stirring speed or reducing viscosity). Conversely, a very low STCF value indicates either that the microbial community itself has very low synergy (small facet), making mass transfer improvement largely meaningless, or that the mass transfer conditions are sufficient (fast mixing, low viscosity), already well supporting the current co-fermentation requirements.

[0110] Step S4: Based on the comparison between the coupling control factor and the target interval, execute the closed-loop iterative control strategy.

[0111] After obtaining the real-time coupling control factor STCF, the system needs to determine whether the current value is good or bad based on an objective benchmark, and then execute corresponding control actions accordingly. This benchmark is the target range determined through offline process optimization experiments.

[0112] Understandably, STCF is a system that integrates biological information (FACE) and engineering information (μ). app(N,T) The comprehensive index of ) has an optimal operating window with respect to the quality of the final product.

[0113] In this embodiment, multiple batches of fermentation experiments with different stirring speeds and temperature combinations were conducted in a small-scale experimental fermenter for the same traditional Chinese medicine formula and strain combination. The relationship between the coupling regulatory factor STCF and the final product quality (such as the yield of active ingredients, product activity, etc.) for each batch was recorded. The range of STCF values ​​corresponding to maintaining optimal product quality was determined as the target interval, i.e., [STCF]. min STCF max This target range provides a clear setpoint for subsequent closed-loop control.

[0114] The controller will use the real-time calculated coupling control factor STCF and the pre-determined target interval [STCF] min STCF max The comparison is performed, and the following control strategies are implemented based on the comparison results:

[0115] Scenario 1: When STCF falls within [STCF] min STCF max Within the specified range, it indicates that the current stirring and temperature conditions match the synergy of the microbial community, and the mass transfer conditions are just able to support the current synergistic fermentation requirements. Therefore, the controller does not intervene, and the fermentation process continues in the optimal state.

[0116] Scenario 2: When STCF exceeds the upper limit of the target interval, i.e., STCF > STCF max This indicates that mass transfer conditions have become the bottleneck for co-fermentation. At this point, the system can increase the stirring intensity and shorten the mixing time by adjusting the speed of motor 2; or adjust the temperature by changing the flow rate of the heat exchange medium in the semi-tube jacket 6 by adjusting the through valve 7, thereby reducing the viscosity of the fermentation broth and decreasing the mass transfer resistance. The goal is to bring the STCF back to the optimal range. The specific action to choose depends on a comprehensive judgment based on the characteristics of the fermentation broth (such as whether it is shear-sensitive) and the characteristics of the microbial strain (such as whether it is heat-resistant).

[0117] Scenario 3: When STCF is lower than the lower limit of the target interval, i.e., STCF < STCF min If the FACE index is below the preset threshold, it indicates that the system is not sensitive to mass transfer conditions. In this case, it is necessary to first determine whether the FACE index is below the preset threshold. If so, it means that the microbial community itself is already unbalanced, and mass transfer improvement cannot solve the problem. In this case, the temperature should be adjusted by changing the flow rate of the heat exchange medium in the half-pipe jacket 6 through the through-body valve 7 to restore microbial community activity (FACE). If the FACE index is not below the threshold, it may mean over-mixing (excessive shear force generated by excessive stirring) or excessive viscosity. In this case, the mass transfer conditions should be optimized by adjusting the speed of the motor 2.

[0118] The controller is also used to re-trigger the metabolic intensity fusion calculation unit, the synergistic entropy calculation unit, and the coupling regulatory factor calculation unit after each of the above-mentioned regulatory actions to perform iterative calculations, obtain new STCF values, and compare them with the target interval again. This process continues until the coupling regulatory factor converges to the target interval.

[0119] This completes a full closed-loop iterative optimization process. This process transforms the complex problem of controlling the synergistic fermentation of multiple microbial strains of traditional Chinese medicine into an optimization problem for the comprehensive index STCF, achieving intelligent and automated control of a multivariable, strongly coupled system, and significantly improving the controllability and batch-to-batch consistency of the fermentation process.

[0120] It should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should also be within the scope of protection of this invention.

Claims

1. A multi-strain synergistic fermentation control system for traditional Chinese medicine, characterized in that, The tank body (1) includes a motor (2) fixedly installed on the top of the tank body (1), and a stirring shaft (8) fixedly connected to the output end of the motor (2). A central turbine stirring blade (9), a staggered stirring blade (10) and a bottom turbine stirring blade (11) are fixedly installed on the stirring shaft (8). A temperature sensor (18) is installed in the middle of the tank body (1), and an electrode sensor (19) is installed on the top of the tank body (1). The multi-strain synergistic fermentation control system for traditional Chinese medicine also includes a controller, which is configured as follows: The metabolic intensity fusion calculation unit is used to obtain the real-time concentration of various substrates in the main body of the tank (1) through the near-infrared spectral probe, and combine it with the real-time temperature of the temperature sensor (18) to obtain the metabolite transient concentration weighted model. The metabolite transient concentration weighted model characterizes the overall activity level of substrate consumption. The synergistic entropy calculation unit is used to obtain the real-time pH and real-time dissolved oxygen of the electrode sensor (19), and combine them with the weighted model of the transient concentration of the metabolites to obtain the synergistic index, which characterizes the degree of synergy among multiple bacterial species. The coupling control factor calculation unit is used to obtain the real-time speed of the motor (2) and the real-time temperature of the temperature sensor (18), and combine them with the synergy index to obtain the coupling control factor, which characterizes the support capability of the current mass transfer conditions for synergistic fermentation. The coupling regulatory factor is used to compare with a predetermined target interval for closed-loop iterative regulation; the synergistic index is used to quantify the degree of balance of metabolic activities of multiple bacterial species rather than to quantify process uncertainty; and the controller aims at active regulation rather than abnormal diagnosis.

2. The multi-strain synergistic fermentation control system for traditional Chinese medicine according to claim 1, characterized in that, The metabolic intensity fusion calculation unit includes: The substrate consumption rate calculation module is used to obtain the ratio of the concentration difference of each substrate at adjacent sampling times to the sampling time interval, and thus obtain the consumption rate of each substrate. The weight mapping module is used to obtain the activity weight coefficient of each substrate based on the real-time temperature of the temperature sensor (18) through a pre-calibrated temperature weight lookup table function; The weighted fusion module is used to multiply the consumption rate of each substrate by the corresponding activity weight coefficient to obtain the weighted consumption rate of each substrate. The module for calculating the modulus length is used to square the weighted consumption rates of each substrate, sum them up, and then take the square root of the summation to obtain the weighted modulus of the transient concentration of the metabolite.

3. The multi-strain synergistic fermentation control system for traditional Chinese medicine according to claim 2, characterized in that, The temperature weighted lookup table function was calibrated through an offline shake-flask experiment: The consumption rate of each substrate at different temperatures was measured, and the consumption rate at each temperature was divided by the maximum consumption rate of the substrate at all temperatures to obtain the activity weighting coefficient corresponding to each temperature. When the real-time temperature of the temperature sensor (18) is between two calibration points, the active weight coefficient is obtained by linear interpolation.

4. A multi-strain synergistic fermentation control system for traditional Chinese medicine according to any one of claims 1-3, characterized in that, The collaborative entropy calculation unit includes: The microbial community activity inversion calculation module is used to obtain the specific metabolic rate of each microbial species based on the real-time pH and real-time dissolved oxygen of the electrode sensor (19) through a pre-calibrated binary lookup table model of specific metabolic rate. The metabolic flux weighting module is used to multiply the specific metabolic rate of each bacterial species by the weighting modulus of the transient concentration of the metabolite to obtain the actual metabolic flux of each bacterial species. The percentage probability calculation module is used to divide the actual metabolic flux of each bacterial species by the sum of the actual metabolic fluxes of all bacterial species to obtain the metabolic percentage probability of each bacterial species. The collaborative computing module is used to multiply the metabolic proportion probability of each bacterial species by its natural logarithm, sum all the multiplications, and then take the negative value to obtain the collaborative index.

5. The multi-strain synergistic fermentation control system for traditional Chinese medicine according to claim 4, characterized in that, The specific metabolic rate binary lookup table model was calibrated through offline shake-flask experiments: Under different combinations of pH and dissolved oxygen, the specific metabolic rate of each bacterial species was measured, and a two-dimensional response surface lookup table was established with pH and dissolved oxygen as inputs and specific metabolic rate as output. When the real-time pH and real-time dissolved oxygen of the electrode sensor (19) are between the calibration points, the specific metabolic rate of each bacterial species is obtained by bilinear interpolation.

6. The multi-strain synergistic fermentation control system for traditional Chinese medicine according to claim 1, characterized in that, The coupling regulation factor calculation unit includes: The mixing time calculation module is used to obtain the effective volume of the tank body (1) and the blade diameter of the turbine agitator (9) in the middle. Combined with the real-time stirring speed of the motor (2), the homogenization time constant of the whole tank is calculated according to the mixing time correlation formula. The mixing time correlation formula is expressed as the effective volume divided by the product of the stirring speed and the cube of the blade diameter. The temperature viscosity correction module is used to obtain a temperature viscosity correction factor based on the fermentation broth temperature of the temperature sensor (18), combined with the pre-calibrated ideal gas constant, activation energy parameter and fermentation broth reference temperature, through the Arrhenius type temperature correction formula. The temperature viscosity correction factor characterizes the influence ratio of temperature on the viscosity of the fermentation broth. The shear viscosity correction module is used to obtain the shear viscosity correction factor by power-law shear correction formula based on the stirring speed of the motor (2), combined with the pre-calibrated reference speed and shear thinning index. The shear viscosity correction factor characterizes the ratio of the influence of stirring shear on the apparent viscosity of the fermentation liquid. The apparent viscosity fusion calculation module is used to multiply the pre-calibrated basic viscosity coefficient by the temperature viscosity correction factor and the shear viscosity correction factor to obtain the dynamic apparent viscosity. The coupling control factor calculation module is used to multiply the synergy index by the whole tank homogenization time constant and then divide by the dynamic apparent viscosity to obtain the coupling control factor.

7. The multi-strain synergistic fermentation control system for traditional Chinese medicine according to claim 1, characterized in that, The tank body (1) has a manhole (3) and an elliptical pull cover (4) installed on the manhole (3). The outer wall of the tank body (1) is fixedly equipped with a lifting lug (5). The outer wall of the tank body (1) is surrounded by a half-pipe sleeve (6). The two ends of the half-pipe sleeve (6) are respectively fixedly connected to a through valve (7). The upper spray ring pipe (12) is provided inside the upper part of the tank body (1). The upper spray ring pipe (12) is fixed by an upper ring fixing bracket (13). The lower spray ring pipe (14) is provided inside the lower part of the tank body (1). The lower spray ring pipe (14) is fixed by a fixing bracket (16). The tank body (1) is equipped with a sampling valve (17). The bottom of the tank body (1) is equipped with a discharge valve (20).

Images