A control system and method for purifying tryptophan in peanut sprouts

Abstract

This invention belongs to the field of parameter control technology and discloses a parameter control system and method for tryptophan purification from peanut sprouts. The method includes: acquiring the frequency shift of the chromatography column by ultrasound, analyzing the flow velocity, and obtaining the flow field uniformity and eddy current intensity based on the flow velocity analysis; constructing an optogenetic probe, activating it with blue light, and acquiring the initial fluorescence intensity; when the flow field uniformity is less than the uniformity threshold, compensating the initial fluorescence intensity with eddy current intensity to obtain the fluorescence intensity; simultaneously acquiring temperature and ORP potential to construct a separation kinetic model; obtaining the enantiomeric impurity concentration to determine the process stage; setting a two-stage optimization target according to the process stage, optimizing the purification parameters by combining the separation kinetic model, and generating corresponding control commands by combining the control barrier function. This invention achieves precise control and green and efficient production of the tryptophan purification process through interdisciplinary multiphysics modeling and intelligent optimization framework.

Description

Technical Field

[0001] This invention relates to the field of parameter control technology, and more specifically, to a parameter control system and method for purifying tryptophan from peanut sprouts. Background Technology

[0002] Tryptophan, an important amino acid, has wide applications in food, medicine, and feed. Purifying tryptophan from peanut sprouts has significant economic value and market potential. However, traditional purification processes focus only on a single chemical separation process, completely ignoring the synergistic effects of dynamic changes in the flow field and the oxidation environment on tryptophan separation. This results in the inability to quantify the oxidation loss and decreased mass transfer efficiency caused by the coupling of these two factors. Simultaneously, traditional detection technologies rely on offline HPLC and other methods, which have long detection cycles and cannot be dynamically matched with process stages. This makes it difficult to provide real-time feedback on tryptophan concentration and impurity distribution, leading to a lag in the switching between impurity elution and collection periods, further exacerbating low separation efficiency and insufficient purity. Furthermore, existing optimization algorithms only target a single objective (such as purity or time), failing to consider the cross-influence of parameters such as flow rate and temperature on multiple physical fields, and lack dynamic safety constraints. This prevents adaptive parameter control when the flow field is abnormal or the oxidation risk increases, resulting in a persistently high tryptophan oxidation loss rate, severely restricting purification efficiency and product quality.

[0003] In view of this, the present invention proposes a parameter control system and method for tryptophan purification in peanut sprouts to solve the above problems. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a method for controlling the purification parameters of tryptophan in peanut sprouts, comprising the following steps:

[0005] The flow velocity is obtained by collecting the frequency shift of the chromatography column by ultrasound, and the flow field uniformity and eddy intensity are obtained by analyzing the flow velocity.

[0006] An optogenetic probe was constructed and fixed to the biomembrane at the injection port of a simulated moving bed. The probe was activated by blue light to collect the initial fluorescence intensity. When the flow field uniformity was less than the uniformity threshold, the fluorescence intensity was compensated by the eddy current intensity to obtain the fluorescence intensity.

[0007] Temperature and ORP potential were collected simultaneously, and a separation kinetic model was constructed by combining fluorescence intensity, activation energy, and diffusion penalty term.

[0008] Obtain the enantiomeric impurity concentration and determine the process stage by combining fluorescence intensity and ORP potential;

[0009] Based on the process stage, a two-stage optimization objective is set, and constraints are set. Rolling time-domain optimization is performed in conjunction with the separation kinetics model to obtain optimized purification parameters. The corresponding control commands are then generated in conjunction with the control barrier function.

[0010] Furthermore, the diffusion penalty term is calculated based on the second partial derivative of fluorescence intensity along the axial direction of the chromatography column. The diffusion penalty term is obtained by combining the molecular diffusion coefficient, eddy current intensity, and cross-sectional area of ​​the chromatography column with empirical constants.

[0011] Furthermore, methods for determining the process stage include:

[0012] When the sum of the ratio of fluorescence intensity to enantiomeric impurity concentration and the absolute value of the first partial derivative of fluorescence intensity along the column axis is less than or equal to the dynamic threshold, it is determined that the column is in the impurity elution period.

[0013] When the ratio of fluorescence intensity to enantiomeric impurity concentration is greater than a preset threshold and the ORP potential is greater than a dynamic threshold, it is determined that the body is in the tryptophan collection period.

[0014] The dynamic threshold is calculated based on the real-time ORP potential and the preset potential threshold.

[0015] Furthermore, methods for obtaining optimized purification parameters include:

[0016] For the impurity elution period, the flow rate is integrated over the elution time period to obtain the time cost. Combined with the flow field penalty term, the first objective is to minimize the sum of the time cost and the flow field penalty term. Constraints are set on the separation kinetic model, and the value of the separation kinetic model is less than or equal to the upper limit of the elution rate. The purification parameters are obtained by optimization based on the fast gradient method.

[0017] For the tryptophan collection period, the ratio of fluorescence intensity to enantiomeric impurity concentration is calculated, the product of the ratio and the oxidation penalty factor is calculated, and the difference between the product and the diffusion penalty term is calculated, with maximizing the difference as the second objective; flow field uniformity constraints are set; and purification parameters are obtained by optimization based on adaptive simulated annealing.

[0018] Furthermore, methods for obtaining the flow field penalty term include:

[0019] Determine whether the uniformity index is less than a preset index threshold. If so, calculate the first penalty function based on the uniformity index. If not, the value of the first penalty function is 0.

[0020] Determine whether the eddy current intensity is greater than a preset intensity threshold. If so, calculate the second penalty function based on the eddy current intensity and the preset intensity threshold. If not, the value of the second penalty function is 0.

[0021] The flow field penalty term is obtained by weighted summation of the first penalty function and the second penalty function.

[0022] Furthermore, methods for obtaining purification parameters based on the fast gradient method include:

[0023] The gradient of the first objective with respect to the purification parameters is calculated using the finite difference method;

[0024] The purification parameters, including flow rate and temperature, are updated according to the gradient descent direction. During the update process, it is checked whether the updated purification parameters meet the constraints. If not, the purification parameters are adjusted until the constraints are met.

[0025] Calculate the difference between the current first target value and the first target value of the previous iteration to obtain the first difference. If the first difference is less than the preset difference threshold, the algorithm is considered to have converged and the optimized purification parameters are obtained. Otherwise, return to continue calculating the gradient and updating the purification parameters.

[0026] Furthermore, methods for obtaining the oxidation penalty factor include:

[0027] The oxidation penalty coefficient is pre-trained, the ORP potential is collected, and the oxidation penalty factor is calculated by combining it with a preset potential threshold.

[0028] Furthermore, methods for obtaining purification parameters based on adaptive simulated annealing include:

[0029] Based on the current solution of the purified parameters, a neighborhood solution is generated through random perturbation;

[0030] Calculate the second objective value for the current solution and the neighborhood solution respectively, and calculate the difference between the second objective value of the neighborhood solution and the second objective value of the current solution to obtain the second difference. If the second difference is greater than 0, accept the neighborhood solution and update the current solution to the neighborhood solution; if the second difference is less than or equal to 0, calculate the acceptance probability P according to the Metropolis criterion, combined with the second difference and the current temperature, and generate a random number p between [0,1]. If the random number p is less than P, accept the neighborhood solution and update the current solution; otherwise, keep the current solution unchanged.

[0031] After L iterations at each temperature, the purification parameters are updated according to the parameter change rate r. When the updated purification parameters are less than or equal to the preset parameter change threshold, the algorithm is considered to have converged and the optimized purification parameters are obtained. Otherwise, the algorithm returns to continue generating neighborhood solutions and making judgments.

[0032] Furthermore, methods for generating corresponding control commands by combining control barrier functions include:

[0033] When the eddy current intensity is greater than the intensity barrier threshold or the ORP potential value is greater than the potential barrier threshold, the rate of change of the purification parameters is controlled to be less than or equal to the corresponding rate of change threshold.

[0034] Furthermore, methods for obtaining enantiomeric impurity concentrations include:

[0035] Prepare N sets of enantiomeric impurity standard solutions with known concentrations, covering all concentration ranges of enantiomeric impurities appearing in the sample. According to the preset chromatographic conditions, perform injection analysis in sequence, record the chromatographic peak area of ​​each standard solution, plot the standard curve with the concentration of enantiomeric impurities as the abscissa and the corresponding peak area as the ordinate, and obtain the standard curve equation through linear regression.

[0036] The pretreated peanut sprout tryptophan extract sample was injected into a pre-set HPLC system for detection, and the chromatogram was recorded. Based on the retention time of the chromatographic peaks in the chromatogram, the chromatographic peaks of enantiomeric impurities were determined.

[0037] The chromatographic peaks of enantiomeric impurities are integrated to obtain the peak areas. The peak areas are then substituted into the standard curve equation to calculate the concentration of enantiomeric impurities in the sample.

[0038] Furthermore, methods for obtaining flow rate include:

[0039] The flow velocity is calculated based on the frequency shift, the angle between the ultrasonic beam and the flow direction, and the sound velocity.

[0040] Methods for obtaining flow field uniformity and eddy intensity include:

[0041] The average velocity and standard deviation of the velocity are obtained by statistical analysis, and the flow field uniformity is calculated based on the average velocity and standard deviation of the velocity.

[0042] The velocity curl is calculated by taking the partial derivative of the velocity component based on the flow velocity. The magnitude of the velocity curl is then calculated based on the vector magnitude. Finally, the eddy intensity is obtained by integrating the magnitude of the velocity curl over the cross-sectional area of ​​the chromatography column.

[0043] A parameter control system for tryptophan purification in peanut sprouts, comprising the following steps:

[0044] First analysis module: The frequency shift of the chromatography column is acquired by ultrasonic waves, the flow velocity is analyzed, and the flow field uniformity and eddy intensity are obtained based on the flow velocity analysis.

[0045] The second analysis module constructs an optogenetic probe, fixes the probe to the biomembrane at the injection port of the simulated moving bed, activates it with blue light, and collects the initial fluorescence intensity; when the flow field uniformity is less than the uniformity threshold, the fluorescence intensity is obtained by compensating the initial fluorescence intensity with eddy current intensity.

[0046] Model building module: Simultaneously collect temperature and ORP potential, and construct a separation kinetic model by combining fluorescence intensity, activation energy and diffusion penalty term;

[0047] Stage determination module: Obtains enantiomeric impurity concentration and determines the process stage by combining fluorescence intensity and ORP potential;

[0048] Parameter optimization module: Set two-stage optimization objectives and constraints according to the process stage, perform rolling time-domain optimization in combination with the separation kinetics model to obtain optimized purification parameters, and generate corresponding control commands in combination with the control barrier function.

[0049] The technical effects and advantages of the control system and method for purifying tryptophan in peanut sprouts according to the present invention are as follows:

[0050] This invention reduces fluorescence detection errors and precisely solves the detection distortion problem caused by flow field fluctuations by using real-time ultrasonic acquisition of flow field parameters and combining it with the eddy current intensity compensation mechanism of optogenetic probes. It constructs a separation kinetic model and, combined with two-stage rolling time-domain optimization and a control barrier function, achieves quantitative control of multi-physics coupling effects, reducing tryptophan oxidation loss and improving tryptophan purity, overcoming the technical bottleneck of traditional single-parameter optimization's inability to handle multi-field synergistic effects. Furthermore, it shortens the switching delay between impurity elution and collection periods through a process stage judgment method that integrates HPLC standard curves and multi-parameter dynamic thresholds. Combined with two-stage optimization using adaptive simulated annealing algorithm and fast gradient method, it improves purification rate while reducing energy consumption. Through an interdisciplinary multi-physics modeling and intelligent optimization framework, this invention systematically solves problems such as detection lag, model inaccuracy, high oxidation loss, and difficulty in parameter coupling control in existing technologies, achieving precise control and green, efficient production of the tryptophan purification process. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of a method for controlling the purification parameters of tryptophan in peanut sprouts according to the present invention.

[0052] Figure 2 This is a schematic diagram of the data flow in this invention;

[0053] Figure 3 This is a schematic diagram of the separation kinetics model construction process of the present invention;

[0054] Figure 4 This is a schematic diagram of the control system for the purification parameters of tryptophan in peanut sprouts according to the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Example 1

[0057] Please see Figure 1 As shown in the figure, this embodiment provides a method for controlling the purification parameters of tryptophan in peanut sprouts, including the following steps:

[0058] The flow velocity is obtained by collecting the frequency shift of the chromatography column by ultrasound, and the flow field uniformity and eddy intensity are obtained by analyzing the flow velocity.

[0059] Methods for obtaining flow rate include:

[0060] The flow velocity is calculated based on the frequency shift, the angle between the ultrasonic beam and the flow direction, and the sound velocity. For example... ;in, The flow rate of the fluid inside the chromatography column; This represents the frequency shift acquired by the ultrasonic wave. The speed at which ultrasound propagates in a fluid; The frequency at which the ultrasound is emitted; The angle between the ultrasonic beam and the direction of the flow velocity;

[0061] Methods for obtaining flow field uniformity and eddy intensity include:

[0062] The average velocity and standard deviation of the velocity are obtained by statistical analysis, and the flow field uniformity is calculated based on the average velocity and standard deviation of the velocity. ;in, For process uniformity; The average flow velocity during the statistical period; The standard deviation of the flow rate;

[0063] The velocity curl is calculated by taking the partial derivative of the velocity component based on the flow velocity. The magnitude of the velocity curl is then calculated based on the vector magnitude. Finally, the eddy intensity is obtained by integrating the magnitude of the velocity curl over the cross-sectional area of ​​the chromatography column.

[0064] By acquiring the frequency shift of the chromatography column using ultrasound and analyzing the flow velocity accordingly, and then calculating the flow field uniformity based on flow velocity statistics to obtain the average flow velocity and standard deviation, the eddy current intensity is obtained by calculating the partial derivatives of the flow velocity components and combining them with the vector modulus integral. This allows for real-time and non-invasive acquisition of dynamic flow field parameters. During the impurity elution period, flow field uniformity and eddy current intensity can be used to construct a flow field penalty term, which, along with time cost, serves as an optimization objective. The flow velocity and temperature are adjusted using the fast gradient method to ensure efficient impurity elution while avoiding a decrease in mass transfer efficiency caused by flow field anomalies. During the tryptophan collection period, these parameters can be embedded into the separation kinetics model and combined with redox potential to construct a linked penalty function. The parameter combination is optimized using an adaptive simulated annealing algorithm to suppress the oxidation risk and diffusion loss caused by eddies. This achieves multiple objectives, including improving tryptophan purity, reducing oxidation loss, and optimizing process energy consumption. It effectively solves the problems of flow field control lag and difficulty in multi-parameter coupling optimization in traditional purification methods, providing a quantitative basis for accurate judgment and dynamic optimization of the process stages.

[0065] An optogenetic probe was constructed and fixed to the biofilm at the inlet of a simulated moving bed. Activated by blue light, the initial fluorescence intensity was collected. When the flow field homogeneity was less than a homogeneity threshold, determined through a flow field-fluorescence signal calibration experiment: a fluid containing the target analyte and the same biofilm probe as the actual process was selected. A gradient flow field homogeneity environment was constructed at the inlet of the simulated moving bed, which could be achieved by adjusting the eddy current intensity using a fluid turbulence diffuser. The fluorescence signal deviation rate was collected under different homogeneities and compared with the ideal flow field fluorescence value. When the deviation rate first exceeded the process tolerance (e.g., 5% in tryptophan purification), the corresponding flow field homogeneity value was taken as the homogeneity threshold. The fluorescence intensity was obtained by compensating for the initial fluorescence intensity with eddy current intensity. The optogenetic probe was constructed and immobilized on the biomembrane at the injection port of the simulated moving bed, and the initial fluorescence intensity was collected after activation by blue light. The flow field uniformity U is calculated in real time using the flow field monitoring module. When U is less than the experimentally determined threshold, the eddy current intensity Ω is substituted into the compensation model. Calculate compensation amount ,in, The flow-fluorescence response coefficient was determined by probe calibration experiments. The base compensation intercept is used to offset the background bias under a uniform flow field; the final compensation fluorescence intensity is... This addresses the signal distortion problem caused by flow field fluctuations.

[0066] An optogenetic probe is constructed and fixed to a biofilm at the injection port of a simulated moving bed. Initial fluorescence intensity is collected via blue light activation. When the flow field uniformity is less than a uniformity threshold, eddy current intensity is used to compensate for the initial fluorescence intensity. This technical solution addresses the key problem of fluorescence detection signal distortion caused by flow field fluctuations in existing technologies. Traditional optogenetic detection methods do not consider the interference of flow field inhomogeneity on probe distribution and fluorescence signals, and are prone to fluorescence intensity measurement deviations due to fluid particle mixing caused by eddies, leading to misjudgment of tryptophan concentration and incorrect switching of process stages. This invention establishes a dynamic mapping relationship between eddy current intensity and fluorescence compensation, correcting the impact of flow field anomalies on the signal in real time and improving fluorescence detection accuracy. This invention breaks through the conventional thinking of those skilled in the art who only focus on the chemical properties of the probe, coupling fluid dynamic parameters with biological detection signals for the first time. Through interdisciplinary technology integration, it achieves a fundamental improvement in detection accuracy. Given that existing technologies have not solved the problem of detection distortion caused by flow field interference, this invention fits an eddy current compensation model through extensive experiments and creatively introduces fluid dynamic parameters into the biological detection correction system.

[0067] Temperature and ORP potential were simultaneously acquired, and a separation kinetic model was constructed by combining fluorescence intensity, activation energy, and diffusion penalty term; for details, please refer to [reference needed]. Figure 3 The diffusion penalty term is calculated based on the second partial derivative of fluorescence intensity along the column axis: (e.g.) ,in, The coordinates of the chromatography column are axial coordinates. The axial sampling interval; To determine the axial coordinate position of the chromatography column , Fluorescence intensity collected at any given time; To determine the axial coordinate position of the chromatography column , Fluorescence intensity collected at any given time; To determine the axial coordinate position of the chromatography column , Fluorescence intensity collected at various times; measured by molecular diffusion coefficient (Available from a table) Eddy current intensity Cross-sectional area of ​​chromatography column and the effective length of the chromatography column The effective diffusion impact factor is calculated, such as the effective diffusion impact factor. Combined with empirical constants The value is typically between 0.5 and 1.5. The diffusion penalty term is calculated, such as the diffusion penalty term. The goal was to minimize the mean square error of fluorescence intensity, and other correlation coefficients were fitted using the least squares method. The activation energy was obtained by preliminary experiments combining tryptophan with a chiral stationary phase.

[0068] The above method quantifies the impact of axial diffusion on concentration distribution by introducing a diffusion penalty term, combines the ORP potential to reflect oxidation risk in real time, and couples temperature, flow field parameters and biosensor signals across scales, enabling the model to dynamically predict concentration changes, oxidation losses and diffusion losses during tryptophan separation, thereby reducing prediction errors. It incorporates fluid dynamics parameters, electrochemical parameters and biosensor signals into a unified modeling framework, and forms a multi-physics collaborative dynamic model by experimentally fitting the diffusion penalty term and activation energy parameters, thus solving the problems of insufficient model accuracy and inability to adapt to flow field fluctuations in existing technologies.

[0069] The concentration of enantiomeric impurities is obtained by detection and analysis, and the process stage is determined by combining fluorescence intensity and ORP potential.

[0070] Methods for obtaining enantiomeric impurity concentrations include:

[0071] Prepare N sets of enantiomeric impurity standard solutions with known concentrations, covering all concentration ranges of enantiomeric impurities appearing in the sample. According to the preset chromatographic conditions, perform injection analysis in sequence, record the chromatographic peak area of ​​each standard solution, plot the standard curve with the concentration of enantiomeric impurities as the abscissa and the corresponding peak area as the ordinate, and obtain the standard curve equation through linear regression.

[0072] The pretreated peanut sprout tryptophan extract sample was injected into a pre-set HPLC system for detection, and the chromatogram was recorded. Based on the retention time of the chromatographic peaks in the chromatogram, the chromatographic peaks of enantiomers were determined. Specifically, the interaction forces between tryptophan and enantiomers with the chiral chromatographic column stationary phase are different, resulting in different retention behaviors in the chromatographic column. Based on the retention time of tryptophan and the theoretical difference in the retention times of the two, the chromatographic peaks of enantiomers can be identified in the chromatogram.

[0073] The chromatographic peaks of enantiomeric impurities are integrated to obtain the peak areas. The peak areas are then substituted into the standard curve equation to calculate the concentration of enantiomeric impurities in the sample.

[0074] By preparing multiple sets of enantiomeric impurity standard solutions with known concentrations and constructing standard curves, HPLC combined with a chiral column was used to detect tryptophan extract from peanut sprouts. The difference in retention time between tryptophan and enantiomeric impurities due to the difference in interaction forces on the stationary phase was used to accurately locate the impurity chromatographic peaks, thereby calculating the enantiomeric impurity concentration. This concentration was then combined with fluorescence intensity (reflecting tryptophan concentration) and ORP potential (characterizing oxidation state) to dynamically and accurately determine the process stage. When the ratio of fluorescence intensity to enantiomeric impurity concentration and the sum of the absolute values ​​of the first-order partial derivatives of fluorescence intensity are less than or equal to a dynamic threshold, it is determined to be the impurity elution period. When the ratio is greater than a preset threshold and the ORP potential is greater than the dynamic threshold, it is determined to be the tryptophan collection period. This method can dynamically switch optimization targets based on real-time monitoring of impurity concentration, tryptophan concentration, and oxidation potential. During the elution period, the goal is to minimize the sum of time cost and flow field penalty term for efficient impurity elution. During the collection period, the goal is to maximize the purity correction value to inhibit oxidation and improve purity. (e.g., purity correction value) ,in, To calculate the peak area ratio of tryptophan by integrating chiral chromatographic peaks using HPLC; To utilize the standard curve, the real-time fluorescence intensity is used to extrapolate the theoretical purity, thereby achieving precise control of the entire process, effectively improving the yield and purity of tryptophan, and reducing oxidation loss and impurity residue.

[0075] Methods for determining the process stage include:

[0076] When the sum of the ratio of fluorescence intensity to enantiomeric impurity concentration and the absolute value of the first partial derivative of fluorescence intensity along the column axis is less than or equal to the dynamic threshold, it is determined that the column is in the impurity elution period.

[0077] When the ratio of fluorescence intensity to enantiomeric impurity concentration is greater than a preset threshold and the ORP potential is greater than a dynamic threshold, it is determined that the process is in the tryptophan collection phase. The preset threshold is determined by collecting fluorescence intensity, enantiomeric impurity concentration, and ORP potential data at different process stages (impurity elution phase and tryptophan collection phase) through a full-process simulation experiment, plotting the distribution of fluorescence intensity / enantiomeric impurity concentration and ORP potential, and taking the data boundary point of the elution-collection phase as the preset threshold.

[0078] The dynamic threshold is calculated based on the real-time ORP potential and a preset potential threshold. For example, the dynamic threshold... ,in, The preset reference potential is determined experimentally, such as the tryptophan stability region potential of 300mV; The response coefficient, with a value of 0.8-1.2, is used to adjust the sensitivity of the threshold to potential fluctuations. This is the real-time ORP potential. The dynamic threshold can also be adaptively adjusted according to the actual situation when the impurities are different.

[0079] Two-stage optimization objectives are set according to the process stages, and constraints are set to limit the adjustment range of purification parameters (including flow rate, temperature, and OPR) to avoid flow field / oxidation runaway; such as ,in, The reference flow rate; For real-time flow rate; such as ,in, The optimal temperature for this stage; This is the real-time temperature; if adjusting parameters, To avoid drastic oxidation caused by potential abrupt changes; rolling time-domain optimization is performed using a separation kinetics model to obtain optimized purification parameters, and corresponding control commands are generated by combining the control barrier function.

[0080] This method determines the process stage by comparing the ratio of fluorescence intensity to enantiomeric impurity concentration, the sum of the absolute values ​​of the first-order partial derivatives of fluorescence intensity along the column axis, and a dynamic threshold. It also considers whether this ratio meets preset conditions in conjunction with the ORP potential. The dynamic threshold can be flexibly adjusted based on real-time ORP potential, preset potential thresholds, and impurity characteristics. This allows for precise and dynamic identification of the impurity elution period and tryptophan collection period, providing an effective solution to the problems of delayed stage judgment and fixed thresholds lacking adaptability in traditional processes. Based on the judgment results, a two-stage optimization objective is set, prioritizing time cost and flow rate optimization during the impurity elution period. By minimizing the sum of field penalty terms and incorporating constraints from the separation kinetics model, and maximizing the purity correction value during the collection period while setting constraints such as flow field uniformity, the purification parameters are dynamically adjusted through rolling time-domain optimization. Furthermore, by combining this with a control barrier function to ensure that parameter changes remain within a safe range and generating control commands, the above method achieves closed-loop control of the entire process from precise stage judgment to dynamic parameter optimization. This effectively solves technical problems such as incomplete impurity elution, low purity of tryptophan collection, lag in parameter control, and system operation risks during the purification process, significantly improving the efficiency, purity, and process stability of tryptophan purification.

[0081] Methods for obtaining optimized purification parameters include:

[0082] For the impurity elution period, the flow rate is integrated over the elution time period to obtain the time cost. Combined with the flow field penalty term, the first objective is to minimize the sum of the time cost and the flow field penalty term. Constraints are set on the separation kinetic model, specifically, the value of the separation kinetic model is less than or equal to the upper limit of the elution rate. The purification parameters are obtained by optimization based on the fast gradient method.

[0083] Methods for obtaining flow field penalty terms include:

[0084] Determine whether the uniformity index is less than the preset index threshold. If it is, calculate the first penalty function based on the uniformity index; otherwise, the value of the first penalty function is 0. The preset index threshold is determined by a flow field calibration experiment. A tracer with properties consistent with the tryptophan solution is injected into a simulated moving bed, and the axial velocity distribution under different flow field disturbances is collected to calculate the uniformity index. When the flow field uniformity decreases, resulting in a 5% reduction in the purity of tryptophan separation, the uniformity index at this time is recorded as the preset index threshold.

[0085] Determine whether the eddy current intensity is greater than the preset intensity threshold. If so, calculate the second penalty function based on the eddy current intensity and the preset intensity threshold. If not, the value of the second penalty function is 0. The preset intensity threshold is determined by generating a gradient eddy current intensity Ω in the same simulated moving bed through a fluid turbulence generator, collecting chromatographic peak broadening data under different eddy current intensities, and using the corresponding eddy current intensity as the preset intensity threshold when the half-peak width increases by 20%.

[0086] The flow field penalty term is obtained by weighted summation of the first penalty function and the second penalty function. The weighting coefficients are determined by experiments or experience based on specific process requirements and flow field characteristics.

[0087] Methods for obtaining purification parameters based on the fast gradient method include:

[0088] The gradient of the first objective with respect to the purification parameters is calculated using the finite difference method;

[0089] The purification parameters, including flow rate and temperature, are updated according to the gradient descent direction. During the update process, it is checked whether the updated purification parameters meet the constraints. If not, the purification parameters are adjusted until the constraints are met.

[0090] Calculate the difference between the current first target value and the first target value of the previous iteration to obtain the first difference. If the first difference is less than the preset difference threshold, the algorithm is considered to have converged, and the optimized purification parameters are obtained. Otherwise, return to continue calculating the gradient and updating the purification parameters. In the simulation experiment of tryptophan purification, the convergence process of the first target value during the fast gradient method iteration is recorded. The first target value is such as the purity correction value, yield, etc. When the iteration is stable, the difference between the target values ​​of adjacent iterations with process parameter fluctuations of <1% is statistically analyzed, and the upper limit of its 95% confidence interval is taken as the preset difference threshold.

[0091] For the tryptophan collection period, the ratio of fluorescence intensity to enantiomeric impurity concentration is calculated, the product of the ratio and the oxidation penalty factor is calculated, and the difference between the product and the diffusion penalty term is calculated, with maximizing the difference as the second objective; flow field uniformity constraints are set; and purification parameters are obtained by optimization based on adaptive simulated annealing.

[0092] Methods for obtaining oxidation penalty factors include:

[0093] Pre-trained oxidation penalty coefficients, such as those fitted to purification experiments at different polyphenol concentrations, involve collecting ORP potentials and calculating the oxidation penalty factor based on a preset potential threshold; for example, the oxidation penalty factor... ,in, The oxidation potential of tryptophan is low, as determined experimentally, such as 250 mV. This is the potential at which oxidation loss increases significantly, such as 400mV;

[0094] Methods for obtaining purification parameters based on adaptive simulated annealing include:

[0095] Based on the current solution of the purified parameters, a neighborhood solution is generated through random perturbation;

[0096] Calculate the second objective value for the current solution and the neighborhood solution respectively, and calculate the difference between the second objective value of the neighborhood solution and the second objective value of the current solution to obtain the second difference. If the second difference is greater than 0, accept the neighborhood solution and update the current solution to the neighborhood solution; if the second difference is less than or equal to 0, calculate the acceptance probability P according to the Metropolis criterion, combined with the second difference and the current temperature, and generate a random number p between [0,1]. If the random number p is less than P, accept the neighborhood solution and update the current solution; otherwise, keep the current solution unchanged.

[0097] After L iterations at each temperature, the purification parameters are updated according to the parameter change rate r. When the updated purification parameters are less than or equal to the preset parameter change threshold, the algorithm is considered to have converged and the optimized purification parameters are obtained. Otherwise, the algorithm returns to generate neighborhood solutions and makes a judgment. The preset parameter change threshold is determined by perturbation experiments to test the impact of the parameter change rate on process stability. The critical value at which the flow field fluctuation / oxidation rate begins to exceed the limit is taken as the preset parameter change threshold.

[0098] By minimizing the sum of time cost and flow field penalty term during the impurity elution period, and combining separation kinetics model constraints with the fast gradient method to optimize flow velocity and temperature, while calculating the flow field penalty term based on the homogeneity index and eddy current intensity to suppress flow field anomalies, this method solves the problem of incomplete impurity elution or flow field disturbance caused by unreasonable flow velocity in traditional elution processes, achieving a balance between efficient elution and flow field stability. During the tryptophan collection period, the goal is to maximize the difference between the ratio of fluorescence intensity to enantiomeric impurity concentration combined with the oxidation penalty factor and diffusion penalty term. This is achieved using an adaptive simulated annealing algorithm with flow field homogeneity constraints, and through pre-training... By dynamically adjusting the oxidation penalty factor in conjunction with the ORP potential, the oxidation penalty factor can be precisely addressed to mitigate the risk of tryptophan oxidation and diffusion loss during the collection period, resolving the contradiction between purity improvement and oxidation loss in traditional methods. The entire optimization method achieves precise parameter control throughout the entire process from impurity elution to tryptophan collection through dynamic adaptation of a two-stage objective function and constraints, combined with efficient search using the fast gradient method and adaptive simulated annealing algorithm. This effectively solves the technical problems in traditional purification processes, such as single optimization objective, difficulty in parameter coupling and control, high oxidation loss, and unstable flow field, significantly improving tryptophan yield, purity, and process stability.

[0099] Methods for generating corresponding control commands by combining control barrier functions include:

[0100] When the eddy current intensity exceeds the intensity barrier threshold or the ORP potential value exceeds the potential barrier threshold, the rate of change of the purification parameters is controlled to be less than or equal to the corresponding rate of change threshold. Specifically, in the pilot-scale unit for tryptophan purification, the eddy current intensity is gradually increased, such as by adjusting it with a flow disruptor. The eddy current intensity value when the flow field fluctuation causes the chromatographic peak broadening to be ≥15%, which exceeds the separation accuracy requirement, is collected as the intensity barrier threshold. Through oxidation loss experiments, the oxidation rate of tryptophan at different ORP potentials is recorded. When the oxidation rate is ≥3% / h, which exceeds the allowable value of process loss, the corresponding ORP potential is used as the potential barrier threshold. Under stable process conditions, the maximum safe rate of change of the test parameters (flow rate, temperature, etc.) is measured. When the parameter change rate causes a sudden change in the ratio of flow field fluctuation to oxidation rate, 80% of this critical rate of change is taken, i.e., a safety margin is reserved as the rate of change threshold.

[0101] By controlling the rate of change of purification parameters within the corresponding threshold when the eddy current intensity exceeds the intensity barrier threshold or the ORP potential value exceeds the potential barrier threshold using a control barrier function, the process stability problem caused by drastic flow field fluctuations or increased oxidation risk during tryptophan purification can be effectively solved. When the eddy current intensity exceeds the limit, limiting the rate of change of parameters such as flow rate and temperature can prevent further flow field disturbance, prevent a decrease in mass transfer efficiency and chromatographic peak broadening, and ensure the separation effect of impurity elution and tryptophan collection. When the ORP potential exceeds the limit, constraining the adjustment range of parameters can inhibit the aggravation of oxidation reaction, reduce tryptophan oxidation loss, and avoid system operation risks caused by parameter mutations. Thus, a safety boundary is set for process parameters during dynamic optimization, achieving robust control of the purification process, ensuring the stability of tryptophan yield and purity, and solving the technical problem of lacking a dynamic protection mechanism for abnormal operating conditions in traditional control methods.

[0102] Example 2

[0103] This embodiment proposes a second objective optimization method applied to Embodiment 1, including the following steps:

[0104] In the second objective, an eddy current-oxidation linkage penalty term is added. The ratio of fluorescence intensity to enantiomeric impurity concentration is calculated, the product of the ratio and the oxidation penalty factor is calculated, the difference between the product and the diffusion penalty term is calculated, and the difference between the difference and the eddy current-oxidation linkage penalty term is calculated to obtain the optimized difference, so as to maximize the optimized difference as the second objective.

[0105] Methods for obtaining the eddy current-oxidation linkage penalty term include:

[0106] The eddy current-oxidation linkage penalty term is calculated by eddy current intensity, preset intensity threshold, ORP potential deviation value (obtained by subtracting the preset potential threshold from the real-time collected ORP value), and maximum allowable deviation value of ORP potential (obtained through pre-experimentation). The weight coefficients involved are determined by experimental fitting or optimization using the NSGA-Ⅲ algorithm.

[0107] In the second objective, an eddy current-oxidation linkage penalty term is added. This is achieved by multiplying the ratio of fluorescence intensity to enantiomeric impurity concentration by an oxidation penalty factor, subtracting a diffusion penalty term, and then subtracting the linkage penalty term calculated based on eddy current intensity, a preset intensity threshold, ORP potential difference, and the maximum allowable deviation. The goal is to maximize the optimization difference, which can accurately quantify the synergistic effect of eddy current intensity and redox potential during the tryptophan collection period. This addresses the problem that traditional optimization objectives do not consider the multi-physics coupling effect. The eddy current-oxidation linkage penalty term dynamically adjusts the penalty intensity of eddy current and oxidation deviation through weighting coefficients. When eddy current abnormalities or oxidation risk increases, the value of the penalty term increases significantly, driving the optimization algorithm to prioritize the parameter combination of reducing flow velocity to weaken eddy current and adjusting temperature to improve ORP potential. This suppresses tryptophan degradation caused by eddy current-induced oxidation. At the same time, the control barrier function limits the parameter change rate to avoid abrupt flow field changes caused by drastic parameter adjustments. This achieves multiple objectives of improving tryptophan purity, reducing oxidation loss, and enhancing process stability during the collection period, effectively solving the problems of disconnect between purity optimization and oxidation control and the difficulty of multi-parameter coupling and regulation in traditional methods.

[0108] Example 3

[0109] Please see Figure 4 As shown, this embodiment provides a parameter control system for tryptophan purification in peanut sprouts, including:

[0110] First analysis module: The frequency shift of the chromatography column is acquired by ultrasonic waves, the flow velocity is analyzed, and the flow field uniformity and eddy intensity are obtained based on the flow velocity analysis.

[0111] The second analysis module constructs an optogenetic probe, fixes the probe to the biomembrane at the injection port of the simulated moving bed, activates it with blue light, and collects the initial fluorescence intensity; when the flow field uniformity is less than the uniformity threshold, the fluorescence intensity is obtained by compensating the initial fluorescence intensity with eddy current intensity.

[0112] Model building module: Simultaneously collect temperature and ORP potential, and construct a separation kinetic model by combining fluorescence intensity, activation energy and diffusion penalty term;

[0113] Stage determination module: Obtains enantiomeric impurity concentration and determines the process stage by combining fluorescence intensity and ORP potential;

[0114] Parameter optimization module: Set two-stage optimization objectives and constraints according to the process stage, perform rolling time-domain optimization in combination with the separation kinetics model to obtain optimized purification parameters, and generate corresponding control commands in combination with the control barrier function.

[0115] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0116] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for controlling parameters in the purification of tryptophan from peanut sprouts, characterized in that, Includes the following steps: The flow velocity is obtained by collecting the frequency shift of the chromatography column by ultrasound, and the flow field uniformity and eddy intensity are obtained by analyzing the flow velocity. An optogenetic probe was constructed, fixed to the biomembrane at the injection port of a simulated moving bed, and activated by blue light to collect the initial fluorescence intensity. When the flow field uniformity is less than the uniformity threshold, the fluorescence intensity is obtained by compensating the initial fluorescence intensity with eddy current intensity. Temperature and ORP potential were collected simultaneously, and a separation kinetic model was constructed by combining fluorescence intensity, activation energy, and diffusion penalty term. The diffusion penalty term was calculated based on the second partial derivative of fluorescence intensity along the column axis, molecular diffusion coefficient, eddy current intensity, cross-sectional area of ​​the column, and effective length of the column, combined with empirical constants. Obtain the enantiomeric impurity concentration and determine the process stage by combining fluorescence intensity and ORP potential: when the sum of the ratio of fluorescence intensity to enantiomeric impurity concentration and the absolute value of the first partial derivative of fluorescence intensity along the column axis is less than or equal to the dynamic threshold, it is determined that the process is in the impurity elution period. When the ratio of fluorescence intensity to enantiomeric impurity concentration is greater than a preset threshold and the ORP potential is greater than a dynamic threshold, it is determined that the body is in the tryptophan collection period. Based on the process stages, a two-stage optimization objective is set, and constraints are defined. Rolling time-domain optimization is performed using a separation kinetics model to obtain optimized purification parameters: For the impurity elution period, it is determined whether the uniformity index is less than a preset index threshold. If so, a first penalty function is calculated based on the uniformity index; otherwise, the first penalty function value is 0. It is also determined whether the eddy current intensity is greater than a preset intensity threshold. If so, a second penalty function is calculated based on the eddy current intensity and the preset intensity threshold; otherwise, the second penalty function value is 0. The first and second penalty functions are weighted and summed to obtain the flow field penalty term. The flow velocity is integrated over the elution time period to obtain the time cost. The first objective is to minimize the sum of the time cost and the flow field penalty term. Set constraints for the separation kinetics model, where the value of the separation kinetics model is less than or equal to the upper limit of the elution rate; Purification parameters are obtained by optimization based on the fast gradient method; For the tryptophan collection period, the eddy current-oxidation linkage penalty term is calculated by eddy current intensity, preset intensity threshold, ORP potential deviation value, and maximum allowable deviation value of ORP potential; the oxidation penalty coefficient is pre-trained, the ORP potential is collected, and the oxidation penalty factor is calculated by combining the preset potential threshold. The ratio of fluorescence intensity to enantiomeric impurity concentration is calculated, the product of the ratio and the oxidation penalty factor is calculated, the difference between the product and the diffusion penalty term is calculated, and the difference between the difference and the eddy current-oxidation linkage penalty term is calculated to obtain the optimized difference, with maximizing the difference as the second objective. Set flow field uniformity constraints; Purification parameters were obtained by optimization based on adaptive simulated annealing. The corresponding control commands are generated by combining the control barrier function.

2. The method for controlling the purification parameters of tryptophan in peanut sprouts according to claim 1, characterized in that, The dynamic threshold is calculated based on the real-time ORP potential and the preset potential threshold.

3. The method for controlling the purification parameters of tryptophan in peanut sprouts according to claim 1, characterized in that, Methods for obtaining purification parameters based on the fast gradient method include: The gradient of the first objective with respect to the purification parameters is calculated using the finite difference method; The purification parameters, including flow rate and temperature, are updated according to the gradient descent direction. During the update process, it is checked whether the updated purification parameters meet the constraints. If not, the purification parameters are adjusted until the constraints are met. Calculate the difference between the current first target value and the first target value of the previous iteration to obtain the first difference. If the first difference is less than the preset difference threshold, the algorithm is considered to have converged and the optimized purification parameters are obtained. Otherwise, return to continue calculating the gradient and updating the purification parameters.

4. The method for controlling the purification parameters of tryptophan in peanut sprouts according to claim 1, characterized in that, Methods for obtaining purification parameters based on adaptive simulated annealing include: Based on the current solution of the purified parameters, a neighborhood solution is generated through random perturbation; Calculate the second objective value for the current solution and the neighborhood solution respectively, and calculate the difference between the second objective value of the neighborhood solution and the second objective value of the current solution to obtain the second difference. If the second difference is greater than 0, accept the neighborhood solution and update the current solution to the neighborhood solution; if the second difference is less than or equal to 0, calculate the acceptance probability P according to the Metropolis criterion, combined with the second difference and the current temperature, and generate a random number p between [0,1]. If the random number p is less than P, accept the neighborhood solution and update the current solution; otherwise, keep the current solution unchanged. After L iterations at each temperature, the purification parameters are updated according to the parameter change rate r. When the updated purification parameters are less than or equal to the preset parameter change threshold, the algorithm is considered to have converged and the optimized purification parameters are obtained. Otherwise, the algorithm returns to continue generating neighborhood solutions and making judgments.

5. The method for controlling the purification parameters of tryptophan in peanut sprouts according to claim 1, characterized in that, Methods for generating corresponding control commands by combining control barrier functions include: When the eddy current intensity is greater than the intensity barrier threshold or the ORP potential value is greater than the potential barrier threshold, the rate of change of the purification parameters is controlled to be less than or equal to the corresponding rate of change threshold.

6. The method for controlling the purification parameters of tryptophan in peanut sprouts according to claim 1, characterized in that, Methods for obtaining enantiomeric impurity concentrations include: Prepare N sets of enantiomeric impurity standard solutions with known concentrations, covering all concentration ranges of enantiomeric impurities appearing in the sample. According to the preset chromatographic conditions, perform injection analysis in sequence, record the chromatographic peak area of ​​each standard solution, plot the standard curve with the concentration of enantiomeric impurities as the abscissa and the corresponding peak area as the ordinate, and obtain the standard curve equation through linear regression. The pretreated peanut sprout tryptophan extract sample was injected into a pre-set HPLC system for detection, and the chromatogram was recorded. Based on the retention time of the chromatographic peaks in the chromatogram, the chromatographic peaks of enantiomeric impurities were determined. The peak area of ​​the enantiomeric impurity is obtained by integrating the chromatographic peaks of the enantiomeric impurities. The peak area is then substituted into the standard curve equation to calculate the concentration of the enantiomeric impurities in the sample.

7. The method for controlling the purification parameters of tryptophan in peanut sprouts according to claim 1, characterized in that, Methods for obtaining flow rate include: The flow velocity is calculated based on the frequency shift, the angle between the ultrasonic beam and the flow direction, and the sound velocity. Methods for obtaining flow field uniformity and eddy intensity include: Statistical flow velocity is used to obtain the average flow velocity and the standard deviation of the flow velocity. The flow field uniformity is then calculated based on the average flow velocity and the standard deviation of the flow velocity. The velocity curl is calculated by taking the partial derivative of the velocity component based on the flow velocity. The magnitude of the velocity curl is then calculated based on the vector magnitude. Finally, the eddy intensity is obtained by integrating the magnitude of the velocity curl over the cross-sectional area of ​​the chromatography column.

8. A parameter control system for tryptophan purification in peanut sprouts, implementing the parameter control method for tryptophan purification in peanut sprouts as described in any one of claims 1-7, characterized in that, include: First analysis module: The frequency shift of the chromatography column is acquired by ultrasonic waves, the flow velocity is analyzed, and the flow field uniformity and eddy intensity are obtained based on the flow velocity analysis. The second analysis module: Constructing optogenetic probes, fixing the probes to the biomembrane at the injection port of the simulated moving bed, activating them with blue light, and collecting the initial fluorescence intensity; When the flow field uniformity is less than the uniformity threshold, the fluorescence intensity is obtained by compensating the initial fluorescence intensity with eddy current intensity. Model building module: Simultaneously collect temperature and ORP potential, and construct a separation kinetic model by combining fluorescence intensity, activation energy and diffusion penalty term; Stage determination module: Obtains enantiomeric impurity concentration and determines the process stage by combining fluorescence intensity and ORP potential; Parameter optimization module: Set two-stage optimization objectives and constraints according to the process stage, perform rolling time-domain optimization in combination with the separation kinetics model to obtain optimized purification parameters, and generate corresponding control commands in combination with the control barrier function.

Images