A microbial fermentation agitation rate control system and method

Abstract

This application discloses a microbial fermentation stirring rate control system and method. It acquires historical fermentation batch process detection values, control setpoints, reactor structural parameters, and fermentation effect data. Historical time-series characteristics are extracted according to fermentation physiological stages and overlapping sliding time windows. A historical state mapping relationship is established between dissolved oxygen supply and demand deviation, impeller shear risk, product synthesis rate, and stirring rate setpoint. The process detection values, control setpoints, and reactor structural parameters of the current fermentation batch are collected to determine the current dissolved oxygen supply and demand deviation, current impeller shear risk, and dynamic adjustable range. Mass transfer coordination constraints are generated based on the aeration rate setpoint, tank pressure setpoint, and dissolved oxygen control valve opening. Within the dynamic adjustable range, a target stirring rate value is determined and closed-loop correction is performed. Using this technical solution can reduce excessive impeller shear, increased stirring power, and batch fluctuations caused by simply increasing the stirring rate to restore dissolved oxygen.

Description

Technical Field

[0001] This application relates to the field of fermentation process control technology, and more specifically, to a microbial fermentation stirring rate control system and method. Background Technology

[0002] Microbial fermentation processes typically maintain the growth and metabolic environment within the fermenter through temperature, pH, dissolved oxygen, feed, and pressure loops. The stirring rate is an adjustable control input, primarily used to alter the mixing state of the fermentation broth and the gas-liquid mass transfer capacity, and simultaneously affects the shear stress on the cell walls. Actual dissolved oxygen, temperature, pH, pressure, and cell concentration are process monitoring values ​​used to characterize the actual state within the fermenter.

[0003] In existing industrial fermentation processes, stirring rates are typically adjusted based on fixed-stage curves, manual experience, or dissolved oxygen cascade control. Fixed-stage curves rely on preset time points and struggle to reflect the actual increase in oxygen demand as cell concentration rises. Conventional dissolved oxygen cascade control primarily adjusts aeration rate, tank pressure, or stirring rate sequentially based on dissolved oxygen deviations. While this can improve dissolved oxygen recovery rates, in fermentation stages with high cell concentrations, high viscosity, or shear-sensitive strains, the influence of impeller diameter, impeller type, and impeller tip velocity on shear forces is not taken into account. This can easily lead to excessive impeller tip shear, increased stirring power, and amplified batch fluctuations in pursuit of dissolved oxygen recovery. The essence of these problems lies in the fact that existing stirring control typically relies on dissolved oxygen deviation recovery as the primary criterion, lacking synchronous constraints on impeller tip shear forces and power changes caused by stirring adjustments. This makes it difficult to simultaneously achieve dissolved oxygen recovery, cell shear protection, and batch stability in high cell concentration, high viscosity, or shear-sensitive stages. Summary of the Invention

[0004] This application provides a microbial fermentation stirring rate control system and method, which can solve the problem that existing stirring control relies mainly on dissolved oxygen deviation recovery as the judgment criterion in high cell concentration, high viscosity or shear-sensitive stages, and lacks synchronous constraints on impeller shearing action and power changes, resulting in difficulty in simultaneously ensuring dissolved oxygen recovery, cell shear protection and batch stability.

[0005] In a first aspect, this application provides a method for controlling the stirring rate of microbial fermentation, applied to the stirring control of a microbial fermentation system, the method comprising the following steps: Acquire process detection values, control setpoints, reactor structural parameters, and fermentation effect data of historical fermentation batches, and acquire the reactor structural parameters corresponding to the current fermenter; Historical time-series features of the process detection values ​​and corresponding control setpoints were extracted according to the fermentation physiological stages and overlapping sliding time windows. Based on the historical time-series features, the control setpoints, and the reactor structural parameters, a historical state mapping relationship was established between dissolved oxygen supply and demand deviation, impeller shear risk, product synthesis rate, and stirring rate setpoints. Collect process detection values ​​and control set values ​​of the current fermentation batch, extract the current time series features according to the same overlapping sliding time window, determine the basic adjustable range of stirring rate based on the mapping relationship between the current time series features and the historical state, and narrow or shift the range based on the current dissolved oxygen supply and demand deviation and the current impeller tip shear risk to obtain the dynamic adjustable range. With the dissolved oxygen circuit as the main control circuit, mass transfer coordination constraints are generated based on the setpoint of the aeration rate, the setpoint of the tank pressure, and the opening degree of the dissolved oxygen control valve. The target value of the stirring rate is determined within the dynamically adjustable range. The target stirring rate is converted into an executable instruction for the fermentation control device, and a closed-loop correction is performed based on the feedback deviation of the actual dissolved oxygen value.

[0006] In some embodiments, the fermentation effect data includes product synthesis rate, substrate consumption rate, cell concentration change rate, and batch endpoint product concentration; the process detection value is the actual detection value recorded by the fermenter sensor, offline detection instrument, or central control system; the control setting value is the control input value received or executed by the fermentation control equipment; the reactor structural parameters are fixed equipment parameters of the same fermenter or isomorphic fermenters; the reactor structural parameters include impeller diameter, impeller type identification, fermenter working volume, and stirring motor operating limits; and the process detection value, the control setting value, the reactor structural parameters, and the fermentation effect data are all bound to the fermentation batch identifier, fermentation physiological stage, and fermentation timestamp.

[0007] In some embodiments, extracting the historical time-series features of the process detection values ​​and corresponding control setpoints according to the fermentation physiological stage and overlapping sliding time windows specifically includes: Historical fermentation batches were labeled according to the lag phase, logarithmic growth phase, product synthesis phase, and stationary phase in the fermentation batch. In each physiological stage of fermentation, an overlapping sliding time window with a window duration greater than the sliding step length is set, and the process detection value and the corresponding control setting value are divided according to the overlapping sliding time window to obtain the historical operating status data segment. Extract the average dissolved oxygen deviation, integral dissolved oxygen deviation, negative rate of change of actual dissolved oxygen value, temperature fluctuation amplitude, tank pressure fluctuation amplitude, and cell concentration change rate from the historical operating status data segment, and extract the stirring rate change rate from the corresponding control setpoint. The historical time series features are obtained by arranging the features of each historical operating state data segment within the same fermentation physiological stage in chronological order.

[0008] In some embodiments, establishing a historical state mapping relationship between dissolved oxygen supply and demand deviation, impeller tip shear risk, product synthesis rate, and stirring rate setpoint specifically includes: The dissolved oxygen deviation is determined based on the difference between the dissolved oxygen setpoint and the actual dissolved oxygen value. The dissolved oxygen supply-demand deviation is determined based on the integral of the dissolved oxygen deviation, the negative change rate of the actual dissolved oxygen value, and the change rate of the bacterial cell concentration. The impeller tip linear velocity is calculated based on the impeller diameter and stirring rate setpoint, and the impeller tip shear risk is determined based on the impeller tip linear velocity, stirring rate change rate, impeller type identification, fermentation broth viscosity grade, and cell shear sensitivity grade. According to the fermentation physiological stage, the dissolved oxygen supply and demand deviation, the impeller tip shear risk, the product synthesis rate, and the stirring rate setting value in the corresponding window are stored to form the historical state mapping relationship. The historical state mapping relationship is used to find historical samples in the current fermentation batch where the dissolved oxygen supply and demand deviation is close, the impeller shear risk does not exceed the process limit, the product synthesis rate meets the production requirements, and there are corresponding stirring rate set values. The historical stirring rate value range is determined by the historical samples.

[0009] In some embodiments, the basic adjustable range is narrowed or shifted based on the current dissolved oxygen supply-demand deviation and the current propeller tip shear risk to obtain a dynamically adjustable range, specifically including: If the current dissolved oxygen supply and demand deviation is within the allowable range and the current impeller tip shear risk is below the risk limit, maintain the basic adjustable range or narrow the basic adjustable range towards the current stirring rate; If the current dissolved oxygen supply and demand deviation exceeds the allowable range and the current impeller tip shear risk does not exceed the risk limit, the basic adjustable range will be shifted upward; if the current impeller tip shear risk reaches the risk limit or the current stirring rate change rate exceeds the allowable change rate, the basic adjustable range will be shifted downward or narrowed. The modified interval boundary is restricted within the operating limits of the stirring motor, the linear velocity limit of the impeller tip, and the safety limits of the fermentation process to obtain the dynamically adjustable interval.

[0010] In some embodiments, generating mass transfer coordination constraints and determining a target stirring rate value within the dynamically adjustable range specifically includes: The setpoints for ventilation rate, tank pressure, and dissolved oxygen control valve opening are used as cooperative manipulation variables within the dissolved oxygen circuit. The executable boundaries and adjustment directions of these cooperative manipulation variables are determined, and the mass transfer cooperative constraints are generated. Multiple candidate stirring rate values ​​are set within the dynamically adjustable range. Based on the historical state mapping relationship, the expected dissolved oxygen supply and demand deviation, expected impeller shear risk, stirring power increment, and stirring rate change amplitude corresponding to each candidate stirring rate are determined respectively. After normalizing the expected dissolved oxygen supply and demand deviation, the expected impeller shear risk, the stirring power increment, and the stirring rate change amplitude, an objective function is formed. The candidate stirring rate with the smallest objective function value that satisfies the mass transfer coordination constraint is selected as the stirring rate target value.

[0011] In some embodiments, converting the target stirring rate value into an executable instruction for the fermentation control device and performing closed-loop correction based on the feedback deviation of the actual dissolved oxygen value specifically includes: According to the communication address, data format and allowed update frequency of the fermentation control equipment, the target value of the stirring rate is converted into a speed setting command that can be recognized by the stirring motor controller. After the speed setting command is executed, the actual value of dissolved oxygen and the actual value of stirring rate are collected. If the feedback deviation of the actual value of dissolved oxygen still exceeds the allowable range of the process, the process returns to the step of determining the dynamic adjustable range for further correction.

[0012] Secondly, this application provides a microbial fermentation stirring rate control system, the system comprising: The data acquisition module is used to acquire process detection values, control setpoints, reactor structural parameters and fermentation effect data of historical fermentation batches, and to collect process detection values, control setpoints and reactor structural parameters corresponding to the current fermentation tank for the current fermentation batch. The time-series processing module is used to extract historical time-series features and current time-series features according to the fermentation physiological stage and overlapping sliding time windows. The historical time-series features are extracted from the process detection values ​​and corresponding control setpoints. The interval determination module is used to establish a historical state mapping relationship between dissolved oxygen supply and demand deviation, impeller tip shear risk, product synthesis rate and stirring rate setpoint. Based on the current time series characteristics and the historical state mapping relationship, the basic adjustable interval of stirring rate is determined, and a dynamic adjustable interval is obtained based on the current dissolved oxygen supply and demand deviation and the current impeller tip shear risk. The target value generation module is used to generate mass transfer coordination constraints based on the dissolved oxygen circuit as the main control circuit, the aeration rate set value, the tank pressure set value and the opening degree of the dissolved oxygen control valve, and to determine the stirring rate target value within the dynamically adjustable range. The execution correction module is used to convert the target stirring rate value into an executable instruction for the fermentation control device, and to perform closed-loop correction based on the feedback deviation of the actual dissolved oxygen value.

[0013] Thirdly, this application provides a computer device, the computer device including a memory and a processor, the memory storing code, and the processor being configured to acquire the code and execute the microbial fermentation stirring rate control method disclosed in this application.

[0014] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the microbial fermentation stirring rate control method disclosed in this application.

[0015] The technical solution disclosed in this application has the following beneficial effects: This application addresses the problem that existing stirring controls rely primarily on dissolved oxygen deviation recovery during high cell concentration, high viscosity, or shear-sensitive stages. This lack of simultaneous constraints on impeller tip shearing and power changes makes it difficult to simultaneously achieve dissolved oxygen recovery, cell shear protection, and batch stability. The application defines stirring rate as a manipulated variable within the dissolved oxygen loop, rather than using it as a cross-loop clustering center. Through the separate acquisition of process detection values, control setpoints, and reactor structural parameters, the application clarifies that actual dissolved oxygen, temperature, pH, tank pressure, and cell concentration are used to characterize the fermentation state; stirring rate setpoints, aeration rate setpoints, tank pressure setpoints, feed flow rate setpoints, and cooling water flow rate setpoints are used to characterize the control inputs; and impeller diameter and impeller type identifiers are used to characterize the equipment basis for impeller tip shearing. This ensures that the historical state mapping relationship is established on a clear basis of controlled variables, manipulated variables, and reactor structural boundaries.

[0016] By extracting historical time-series features through fermentation physiological stages and overlapping sliding time windows, continuous changes in fermentation status between adjacent windows can be preserved. A basic adjustable range is determined by mapping historical states between dissolved oxygen supply-demand deviation, impeller shear risk, product synthesis rate, and stirring rate setpoints. This range is then narrowed or shifted based on the current dissolved oxygen supply-demand deviation and impeller shear risk, ensuring stability or slight narrowing when fermentation is in good condition, moderately increasing the upper limit of stirring rate when dissolved oxygen supply is insufficient, and decreasing or narrowing the stirring rate range when impeller shear risk increases. Thus, the target stirring rate value is generated under the combined constraints of dissolved oxygen recovery requirements, impeller shear constraints, and the increment and magnitude of stirring power changes. This improves the matching degree between stirring regulation and the current fermentation state, reducing excessive impeller shear, increased stirring power, and batch fluctuations caused by simply increasing the stirring rate to restore dissolved oxygen. Attached Figure Description

[0017] Figure 1 This is an exemplary flowchart of a microbial fermentation stirring rate control method according to some embodiments of this application; Figure 2This is a schematic diagram illustrating the extraction process of historical time-series features according to some embodiments of this application; Figure 3 This is a schematic diagram illustrating the process of establishing historical state mapping relationships according to some embodiments of this application; Figure 4 This is a schematic diagram of the structure of a microbial fermentation stirring rate control system according to some embodiments of this application; Figure 5 This is a schematic diagram of the structure of a computer device for implementing a microbial fermentation stirring rate control method according to some embodiments of this application. Detailed Implementation

[0018] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] refer to Figure 1 The figure is an exemplary flowchart of a microbial fermentation stirring rate control method according to some embodiments of this application. The microbial fermentation stirring rate control method mainly includes the following steps: In step 101, process detection values, control setpoints, reactor structural parameters and fermentation effect data of historical fermentation batches are obtained, and reactor structural parameters corresponding to the current fermenter are obtained.

[0020] Process monitoring values ​​are the results of actual operating conditions within the fermenter, including actual dissolved oxygen, temperature, pH, tank pressure, and cell concentration. Control setpoints are the control inputs received or executed by the fermentation control equipment, including stirring rate setpoints, aeration rate setpoints, tank pressure setpoints, feed flow rate setpoints, and cooling water flow rate setpoints. Reactor structural parameters include impeller diameter, impeller type identification, fermenter working volume, and stirring motor operating limits; these parameters are linked to the fermenter identifier and fermentation batch identifier. Fermentation performance data includes product synthesis rate, substrate consumption rate, cell concentration change rate, and batch endpoint product concentration. Process monitoring values, control setpoints, and fermentation performance data are all linked to the fermentation batch identifier, fermentation physiological stage, and fermentation timestamp to distinguish the actual process state, control inputs, and equipment structural boundaries in subsequent processing.

[0021] In step 102, the historical time-series features of the process detection values ​​and corresponding control setpoints are extracted according to the fermentation physiological stages and overlapping sliding time windows. Based on the historical time-series features, the control setpoints, and the reactor structural parameters, a historical state mapping relationship is established between dissolved oxygen supply and demand deviation, impeller shear risk, product synthesis rate, and stirring rate setpoints.

[0022] refer to Figure 2The figure is a schematic diagram of the historical time series feature extraction process in some embodiments of this application. In this embodiment, the historical time series features of the process detection values ​​and corresponding control setpoints extracted according to the fermentation physiological stage and overlapping sliding time window can be obtained through the following steps: In step 1021, historical fermentation batches are labeled according to the lag phase, logarithmic growth phase, product synthesis phase and stationary phase in the fermentation batch; In step 1022, an overlapping sliding time window with a window duration greater than the sliding step length is set in each fermentation physiological stage, and the process detection value and the corresponding control setting value are divided according to the overlapping sliding time window to obtain the historical operating status data segment. In step 1023, the average dissolved oxygen deviation, the integral of dissolved oxygen deviation, the negative change rate of actual dissolved oxygen value, the temperature fluctuation amplitude, the tank pressure fluctuation amplitude, and the cell concentration change rate are extracted from the historical operating status data segment, and the stirring rate change rate is extracted from the corresponding control setpoint. In step 1024, the features of each historical operating state data segment within the same fermentation physiological stage are arranged in chronological order to obtain the historical time series features.

[0023] In practice, the overlapping sliding time window is determined by the window duration and the sliding step size, with the window duration being longer than the sliding step size. Therefore, there is an overlap between adjacent windows. For example, if the fermentation parameter acquisition interval is one minute, the window duration can be set to six minutes, and the sliding step size to two minutes, allowing two adjacent windows to share four minutes of data. The overlapping time period is used to maintain the continuity of states between adjacent windows, preventing the erroneous segmentation of continuous changes such as decreased dissolved oxygen or increased cell concentration into unrelated fragments.

[0024] When dividing the historical operating status data segment, the process detection values ​​and corresponding control setpoints are first sorted according to the fermentation timestamp. Then, a data segment of one window duration is extracted starting from the start time of each fermentation physiological stage. Subsequently, the starting point of the window is moved by a sliding step size, and data segments are extracted continuously until the end of the fermentation physiological stage. If the number of data points in the end window is less than the predetermined minimum number, the end window is merged into the previous window or discarded. Each historical operating status data segment records the fermentation batch identifier, fermentation physiological stage, start time, end time, and the number of data points contained.

[0025] Within each historical operating status data segment, the average dissolved oxygen deviation is obtained by subtracting the actual dissolved oxygen value from the setpoint and averaging the results within the window; the integral of the dissolved oxygen deviation is obtained by multiplying the dissolved oxygen deviation at each sampling time within the window by the sampling interval and then summing the results; the negative rate of change of the actual dissolved oxygen value is the non-negative slope of the actual dissolved oxygen value as it decreases over time, and is set to zero when the actual dissolved oxygen value does not decrease; the temperature fluctuation amplitude and the tank pressure fluctuation amplitude are obtained by the difference between the maximum and minimum values ​​within the window, respectively; the rate of change of bacterial concentration is obtained by dividing the increase in bacterial concentration by the window duration; and the rate of change of stirring rate is obtained by dividing the increase in the stirring rate setpoint in the corresponding control setpoint by the corresponding time interval.

[0026] To reduce the interference of invalid features on subsequent judgments, the fluctuation degree of each feature within the same fermentation physiological stage was statistically analyzed, and features that remained unchanged for a long time in all historical batches and could not reflect the deviation of dissolved oxygen supply and demand or the risk of impeller shearing were removed. Features with different dimensions were normalized so that each feature was converted into a dimensionless value. The normalized features were stored according to the fermentation batch identifier, fermentation physiological stage and window number to form historical time-series features.

[0027] It should be noted that the historical time series characteristics do not simply record the historical values ​​of the stirring rate, but rather record the state characteristics of dissolved oxygen demand, cell growth, temperature and tank pressure fluctuations over time within the same fermentation physiological stage; this characteristic is used to determine the dissolved oxygen supply and demand deviation and impeller shear risk state of the current fermentation batch.

[0028] refer to Figure 3 The figure is a schematic diagram of the process of establishing historical state mapping relationships in some embodiments of this application. In this embodiment, the historical state mapping relationship between dissolved oxygen supply and demand deviation, impeller shear risk, product synthesis rate and stirring rate setpoint is established based on the historical time series characteristics, the control setpoint and the reactor structural parameters through the following steps: In step 1031, the dissolved oxygen deviation is determined based on the difference between the dissolved oxygen setpoint and the actual dissolved oxygen value. The dissolved oxygen supply and demand deviation amount used to characterize the degree of enhancement on the oxygen demand side is determined based on the dissolved oxygen deviation integral, the negative change rate of the actual dissolved oxygen value, and the cell concentration change rate. In step 1032, the blade tip linear velocity is calculated based on the impeller diameter and the set value of the stirring rate, and the blade tip shear risk is determined based on the blade tip linear velocity, the stirring rate change rate, the impeller type identifier, the fermentation broth viscosity level, and the cell shear sensitivity level. In step 1033, the dissolved oxygen supply and demand deviation, the impeller shear risk, the product synthesis rate, and the stirring rate setting value in the corresponding window are stored according to the fermentation physiological stage to form the historical state mapping relationship. The historical state mapping relationship is used to find historical samples in the current fermentation batch where the dissolved oxygen supply and demand deviation is close, the impeller shear risk does not exceed the process limit, the product synthesis rate meets the production requirements, and there are corresponding stirring rate set values. The historical stirring rate value range is determined by the historical samples.

[0029] In practice, the dissolved oxygen supply-demand deviation does not represent the actual oxygen transfer rate from the gas phase to the liquid phase. Instead, it is obtained by summing the normalized dissolved oxygen deviation integral, the negative rate of change of the actual dissolved oxygen value, and the rate of change of cell concentration, all weighted non-negative and summing to one. A larger dissolved oxygen deviation integral indicates a longer duration of oxygen deficiency within that window; a larger negative rate of change of the actual dissolved oxygen value indicates a faster increase in oxygen demand; and a larger rate of change of cell concentration generally indicates a faster increase in biomass per unit time, with a potential synchronous increase in oxygen demand. These three quantities are used together to characterize the intensity of the current fermentation broth's demand for oxygen supply.

[0030] The impeller tip shear risk is obtained by summing the normalized impeller tip linear velocity, the rate of change of stirring speed, the impeller type influence coefficient, the fermentation broth viscosity grade, and the cell shear sensitivity grade according to their weights. Each weight is a non-negative number and the sum is one. When the stirring speed is expressed in revolutions per minute, the impeller tip linear velocity is calculated by multiplying pi by the impeller diameter, then by the stirring speed setpoint, and finally dividing by sixty. The impeller type influence coefficient is used to distinguish the influence of radial flow impellers, axial flow impellers, or other impeller structures on shear action. The fermentation broth viscosity grade can be obtained by classifying online viscosity measurement values ​​or offline viscosity measurement results. The cell shear sensitivity grade can be given in advance by the strain process document.

[0031] When establishing historical state mapping relationships, the stirring rate is not used as the cross-loop cluster center, nor are temperature loops, pH loops, feed loops, and tank pressure loops forcibly merged into the same control group. The historical state mapping relationship is only used to identify the impact of stirring rate on dissolved oxygen supply and demand deviation and impeller tip shear risk within the dissolved oxygen mass transfer correlation range, and the set boundaries of each independent control loop are preserved.

[0032] The degree of correlation involved in the historical state mapping is mainly identified through nonlinear relationship identification, supplemented by linear consistency checks. For the relationship between the actual values ​​of stirring rate, aeration rate, and dissolved oxygen, the degree of nonlinear correlation can be determined by the consistency of the changes in values ​​within the window or by the degree of overlap of information after discrete binning. When the linear correlation results and nonlinear correlation results are in opposite directions, the historical segment is marked as requiring manual review or removed from the mapping sample.

[0033] In step 103, the process detection values ​​and control settings of the current fermentation batch are collected, the current time series features are extracted according to the same overlapping sliding time window, the basic adjustable range of the stirring rate is determined according to the mapping relationship between the current time series features and the historical state, and the range is narrowed or shifted based on the current dissolved oxygen supply and demand deviation and the current impeller tip shear risk to obtain the dynamic adjustable range.

[0034] The extraction criteria for current time-series features are consistent with those for historical time-series features. Specifically, the fermentation control system reads the current actual values ​​of dissolved oxygen, temperature, pH, tank pressure, cell concentration, stirring rate, aeration rate, tank pressure, feed flow rate, cooling water flow rate, and reactor structural parameters corresponding to the current fermenter according to the sampling cycle; missing data is supplemented by adjacent valid sampling points, data that significantly exceeds the sensor's measurement range is discarded, and current operating status data segments are generated within the same fermentation physiological stage according to overlapping sliding time windows.

[0035] The current time-series features extracted from the current operating status data segment include the mean dissolved oxygen deviation, the integral of dissolved oxygen deviation, the negative rate of change of the actual dissolved oxygen value, the temperature fluctuation amplitude, the tank pressure fluctuation amplitude, the cell concentration change rate, and the stirring rate change rate. The arrangement order and dimensional normalization method of the current time-series features are consistent with those of the historical time-series features, enabling the current fermentation batch to be matched in the historical state mapping relationship.

[0036] In practice, historical samples of the same stage are first selected from the historical state mapping relationship according to the current fermentation physiological stage. Then, the absolute difference between the current dissolved oxygen supply and demand deviation and the dissolved oxygen supply and demand deviation of each historical sample is calculated. Samples with a difference less than a preset similarity threshold are used as matching samples, and historical samples with impeller tip shear risk exceeding the process limit or product synthesis rate lower than the predetermined lower limit are excluded. The stirring rate settings corresponding to the remaining historical samples are statistically analyzed, and the range that can cover most valid samples and does not exceed the equipment operating limit and impeller tip linear velocity limit is taken as the basic adjustable range.

[0037] Subsequently, the basic adjustable range is adjusted based on the current dissolved oxygen supply-demand deviation and the current impeller tip shear risk. If the current dissolved oxygen supply-demand deviation is within the allowable range and the current impeller tip shear risk is below the risk limit, it indicates that the current fermentation state is close to the process target. In this case, the basic adjustable range is maintained or the range boundary is narrowed towards the current stirring rate to reduce unnecessary speed fluctuations. If the current dissolved oxygen supply-demand deviation exceeds the allowable range and the current impeller tip shear risk does not exceed the risk limit, it indicates that the dissolved oxygen supply is insufficient but there is still a usable shear margin. In this case, the basic adjustable range is shifted upward. When the impeller tip shear risk reaches the risk limit or the stirring rate change rate exceeds the allowable change rate, the basic adjustable range is shifted downward or narrowed.

[0038] The upper boundary of the dynamically adjustable range does not exceed the lower of the following limits: the maximum permissible speed of the stirring motor, the limit of the impeller tip linear velocity, the upper limit of the process to prevent excessive foaming of the fermentation broth, and the limit of cell shear risk. The lower boundary of the dynamically adjustable range is not lower than the minimum stirring rate required to maintain basic mixing uniformity and gas-liquid dispersion. Through the above boundary restrictions, the dynamically adjustable range responds to the current dissolved oxygen supply demand without exceeding equipment safety and impeller tip shear constraints.

[0039] In step 104, the dissolved oxygen circuit is used as the main control circuit. Mass transfer coordination constraints are generated based on the set value of the aeration rate, the set value of the tank pressure, and the opening degree of the dissolved oxygen control valve. The target value of the stirring rate is determined within the dynamically adjustable range.

[0040] The controlled variable in the dissolved oxygen loop is the actual dissolved oxygen value, and the manipulated variables include the stirring rate setpoint, aeration rate setpoint, tank pressure setpoint, and dissolved oxygen control valve opening. The cooling water flow rate in the temperature loop, the acid and alkali addition amounts in the pH loop, and the feed flow rate in the feed loop are not included as manipulated variables in the same group when generating the stirring rate target value; the setting boundaries of the above loops are only used as safety verification conditions to prevent stirring regulation from disrupting the stability of other loops.

[0041] Mass transfer coordination constraints include the executable range of the aeration rate setpoint, the safe range of the tank pressure setpoint, the available margin of the dissolved oxygen control valve opening, and the limitation on the variation range of the stirring rate. When the tank pressure setpoint participates in the mass transfer coordination constraints, it is only used as an upper-level given adjustment amount within the allowable safe range of the tank pressure loop; the actual tank pressure value is still controlled by the closed-loop tank pressure loop. If the actual tank pressure value is close to the safe upper limit or the tank pressure actuator reaches the boundary, the dissolved oxygen supply will no longer be enhanced by increasing the tank pressure setpoint. If the dissolved oxygen control valve opening is already close to the upper limit and the aeration rate setpoint still has a margin, the aeration rate is allowed to be increased within the dissolved oxygen loop. If both the aeration rate and the tank pressure are close to the boundary and the impeller tip shear risk has not exceeded the limit, the stirring rate is allowed to be increased within the dynamically adjustable range. If the impeller tip shear risk is close to the limit, the increase in the stirring rate is restricted, and the executable adjustment of the aeration rate or tank pressure is maintained first.

[0042] The objective function consists of the expected dissolved oxygen supply-demand deviation, the expected impeller shear risk, the stirring power increment, and the stirring rate variation magnitude corresponding to the candidate stirring rates. For candidate stirring rates within the dynamically adjustable range, the expected dissolved oxygen supply-demand deviation and the expected impeller shear risk are determined according to the historical state mapping relationship. The stirring power increment is obtained by subtracting the current stirring motor power from the stirring motor power corresponding to the candidate stirring rate. When the central control system does not directly record the power, it is calculated from the stirring motor current, voltage, and power factor. The four indicators are then normalized and summed according to the process weights corresponding to the fermentation physiological stages to obtain the objective function value of the candidate stirring rate. The candidate stirring rate with the smallest objective function value that satisfies the mass transfer synergy constraint is selected as the target stirring rate value.

[0043] For example, during the logarithmic growth phase, when the actual dissolved oxygen value is consistently below the set lower limit, the dissolved oxygen control valve opening is close to the upper limit, and the impeller shear risk is below the risk limit, the objective function will reduce the evaluation value of the higher candidate stirring rate that meets the dissolved oxygen requirement. During the product synthesis phase, when the cell shear sensitivity level is high and the dissolved oxygen supply and demand deviation is within the allowable range, the objective function will increase the evaluation value of the candidate value corresponding to a larger stirring rate change, thereby selecting the target value that is closer to the current stirring rate.

[0044] In step 105, the target stirring rate value is converted into an executable instruction for the fermentation control device, and a closed-loop correction is performed based on the feedback deviation of the actual dissolved oxygen value.

[0045] In practice, the fermentation control system reads the communication address of the stirring motor controller, the speed setting register address, the numerical scaling ratio, and the allowed update frequency, and converts the target stirring rate into a speed setting command that the controller can recognize. If the target value is in revolutions per minute, but the controller register outputs a percentage, it is first converted to the output percentage based on the maximum speed of the stirring motor, and then written according to the register value format.

[0046] After the speed setting command is executed, the fermentation central control system continues to collect the actual dissolved oxygen value and the actual stirring rate value. If the actual stirring rate value does not reach the target value and the deviation exceeds the equipment's allowable error, the command is reissued or an alarm is triggered. If the actual stirring rate value has reached the target value but the feedback deviation of the actual dissolved oxygen value still exceeds the process's allowable range, the dynamic adjustable range is redefined, and the stirring rate target value is redefined within the new dynamic adjustable range. If the dissolved oxygen deviation still does not return to the allowable range after the number of consecutive corrections reaches the preset maximum number of iterations (e.g., 3 times), an alarm is triggered and the system switches to the preset stage fixed stirring rate curve to ensure safe process operation.

[0047] In some embodiments, the fermentation control device may also be set with a maximum single adjustment range and a minimum adjustment interval. The maximum single adjustment range is used to limit the shear shock caused by excessive changes in rotational speed in a single operation, and the minimum adjustment interval is used to ensure that the dissolved oxygen feedback after one adjustment has sufficient time to manifest. If the difference between the new target stirring rate and the current stirring rate exceeds the maximum single adjustment range, the operation is performed in segments according to the maximum single adjustment range.

[0048] During the closed-loop correction process, the process detection values ​​of the temperature loop, pH loop, feed loop, and tank pressure loop are used as safety verification conditions. If any independent control loop exceeds its limit or approaches the actuator boundary, the action of increasing the stirring rate is suspended, and the dynamically adjustable range is maintained or narrowed in priority.

[0049] When it is necessary to change the aeration rate setpoint, tank pressure setpoint, or dissolved oxygen control valve opening, the fermentation central control system only adjusts within the dissolved oxygen circuit and tank pressure safety closed loop, without changing the control objectives of the temperature circuit, pH circuit, and feeding circuit.

[0050] Through the aforementioned closed-loop process, data acquisition, interval correction, target value generation, command issuance, and feedback verification form a continuous control link. In this link, the historical time-series features obtained in the preceding steps are used to establish a historical state mapping relationship, which is used to determine the basic adjustable range. The current dissolved oxygen supply and demand deviation and the current impeller tip shear risk are used to correct the dynamically adjustable range. The dynamically adjustable range and mass transfer synergistic constraints are used to generate the stirring rate target value. The stirring rate target value is finally converted into equipment commands and further corrected through dissolved oxygen feedback. All intermediate results participate in subsequent steps.

[0051] In another aspect, in some embodiments, this application provides a microbial fermentation stirring rate control system, with reference to... Figure 4 The figure is a schematic diagram of the structure of a microbial fermentation stirring rate control system according to some embodiments of this application. The system includes a data acquisition module 401, a timing processing module 402, an interval determination module 403, a target value generation module 404, and an execution correction module 405.

[0052] The data acquisition module 401 is used to acquire process detection values, control setpoints, reactor structural parameters and fermentation effect data of historical fermentation batches, and to collect process detection values, control setpoints and reactor structural parameters corresponding to the current fermentation tank for the current fermentation batch.

[0053] The time series processing module 402 is used to extract historical time series features and current time series features according to the fermentation physiological stage and overlapping sliding time window.

[0054] The interval determination module 403 is used to establish a historical state mapping relationship between dissolved oxygen supply and demand deviation, impeller tip shear risk, product synthesis rate and stirring rate setpoint. Based on the current time series characteristics and the historical state mapping relationship, it determines the basic adjustable interval of stirring rate, and obtains the dynamic adjustable interval based on the current dissolved oxygen supply and demand deviation and the current impeller tip shear risk.

[0055] The target value generation module 404 is used to generate mass transfer coordination constraints based on the dissolved oxygen circuit as the main control circuit, the aeration rate setpoint, the tank pressure setpoint and the opening degree of the dissolved oxygen control valve, and determine the stirring rate target value within the dynamically adjustable range.

[0056] The execution correction module 405 is used to convert the target stirring rate value into an executable instruction for the fermentation control device, and to perform closed-loop correction based on the feedback deviation of the actual dissolved oxygen value.

[0057] In addition, this application also provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to acquire the code and execute the microbial fermentation stirring rate control method disclosed in this application.

[0058] In some embodiments, reference Figure 5 The figure is a schematic diagram of the structure of a computer device for implementing a microbial fermentation stirring rate control method according to some embodiments of this application. The above-described microbial fermentation stirring rate control method can... Figure 5 The computer device shown is implemented as follows: the computer device 500 includes at least one processor 501, a communication bus 502, a memory 503, and at least one communication interface 504.

[0059] Processor 501 can be a general-purpose central processing unit or an application-specific integrated circuit.

[0060] The communication bus 502 is used to transmit information between the aforementioned components.

[0061] The memory 503 may be a read-only memory, a random access memory, an erasable read-only memory, an optical disc storage device, a magnetic disk storage device, or other storage media that can be read by a computer. The memory 503 may be set up independently and connected to the processor 501 via the communication bus 502; the memory 503 may also be integrated with the processor 501.

[0062] The memory 503 stores program code for executing the scheme of this application and is controlled by the processor 501 for execution. The processor 501 executes the program code stored in the memory 503. The program code may include one or more software modules. The above-described microbial fermentation stirring rate control method can be implemented by the processor 501 and the program code in the memory 503.

[0063] Communication interface 504 is used for communication with other devices or communication networks.

[0064] In a practical implementation, a computer device may include multiple processors, each of which may be a single-core processor or a multi-core processor. Here, a processor may refer to one or more devices, circuits, or processing cores used to process data.

[0065] The aforementioned computer equipment can be general-purpose computer equipment or special-purpose computer equipment. In specific implementations, the computer equipment can be a desktop computer, a portable computer, a network server, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer equipment.

[0066] In addition, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the microbial fermentation stirring rate control method disclosed in this application.

[0067] Although embodiments of this application have been described, those skilled in the art can make changes and modifications to these embodiments without departing from the spirit and scope of this application.

[0068] This application is intended to include modifications and variations that fall within the scope of the technical concept of this application.

Claims

1. A method for controlling the stirring rate in microbial fermentation, applied to the stirring control of a microbial fermentation system, characterized in that, Includes the following steps: Acquire process detection values, control setpoints, reactor structural parameters, and fermentation effect data of historical fermentation batches, and acquire the reactor structural parameters corresponding to the current fermenter; Historical time-series features of the process detection values ​​and corresponding control setpoints were extracted according to the fermentation physiological stages and overlapping sliding time windows. Based on the historical time-series features, the control setpoints, and the reactor structural parameters, a historical state mapping relationship was established between dissolved oxygen supply and demand deviation, impeller shear risk, product synthesis rate, and stirring rate setpoints. Collect process detection values ​​and control set values ​​of the current fermentation batch, extract the current time series features according to the same overlapping sliding time window, determine the basic adjustable range of stirring rate based on the mapping relationship between the current time series features and the historical state, and narrow or shift the range based on the current dissolved oxygen supply and demand deviation and the current impeller tip shear risk to obtain the dynamic adjustable range. With the dissolved oxygen circuit as the main control circuit, mass transfer coordination constraints are generated based on the setpoint of the aeration rate, the setpoint of the tank pressure, and the opening degree of the dissolved oxygen control valve. The target value of the stirring rate is determined within the dynamically adjustable range. The target stirring rate is converted into an executable instruction for the fermentation control device, and a closed-loop correction is performed based on the feedback deviation of the actual dissolved oxygen value.

2. The method according to claim 1, characterized in that, The fermentation effect data includes product synthesis rate, substrate consumption rate, cell concentration change rate, and batch endpoint product concentration; the process detection value is the actual detection value recorded by the fermenter sensor, offline detection instrument, or central control system; the control setting value is the control input value received or executed by the fermentation control equipment; the reactor structural parameters are fixed equipment parameters of the same fermenter or isomorphic fermenters; the reactor structural parameters include impeller diameter, impeller type identification, fermenter working volume, and stirring motor operating limits; and the process detection value, the control setting value, the reactor structural parameters, and the fermentation effect data are all bound to the fermentation batch identifier, fermentation physiological stage, and fermentation timestamp.

3. The method according to claim 1, characterized in that, The historical time-series features of the process detection values ​​and corresponding control setpoints extracted according to the fermentation physiological stages and overlapping sliding time windows specifically include: Historical fermentation batches were labeled according to the lag phase, logarithmic growth phase, product synthesis phase, and stationary phase in the fermentation batch. In each physiological stage of fermentation, an overlapping sliding time window with a window duration greater than the sliding step length is set, and the process detection value and the corresponding control setting value are divided according to the overlapping sliding time window to obtain the historical operating status data segment. Extract the average dissolved oxygen deviation, integral dissolved oxygen deviation, negative rate of change of actual dissolved oxygen value, temperature fluctuation amplitude, tank pressure fluctuation amplitude, and cell concentration change rate from the historical operating status data segment, and extract the stirring rate change rate from the corresponding control setpoint. The historical time series features are obtained by arranging the features of each historical operating state data segment within the same fermentation physiological stage in chronological order.

4. The method according to claim 1, characterized in that, Establishing a historical state mapping relationship between dissolved oxygen supply and demand deviation, impeller tip shear risk, product synthesis rate, and stirring rate setpoint specifically includes: The dissolved oxygen deviation is determined based on the difference between the dissolved oxygen setpoint and the actual dissolved oxygen value. The dissolved oxygen supply-demand deviation is determined based on the integral of the dissolved oxygen deviation, the negative change rate of the actual dissolved oxygen value, and the change rate of the bacterial cell concentration. The impeller tip linear velocity is calculated based on the impeller diameter and stirring rate setpoint, and the impeller tip shear risk is determined based on the impeller tip linear velocity, stirring rate change rate, impeller type identification, fermentation broth viscosity grade, and cell shear sensitivity grade. According to the fermentation physiological stage, the dissolved oxygen supply and demand deviation, the impeller tip shear risk, the product synthesis rate, and the stirring rate setting value in the corresponding window are stored to form the historical state mapping relationship. The historical state mapping relationship is used to find historical samples in the current fermentation batch where the dissolved oxygen supply and demand deviation is close, the impeller shear risk does not exceed the process limit, the product synthesis rate meets the production requirements, and there are corresponding stirring rate set values. The historical stirring rate value range is determined by the historical samples.

5. The method according to claim 1, characterized in that, Based on the current dissolved oxygen supply-demand deviation and the current propeller tip shear risk, the basic adjustable range is narrowed or shifted to obtain the dynamic adjustable range, which specifically includes: If the current dissolved oxygen supply and demand deviation is within the allowable range and the current impeller tip shear risk is below the risk limit, maintain the basic adjustable range or narrow the basic adjustable range towards the current stirring rate; if the current dissolved oxygen supply and demand deviation exceeds the allowable range and the current impeller tip shear risk does not exceed the risk limit, shift the basic adjustable range upward. If the current impeller tip shear risk reaches the risk limit or the current stirring rate change rate exceeds the allowable change rate, the basic adjustable range is shifted downward or narrowed, and the corrected range boundary is limited within the stirring motor operating limit, the impeller tip linear velocity limit, and the fermentation process safety limit to obtain the dynamic adjustable range.

6. The method according to claim 1, characterized in that, Generating mass transfer coordination constraints and determining the target stirring rate within the dynamically adjustable range specifically includes: The setpoints for ventilation rate, tank pressure, and dissolved oxygen control valve opening are used as cooperative manipulation variables within the dissolved oxygen circuit. The executable boundaries and adjustment directions of these cooperative manipulation variables are determined, and the mass transfer cooperative constraints are generated. Multiple candidate stirring rate values ​​are set within the dynamically adjustable range. Based on the historical state mapping relationship, the expected dissolved oxygen supply and demand deviation, expected impeller shear risk, stirring power increment, and stirring rate change amplitude corresponding to each candidate stirring rate are determined respectively. After normalizing the expected dissolved oxygen supply and demand deviation, the expected impeller shear risk, the stirring power increment, and the stirring rate change amplitude, an objective function is formed. The candidate stirring rate with the smallest objective function value that satisfies the mass transfer coordination constraint is selected as the stirring rate target value.

7. The method according to claim 1, characterized in that, Converting the target stirring rate into executable instructions for the fermentation control equipment, and performing closed-loop correction based on the feedback deviation of the actual dissolved oxygen value, specifically includes: According to the communication address, data format and allowed update frequency of the fermentation control equipment, the target value of the stirring rate is converted into a speed setting command that can be recognized by the stirring motor controller, and the speed setting command is sent to the stirring motor controller; After the speed setting command is executed, the actual dissolved oxygen value and the actual stirring rate value are collected. If the feedback deviation of the actual dissolved oxygen value still exceeds the allowable range of the process, the process returns to the step of determining the dynamic adjustable range for further correction.

8. A microbial fermentation stirring rate control system, characterized in that, The system includes: The data acquisition module is used to acquire process detection values, control setpoints, reactor structural parameters and fermentation effect data of historical fermentation batches, and to collect process detection values, control setpoints and reactor structural parameters corresponding to the current fermentation tank for the current fermentation batch. The time-series processing module is used to extract historical time-series features and current time-series features according to the fermentation physiological stage and overlapping sliding time windows. The historical time-series features are extracted from the process detection values ​​and corresponding control setpoints. The interval determination module is used to establish a historical state mapping relationship between dissolved oxygen supply and demand deviation, impeller tip shear risk, product synthesis rate and stirring rate setpoint. Based on the current time series characteristics and the historical state mapping relationship, the basic adjustable interval of stirring rate is determined, and a dynamic adjustable interval is obtained based on the current dissolved oxygen supply and demand deviation and the current impeller tip shear risk. The target value generation module is used to generate mass transfer coordination constraints based on the dissolved oxygen circuit as the main control circuit, the aeration rate set value, the tank pressure set value and the opening degree of the dissolved oxygen control valve, and to determine the stirring rate target value within the dynamically adjustable range. The execution correction module is used to convert the target stirring rate value into an executable instruction for the fermentation control device, and to perform closed-loop correction based on the feedback deviation of the actual dissolved oxygen value.

9. A computer device, characterized in that, The computer device includes a memory and a processor, the memory storing code, and the processor being configured to retrieve the code and execute the microbial fermentation stirring rate control method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the microbial fermentation stirring rate control method according to any one of claims 1 to 7.

Images