Heparin production process optimization method and system based on environmental analysis
By analyzing the multi-source environmental monitoring data and multi-spectral data in the heparin production process, establishing a nonlinear mapping model, optimizing environmental conditions and process parameters, the problems of low production efficiency and unstable product quality in traditional heparin production processes are solved, and efficient and accurate heparin production is achieved.
Patent Information
- Application Number
- CN202510407457.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-02
- Publication Date
- 2025-06-27
AI Technical Summary
Traditional heparin production processes lack a systematic analysis of the relationship between environmental conditions, process parameters and production purity, resulting in low production efficiency and unstable product quality.
By obtaining the historical multi-source environmental monitoring data and multi-spectral data of the enzymatic lysis reactor, the substance generation path under different environmental conditions was determined, and a nonlinear mapping model of environmental condition parameters and heparin production purity was established based on the random forest algorithm, the influence weight of each environmental condition parameter on the purity of heparin production was calculated, and an environmentally adaptive enzymatic lysis reaction environmental parameter optimization scheme was generated.
It realizes precise control and efficient optimization of the heparin production process, improves production purity and product quality, and meets the needs of industrial production.
Smart Images

Figure CN120220846A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of biopharmaceutical production processes, and particularly to an optimization method and system for heparin production processes based on environmental analysis. Background Art
[0002] Heparin is an anticoagulant drug widely used in clinical practice. Its production process is complex, involving multiple enzymatic hydrolysis reactions, and is significantly affected by environmental conditions (such as temperature, pH value, dissolved oxygen concentration, etc.) and process parameters (such as stirring rate, vibration frequency, etc.). Traditional heparin production processes mainly rely on empirical parameters, lacking a systematic analysis of the relationship between environmental conditions, process parameters, and production purity, resulting in low production efficiency and unstable product quality. In addition, the molecular weight distribution, molecular conformation changes, and vibration characteristics of the reaction kettle during the enzymatic hydrolysis reaction have an important impact on heparin purity, but existing technologies are difficult to monitor these parameters in real time and dynamically adjust them, further limiting production optimization.
[0003] In recent years, the development of multi-source environmental monitoring technology, multi-spectral analysis technology, and vibration spectrum analysis technology has provided new ideas for the optimization of heparin production processes. By collecting environmental data, spectral data, and vibration data of the enzymatic hydrolysis reaction kettle, the material generation path, reaction kinetic characteristics, and the impact of mechanical vibration on the reaction can be analyzed more precisely. However, existing methods still have deficiencies in data processing and parameter optimization, such as the failure to effectively integrate multi-modal data, the lack of the ability to model non-linear relationships, and the inability to achieve dynamic matching of environmental parameters, process parameters, and vibration characteristics.
[0004] Therefore, there is an urgent need for an optimization method and system for heparin production processes based on environmental analysis. Through multi-source data fusion, non-linear modeling, and dynamic parameter adjustment, considering environmental conditions, process parameters, and vibration characteristics comprehensively, precise control and efficient optimization of the heparin production process can be achieved, thereby improving production purity and product quality to meet the requirements of industrial production. The present invention aims to solve the above problems and provide a scientific, systematic, and implementable heparin production optimization scheme. Summary of the Invention
[0005] To solve at least one of the above technical problems, the present invention proposes an optimization method and system for heparin production processes based on environmental analysis.
[0006] In the first aspect of the present invention, an optimization method for heparin production processes based on environmental analysis is provided, including: Obtaining historical multi-source environmental monitoring data and historical multi-spectral data of the enzymatic hydrolysis reaction kettle during the heparin production process, and determining the material generation path of different environmental conditions during the heparin enzymatic hydrolysis process according to the multi-source environmental monitoring data and the historical multi-spectral data; Determine the production purity of heparin under different environmental conditions according to the substance generation path, establish a non-linear mapping model between environmental condition parameters and heparin production purity based on the random forest algorithm, and calculate the influence weights of each environmental condition parameter on heparin production purity; Generate an optimized scheme for the enzymatic reaction environment parameters adapted to the environment according to the influence weights; Obtain the molecular weight distribution data of the crude heparin after implementing the optimized scheme for the enzymatic reaction environment parameters. When it is detected that the molecular weight concentration is less than the preset value, collect the vibration spectrum data and molecular conformation change characteristic data of the enzymatic reaction kettle, and adjust the enzymatic reaction parameters according to the vibration spectrum data and molecular conformation change characteristic data to obtain the reaction kettle adjustment parameters; Perform multi-objective optimization matching on the optimized scheme for the enzymatic reaction environment parameters and the reaction kettle adjustment parameters to generate an optimized heparin production scheme including the mapping relationship between environmental parameters and process parameters.
[0007] In this scheme, the historical multi-source environmental monitoring data and historical multi-spectral data of the enzymatic reaction kettle in the heparin production process are obtained, and the substance generation paths under different environmental conditions in the heparin enzymatic hydrolysis process are determined according to the multi-source environmental monitoring data and historical multi-spectral data. Specifically: Obtain the historical multi-source environmental monitoring data of the historical production batches of heparin from the enzymatic reaction kettle, including the time-series data of temperature, pH value, dissolved oxygen concentration, ion concentration, and enzyme substrate feeding ratio, and simultaneously collect the historical multi-spectral data of the substance changes in the enzymatic reaction kettle in the corresponding historical production batches, including the continuous scanning data of infrared spectrum data and Raman spectrum; Align the historical multi-source environmental monitoring data and historical multi-spectral data in time series based on the dynamic time warping algorithm to construct a time-synchronized enzymatic reaction multi-modal database; Extract features from the historical multi-spectral data based on the successive projections algorithm, and output the characteristic wavelengths related to heparin precursors, intermediate products, impurities, and target heparin; Establish a partial least squares regression model, and perform non-linear mapping on the absorbance at the characteristic wavelengths and the measured heparin titer, sulfation degree, and impurity concentration in the historical heparin production batches according to the partial least squares regression model to generate a spectrum-substance concentration relationship model; Input the time-series data of the historical multi-source environmental monitoring data in the time-synchronized enzymatic reaction multi-modal database into the dynamic Bayesian network, and calculate the substance generation path branch probabilities under different environmental condition parameter combinations in combination with the spectrum-substance concentration relationship network; Generate an environmental condition-substance generation path mapping table under different environmental conditions according to the substance generation path branch probabilities, and obtain the substance generation paths under different environmental conditions in the heparin enzymatic hydrolysis process.
[0008] In this solution, to determine the production purity of heparin under different environmental conditions according to the substance generation path, a non-linear mapping model between environmental condition parameters and heparin production purity is established based on the random forest algorithm, and the influence weights of each environmental condition parameter on heparin production purity are calculated. Specifically: Extract the heparin conversion efficiency and impurity content information under different environmental conditions according to the substance generation path, and determine the production purity of heparin under different environmental conditions based on the heparin conversion efficiency and impurity content information; Construct a non-linear mapping model based on the random forest algorithm, and set the parameter data of the non-linear mapping model, including the number of decision trees, the maximum depth of the decision tree, and the number of samples in the minimum leaf node; Import the production purity and the environmental condition parameters corresponding to each production purity into the non-linear mapping model for training, input the environmental condition parameters and output the predicted value of heparin production purity, and stop training until the preset training iteration times are reached; Randomly generate a preset number of environmental condition parameters and import them into the non-linear mapping model for heparin purity prediction. Calculate the contribution degree of each environmental condition to heparin production purity during the prediction process according to the SHAP value, and determine the influence weight of each environmental condition on heparin production purity according to the contribution degree.
[0009] In this solution, to generate an optimized scheme for the enzymatic reaction environment parameters adaptable to the environment according to the influence weight, specifically: Determine the optimal environmental condition parameter range for the enzymatic reaction during heparin production according to the substance generation path under different environmental conditions; Obtain the multi-source environmental monitoring data of the enzymatic reaction during heparin production in real time, compare the multi-source environmental monitoring data with the optimal environmental condition parameter range, and calibrate the environmental conditions not within the optimal environmental condition parameter range as the environmental conditions to be regulated; Determine the regulation priority of each environmental condition to be regulated according to the influence weight, and determine the compensation value of each environmental condition to be regulated according to the deviation degree between the multi-source environmental monitoring data and the optimal environmental condition parameter range; Construct an optimized scheme for the enzymatic reaction environment parameters adaptable to the environment according to the regulation priority and compensation value of each environmental condition to be regulated.
[0010] In this solution, to obtain the molecular weight distribution data of the crude heparin after implementing the optimized scheme for the enzymatic reaction environment parameters, when it is detected that the molecular weight concentration is less than the preset value, collect the vibration spectrum data and molecular conformation change characteristic data of the enzymatic reaction kettle, and adjust the enzymatic reaction parameters according to the vibration spectrum data and molecular conformation change characteristic data to obtain the reaction kettle adjustment parameters. Specifically: Obtain the mass spectrometry data of the crude heparin after implementing the optimized enzyme hydrolysis reaction environment parameter scheme, and determine the molecular weight distribution data of the crude heparin according to the mass spectrometry data; Determine the molecular weight concentration degree of heparin according to the molecular weight distribution data. When the molecular weight concentration degree is less than the preset value, obtain the vibration spectrum data of the enzyme hydrolysis reaction kettle, and simultaneously obtain the real-time Raman spectrum data of the heparin molecular chain during the enzyme hydrolysis reaction process; Based on a convolutional neural network, extract the molecular conformation change characteristics from the real-time Raman spectrum data to generate dynamic evolution sequences of the molecular chain folding degree and the sulfonic acid group orientation degree; Decompose the vibration spectrum data into fundamental frequency components, harmonic components, and random vibration components, and calculate the energy proportion and spectrum entropy value of each component; When the amplitude of the fundamental frequency component in the vibration spectrum data exceeds the first preset critical value, calibrate the real-time enzyme hydrolysis reaction process as a mechanical resonance interference process, obtain the mechanical stirring rate information of the enzyme hydrolysis reaction kettle, obtain the change relationship between the stirring rate and the fundamental frequency component, determine the vibration frequency range causing resonance interference according to the change relationship, and adjust the mechanical stirring rate according to the vibration frequency range to obtain the first reactor adjustment parameter; If the coefficient of variation of the sulfonic acid group orientation degree in the molecular conformation change characteristics is greater than the second preset critical value, it is determined that the molecular chain is abnormally folded. According to the offset of the stretching vibration peak of the C-O-S bond in the Raman spectrum data, adjust the direction of the temperature gradient change in the reaction kettle to obtain the second reactor adjustment parameter; Adjust the enzyme hydrolysis reaction parameters according to the first reactor adjustment parameter and the second reactor adjustment parameter to obtain the reactor adjustment parameter.
[0011] In this scheme, the optimized enzyme hydrolysis reaction environment parameter scheme and the reactor adjustment parameter are subjected to multi-objective optimization matching to generate an optimized heparin production scheme including the mapping relationship between environmental parameters and process parameters, specifically: Maximize the heparin production purity as the optimization goal, use the environmental condition parameters in the optimized enzyme hydrolysis reaction environment parameter scheme as decision variables, use the dynamic adjustment parameters as constraints, and determine the feasible region of environmental parameters and the feasible region of dynamic adjustment parameters according to the decision variables and constraints; Based on the NSGA algorithm, perform parameter search and matching on the feasible region of environmental parameters and the feasible region of dynamic adjustment parameters, and output the optimal parameter combination that meets the optimization goal; Extract the environmental parameters and reactor parameters in the optimal parameter combination to construct a mapping relationship, determine the reactor parameters corresponding to maximizing the heparin production purity under the set environmental parameters, and construct a mapping relationship between environmental parameters and process parameters; Optimize the enzymatic hydrolysis reaction in the heparin production process according to the environmental parameter-process parameter mapping relationship to obtain an optimized heparin production plan.
[0012] In a second aspect of the present invention, there is also provided an optimization system for the heparin production process based on environmental analysis. The system includes: a memory and a processor. The memory includes an optimization method program for the heparin production process based on environmental analysis. When the optimization method program for the heparin production process based on environmental analysis is executed by the processor, the following steps are implemented: Obtain the historical multi-source environmental monitoring data and historical multi-spectral data of the enzymatic hydrolysis reactor in the heparin production process, and determine the material generation paths under different environmental conditions during the heparin enzymatic hydrolysis process according to the multi-source environmental monitoring data and historical multi-spectral data; Determine the production purity of heparin under different environmental conditions according to the material generation paths, establish a non-linear mapping model between environmental condition parameters and heparin production purity based on the random forest algorithm, and calculate the influence weights of each environmental condition parameter on heparin production purity; Generate an optimized plan for the enzymatic hydrolysis reaction environment adapted to the environment according to the influence weights; Obtain the molecular weight distribution data of the crude heparin after implementing the optimized plan for the enzymatic hydrolysis reaction environment parameters. When it is detected that the molecular weight concentration is less than the preset value, collect the vibration spectrum data and molecular conformation change characteristic data of the enzymatic hydrolysis reactor, and adjust the enzymatic hydrolysis reaction parameters according to the vibration spectrum data and molecular conformation change characteristic data to obtain the reactor adjustment parameters; Perform multi-objective optimization matching on the optimized plan for the enzymatic hydrolysis reaction environment parameters and the reactor adjustment parameters to generate an optimized heparin production plan including the environmental parameter-process parameter mapping relationship.
[0013] The present invention discloses an optimization method and system for the heparin production process based on environmental analysis, aiming to improve the purity and quality of heparin production. The method includes: obtaining the historical environmental monitoring data and multi-spectral data of the enzymatic hydrolysis reactor, and determining the material generation paths under different environmental conditions; establishing a non-linear mapping model between environmental parameters and production purity based on the random forest algorithm, and calculating the influence weights of each parameter; generating an optimized enzymatic hydrolysis reaction plan adapted to the environment according to the weights; dynamically adjusting the enzymatic hydrolysis reaction parameters by detecting the molecular weight distribution data of the crude heparin and combining the vibration spectrum and molecular conformation change characteristics; finally, performing multi-objective matching on the environmental optimization plan and the reaction parameters to generate an optimized production plan including the environmental-process parameter mapping relationship. The present invention can effectively improve the heparin production efficiency and quality and is applicable to industrial production optimization. Description of the Drawings
[0014] Figure 1 Shows a flowchart of an optimization method for the heparin production process based on environmental analysis according to the present invention; Figure 2 The flowchart of the present invention for generating an optimized scheme for the environmental parameters of the enzymatic hydrolysis reaction is shown; Figure 3 The flowchart of the present invention for generating an optimized heparin production scheme is shown; Figure 4 The block diagram of an optimized heparin production process system based on environmental analysis according to the present invention is shown. Detailed implementation manners
[0015] In order to more clearly understand the above objects, features and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific implementation manners. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.
[0016] Many specific details are set forth in the following description in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.
[0017] Figure 1 The flowchart of an optimized heparin production process method based on environmental analysis according to the present invention is shown.
[0018] As Figure 1 shown, the first aspect of the present invention provides an optimized heparin production process method based on environmental analysis, including: S102, obtaining historical multi-source environmental monitoring data and historical multi-spectral data of the enzymatic hydrolysis reactor during the heparin production process, and determining the material generation paths under different environmental conditions during the heparin enzymatic hydrolysis process according to the multi-source environmental monitoring data and the historical multi-spectral data; S104, determining the production purity of heparin under different environmental conditions according to the material generation paths, establishing a non-linear mapping model between environmental condition parameters and heparin production purity based on the random forest algorithm, and calculating the influence weights of each environmental condition parameter on the heparin production purity; S106, generating an optimized scheme for the enzymatic hydrolysis reaction environmental parameters adapted to the environment according to the influence weights; S108, obtaining the molecular weight distribution data of the crude heparin after implementing the optimized scheme for the enzymatic hydrolysis reaction environmental parameters. When it is detected that the molecular weight concentration is less than a preset value, collecting the vibration spectrum data and molecular conformation change characteristic data of the enzymatic hydrolysis reactor, and adjusting the enzymatic hydrolysis reaction parameters according to the vibration spectrum data and the molecular conformation change characteristic data to obtain the reactor adjustment parameters; S110, performing multi-objective optimization matching on the optimized scheme for the enzymatic hydrolysis reaction environmental parameters and the reactor adjustment parameters, and generating an optimized heparin production scheme including the mapping relationship between environmental parameters and process parameters.
[0019] It should be noted that by integrating environmental monitoring data and multi-spectral features, the dynamic correlation between the impurity generation path and the target product in the enzymatic hydrolysis reaction can be accurately identified, significantly reducing the impurity ratio and enhancing the selectivity of the target heparin. Based on the random forest model, the non-linear impact of environmental parameters on purity is quantified, the priority of key control parameters is clarified, and the directional optimization of parameters such as temperature and pH value during the production process is guided, so that the crude product purity can be stably increased to a high-standard range. Aiming at the problem of uneven molecular weight distribution, by combining vibration spectrum analysis and molecular conformation characteristics, potential defects such as mechanical interference or abnormal molecular chain folding can be quickly located, and the stirring rate and temperature gradient are dynamically adjusted to ensure that the molecular weight concentration of the product meets the process requirements. Finally, through multi-objective optimization to match environmental parameters and process parameters, a balance is achieved among maximum purity, minimum energy consumption, and production stability, forming a reusable parameter mapping relationship, significantly reducing the quality fluctuation between batches, while reducing the risk of impurity residue, and overall improving the controllability and product consistency of heparin production.
[0020] According to an embodiment of the present invention, the historical multi-source environmental monitoring data and historical multi-spectral data of the enzymatic hydrolysis reactor in the heparin production process are obtained, and the substance generation paths under different environmental conditions in the heparin enzymatic hydrolysis process are determined according to the multi-source environmental monitoring data and historical multi-spectral data, specifically as follows: Obtain the historical multi-source environmental monitoring data of the historical production batches of heparin from the enzymatic hydrolysis reactor, including the time-series data of temperature, pH value, dissolved oxygen concentration, ion concentration, and enzyme substrate feeding ratio, and synchronously collect the historical multi-spectral data of the substance changes in the enzymatic hydrolysis reactor in the corresponding historical production batches, including the continuous scanning data of infrared spectrum data and Raman spectrum; Based on the dynamic time warping algorithm, the historical multi-source environmental monitoring data and historical multi-spectral data are aligned in time series to construct a time-synchronized enzymatic hydrolysis reaction multi-modal database; Based on the successive projections algorithm, feature extraction is performed on the historical multi-spectral data, and the characteristic wavelengths related to heparin precursors, intermediate products, impurities, and target heparin are output; A partial least squares regression model is established, and according to the partial least squares regression model, the absorbance at the characteristic wavelength is non-linearly mapped to the measured heparin titer, sulfation degree, and impurity concentration in the historical production batches of heparin to generate a spectrum-substance concentration relationship model; The time-series data of the historical multi-source environmental monitoring data in the time-synchronized enzymatic hydrolysis reaction multi-modal database are input into the dynamic Bayesian network, and the substance generation path branch probabilities of the enzymatic hydrolysis reaction under different environmental condition parameter combinations are calculated in combination with the spectrum-substance concentration relationship network; Generate an environmental condition - substance generation path mapping table under different environmental conditions according to the substance generation path branch probability, and obtain the substance generation paths under different environmental conditions during the heparin enzymatic hydrolysis process.
[0021] It should be noted that there is a dynamic causal relationship between the temporal changes of environmental parameters (such as temperature, pH value) and the substance generation path in the enzymatic hydrolysis reaction, and it is difficult for traditional static models to capture the multi - parameter coupling effect and time - lag effect. The dynamic Bayesian network combines historical environmental monitoring data (temporal parameters) with the spectral - substance concentration relationship model to construct the probability dependence relationship between environmental parameters and substance concentration nodes, and uses the conditional probability table to quantify the transfer probability of each substance path (such as heparin synthesis, impurity accumulation) under specific environmental conditions. For example, when the dissolved oxygen concentration continues to be low, the network can infer based on historical data that the probability of the intermediate product converting to impurities increases significantly. By dynamically analyzing the cumulative impact of environmental parameter fluctuations on the reaction path, accurately identify the key parameter combinations leading to impurity generation (such as the synergistic effect of high temperature and low pH); quantify the guiding effect of different environmental conditions on the selection of target products through path branch probability, providing a direct basis for process optimization. For example, adjusting a specific parameter range can increase the heparin synthesis path probability to more than 90%; the generated mapping table can guide the real - time avoidance of high - risk environmental condition combinations during the production process, reduce invalid reaction paths, thereby reducing the impurity ratio to below the process - allowed threshold, while improving the purity of the target product and the consistency between batches, significantly shortening the process debugging cycle and reducing raw material waste. The heparin precursor refers to the initial substance of heparin biosynthesis, such as heparan sulfate. The intermediate product is a transitional substance generated during the enzymatic hydrolysis process, such as partially de - acetylated or de - sulfated products. Impurities include incompletely decomposed proteins, other glycosaminoglycans or side - reaction products. The target heparin is the final purified product required. The construction of the dynamic Bayesian network includes network node definition: environmental parameter nodes (temperature, pH, dissolved oxygen), substance concentration nodes (precursor, intermediate product A / B, target heparin, impurity); conditional probability table (CPT) training: based on historical data, calculate the dependence relationship of each node's state at time t on the state of its parent node at time t - 1 through the maximum likelihood estimation method; path probability calculation: under the given environmental parameter time - series conditions, deduce the hidden state sequence probabilities of the main path and the sub - path through the Forward - Backward Algorithm.
[0022] According to the embodiments of the present invention, to determine the production purity of heparin under different environmental conditions according to the substance generation path, a non - linear mapping model of environmental condition parameters and heparin production purity is established based on the random forest algorithm, and the influence weight of each environmental condition parameter on heparin production purity is calculated. Specifically: Extract the heparin conversion efficiency and impurity content information under different environmental conditions according to the described substance generation path, and determine the production purity of heparin under different environmental conditions according to the heparin conversion efficiency and impurity content information; Construct a non-linear mapping model based on the random forest algorithm, and set the parameter data of the non-linear mapping model, including the number of decision trees, the maximum depth of the decision tree, and the number of samples in the minimum leaf node; Import the production purity and the environmental condition parameters corresponding to each production purity into the non-linear mapping model for training, input the environmental condition parameters and output the predicted value of heparin production purity until the preset training iteration times are reached, and stop training; Randomly generate a preset number of environmental condition parameters and import them into the non-linear mapping model for heparin purity prediction. Calculate the contribution degree of each environmental condition to the heparin production purity during the prediction process according to the SHAP value, and determine the influence weight of each environmental condition on the heparin production purity according to the contribution degree.
[0023] It should be noted that the heparin conversion efficiency and impurity content information extracted based on the substance generation path can directly reflect the change law of production purity under different environmental conditions and provide high-quality data for model training. The random forest algorithm captures the complex non-linear relationship between environmental parameters (such as temperature, pH value, dissolved oxygen concentration, etc.) and purity by constructing multiple decision trees, avoiding the limitation of traditional linear models that cannot handle the interaction of multiple parameters. During the model training process, by setting parameters such as the number of decision trees and the maximum depth, it is ensured that the model can fully learn the data characteristics and avoid overfitting, thereby improving the prediction accuracy. The core of the SHAP value (Shapley Additive Explanations) lies in its fair distribution principle based on game theory, which can quantify the contribution degree of each environmental parameter to the purity prediction result. Specifically, the SHAP value calculates the marginal contribution of a certain parameter in all possible parameter combinations to clarify its influence weight on purity. For example, a higher SHAP value of the pH value indicates that its influence on purity is significant, while a lower SHAP value of the dissolved oxygen concentration indicates that its influence is relatively small. This quantification method can not only identify key regulatory parameters (such as the contribution ratio of temperature fluctuation to purity reaches 30%), but also reveal the interaction between parameters (such as high temperature and low pH synergistically lead to impurity generation); through the high-precision prediction of the random forest model, the production purity under different environmental parameter combinations can be evaluated in advance, reducing the trial-and-error cost; based on the contribution degree analysis of the SHAP value, clarify the regulatory priority of each parameter, and guide the key optimization of key parameters during the production process (such as stabilizing the pH value in the optimal range); by quantifying the parameter interaction, avoid the local optimum problem caused by single-parameter optimization, achieve global optimal regulation, and improve the stability of production purity. The higher the contribution degree, the higher the influence weight.
[0024] Figure 2 The flowchart of the optimization scheme for the enzymatic reaction environment parameters generated by the present invention is shown.
[0025] According to an embodiment of the present invention, the optimization scheme for the enzymatic reaction environment parameters adapted to the environment generated according to the influence weight is specifically as follows: S202. Determine the optimal environmental condition parameter range for the enzymatic reaction in the heparin production process according to the substance generation paths under different environmental conditions; S204. Obtain the multi-source environmental monitoring data of the enzymatic reaction in the heparin production process in real time, compare the multi-source environmental monitoring data with the optimal environmental condition parameter range, and label the environmental conditions not within the optimal environmental condition parameter range as the environmental conditions to be regulated; S206. Determine the regulation priority of each environmental condition to be regulated according to the influence weight, and determine the compensation value of each environmental condition to be regulated according to the deviation degree between the multi-source environmental monitoring data and the optimal environmental condition parameter range; S208. Construct an optimization scheme for the enzymatic reaction environment parameters adapted to the environment according to the regulation priority and compensation value of each environmental condition to be regulated.
[0026] It should be noted that determining the regulation priority according to the influence weight can focus on the key parameters that have the greatest impact on the production purity, and adjust these parameters first to quickly restore the production stability. At the same time, calculating the compensation value based on the deviation degree ensures that the regulation measures are accurate and effective, and avoids waste of resources or production fluctuations caused by excessive adjustment. Finally, by integrating the regulation priority and the compensation value, the constructed optimization scheme adapted to the environment can dynamically adapt to the changes in the environmental conditions during the production process, ensuring that the enzymatic reaction is always carried out within the optimal environmental condition parameter range. Controlling the environmental parameter fluctuations within the optimal range reduces the quality differences between batches; by preferentially regulating the key parameters, quickly correcting the production deviation, and shortening the abnormal handling time.
[0027] According to an embodiment of the present invention, for the molecular weight distribution data of the crude heparin after implementing the optimization scheme for the enzymatic reaction environment parameters, when it is detected that the molecular weight concentration is less than the preset value, collect the vibration spectrum data and the molecular conformation change characteristic data of the enzymatic reaction kettle, and adjust the enzymatic reaction parameters according to the vibration spectrum data and the molecular conformation change characteristic data to obtain the reaction kettle adjustment parameters, specifically as follows: Obtain the mass spectrometry data of the crude heparin after implementing the optimization scheme for the enzymatic reaction environment parameters, and determine the molecular weight distribution data of the crude heparin according to the mass spectrometry data; Determine the molecular weight concentration of heparin based on the molecular weight distribution data. When the molecular weight concentration is less than the preset value, obtain the vibration spectrum data of the enzymatic hydrolysis reactor, and simultaneously obtain the real-time Raman spectrum data of the heparin molecular chain during the enzymatic hydrolysis process; Based on a convolutional neural network, extract the molecular conformation change characteristics from the real-time Raman spectrum data to generate the dynamic evolution sequences of the molecular chain folding degree and the sulfonic acid group orientation degree; Decompose the vibration spectrum data into fundamental frequency components, harmonic components and random vibration components, and calculate the energy proportion and spectrum entropy value of each component; It should be noted that during the heparin production process, the molecular weight distribution is one of the key indicators to measure the product quality. The target heparin usually requires the molecular weight to be concentrated within a certain range (such as 12 - 15 kDa) to ensure the consistency of its anticoagulant activity and drug efficacy. The molecular weight concentration (i.e., the dispersion degree of the molecular weight distribution) reflects the uniformity and controllability of the enzymatic hydrolysis reaction during the production process. When the molecular weight concentration is less than the preset value (such as the molecular weight dispersion > 15%), it indicates that there are too many low molecular weight fragments (such as < 10 kDa) or high molecular weight impurities (such as > 20 kDa) in the product, which will not only reduce the biological activity of heparin, but may also lead to pharmacological safety problems. Uneven distribution of enzyme activity or fluctuations in substrate concentration during the enzymatic hydrolysis process may cause some polysaccharide chains to be over-degraded. Too high stirring rate or abnormal mechanical vibration of the reactor may damage the integrity of the heparin molecular chain, resulting in a wider molecular weight distribution. Or fluctuations in temperature or pH value during the reaction process may cause abnormal folding of the heparin molecular chain or disorder of the sulfonic acid group orientation, affecting the enzymatic hydrolysis efficiency and product uniformity. These are important reasons affecting the molecular weight concentration. Therefore, by obtaining the molecular weight distribution data after the optimization plan, the product quality can be monitored in real time, and the problem of insufficient molecular weight concentration can be detected in time. When the detected concentration is lower than the preset value, further collect the vibration spectrum data and Raman spectrum data, which can accurately locate the root cause of the problem (such as mechanical resonance interference or abnormal molecular conformation), and provide a scientific basis for dynamically adjusting the reaction parameters (such as stirring rate, temperature gradient), so as to control the molecular weight distribution within the target range and ensure the product activity and safety. The molecular weight distribution data of the heparin crude product obtained in this step are the molecular weight distribution data of multiple heparin crude products in a preset number of batches. The mass spectrometry data includes the mass-to-charge ratio, peak intensity, isotope pattern, and fragment ion information of the heparin crude product.
[0028] If the amplitude of the fundamental frequency component in the vibration spectrum data exceeds the first preset critical value, calibrate the real-time enzymatic hydrolysis reaction process as a mechanical resonance interference process, obtain the mechanical stirring rate information of the enzymatic hydrolysis reactor, obtain the change relationship between the stirring rate and the fundamental frequency component, determine the vibration frequency range causing the resonance interference according to the change relationship, and adjust the mechanical stirring rate according to the vibration frequency range to obtain the first reactor adjustment parameter; It should be noted that the mechanical stirring of the reactor is a key link to ensure the uniform mixing of reactants and the enzymatic hydrolysis efficiency. However, improper setting of the stirring rate leads to mechanical resonance, that is, the stirring frequency matches the natural frequency of the reactor, resulting in abnormal vibration. Such vibration will not only damage the integrity of the heparin molecular chain, leading to a wider molecular weight distribution (such as the generation of too many low molecular weight fragments); by analyzing the vibration spectrum data, when the amplitude of the fundamental frequency component exceeds the preset critical value, the presence of mechanical resonance can be quickly determined; by obtaining the variation relationship between the stirring rate and the fundamental frequency component, the specific frequency range causing resonance can be clarified (such as when the stirring rate is between 200 - 250 rpm, the fundamental frequency amplitude increases significantly), providing a direct basis for adjusting the stirring parameters; according to the resonance frequency range, the stirring rate is adjusted to a safe range (such as avoiding the 200 - 250 rpm range), eliminating the interference of mechanical vibration on the reaction process, ensuring the integrity of the heparin molecular chain, and restoring the molecular weight concentration to above the preset value.
[0029] If the coefficient of variation of the sulfonic acid group orientation degree in the molecular conformation change characteristics is greater than the second preset critical value, it is determined that the molecular chain is abnormally folded, and according to the shift amount of the stretching vibration peak of the C - O - S bond in the Raman spectrum data, the direction of the temperature gradient change in the reactor is adjusted to obtain the second reactor adjustment parameter; The enzymatic hydrolysis reaction parameters are adjusted according to the first reactor adjustment parameter and the second reactor adjustment parameter to obtain the reactor adjustment parameter.
[0030] It should be noted that the stability of the molecular conformation directly affects the biological activity and quality of the product. The sulfonic acid group is a key active group in the heparin molecule, and its degree of orientation (i.e., the spatial arrangement of the sulfonic acid group on the molecular chain) determines the anticoagulant activity of heparin. When the coefficient of variation of the sulfonic acid group orientation degree is greater than the preset critical value, it indicates that the molecular chain has abnormal folding, resulting in disordered spatial distribution of the sulfonic acid group, thereby reducing the biological activity of heparin. This abnormal folding is usually caused by uneven temperature gradient or pH value fluctuation during the reaction process. For example, too high local temperature may cause the molecular chain to overstretch, while too low temperature may cause local aggregation of the molecular chain. Through the offset of the stretching vibration peak of the C-O-S bond in the Raman spectrum data, the change of the sulfonic acid group orientation degree can be monitored in real time. When the coefficient of variation exceeds the preset critical value, the abnormal folding of the molecular chain can be quickly determined, avoiding the lag of relying on endpoint detection in traditional methods; according to the direction and amplitude of the offset of the C-O-S bond vibration peak, the unevenness of the temperature gradient in the reaction kettle can be clarified (such as too high or too low local temperature); by changing the direction of the temperature gradient change in the reaction kettle (such as adjusting from unidirectional heating to bidirectional circulation heating), the local temperature difference can be eliminated, ensuring uniform folding of the molecular chain and restoring the normal orientation degree of the sulfonic acid group. The second reaction kettle adjusts the parameters to form a negative feedback control between the temperature gradient direction and the molecular chain stretching direction. When it is detected that the heparin molecular chain is overstretched in a certain direction, the system will set a reverse temperature gradient in this direction (such as cooling at the front end and heating at the rear end), and use the thermophoretic effect to shrink the molecular chain; on the contrary, if the molecular chain is curled, a positive temperature (such as heating and cooling at the front and rear ends simultaneously) gradient will be generated to stretch it. This dynamic adjustment ensures that the molecular chain is always in the optimal conformation, thereby improving the enzymatic hydrolysis efficiency and product quality.
[0031] Figure 3 The flowchart showing the optimization scheme for heparin production according to the present invention is presented.
[0032] According to an embodiment of the present invention, the multi-objective optimization matching of the enzymatic hydrolysis reaction environment parameter optimization scheme and the reaction kettle adjustment parameters to generate a heparin production optimization scheme including the mapping relationship between environmental parameters - process parameters is specifically as follows: S302, taking the maximization of heparin production purity as the optimization goal, using the environmental condition parameters in the enzymatic hydrolysis reaction environment parameter optimization scheme as decision variables, and using the dynamic adjustment parameters as constraint conditions, to determine the feasible region of environmental parameters and the feasible region of dynamic adjustment parameters according to the decision variables and constraint conditions; S304, based on the NSGA algorithm, performing parameter search and matching on the feasible region of environmental parameters and the feasible region of dynamic adjustment parameters, and outputting the optimal parameter combination that meets the optimization goal; S306. Extract the environmental parameters and reactor parameters in the optimal parameter combination to construct a mapping relationship, determine the reactor parameters corresponding to the maximized heparin production purity under the set environmental parameters, and construct an environmental parameter - process parameter mapping relationship. S308. Optimize the enzymatic hydrolysis reaction in the heparin production process according to the environmental parameter - process parameter mapping relationship to obtain an optimized heparin production plan.
[0033] It should be noted that the environmental parameters of the enzymatic hydrolysis reaction (such as temperature, pH value, dissolved oxygen concentration) and the reactor process parameters (such as stirring rate, temperature gradient) jointly determine the purity, activity, and molecular weight distribution of the product. However, there are complex interactions between environmental parameters and process parameters. The optimization of a single parameter may lead to suboptimal choices of other parameters and even cause new problems. For example, increasing the temperature may accelerate the enzymatic hydrolysis reaction, but at the same time, it may also increase the generation of impurities or cause mechanical resonance; adjusting the stirring rate can improve the reaction uniformity, but it may affect the temperature gradient distribution. Therefore, it is difficult to achieve the global optimum by independently optimizing only the environmental parameters of the enzymatic hydrolysis reaction or the reactor parameters. By using the NSGA algorithm to search for the optimal parameter combination, it is ensured that the heparin production purity reaches the highest level (such as > 95%), and at the same time, the impurity content is controlled below the pharmacopoeia standard (such as < 0.5%), significantly improving the product quality and the consistency of drug efficacy; comprehensively considering the interaction between environmental parameters and process parameters, it is ensured that the molecular weight of the product is concentrated within a certain range (such as 12 - 15 kDa), avoiding the generation of low molecular weight fragments or high molecular weight impurities, and improving the biological activity and batch - to - batch consistency of heparin. NSGA (Non - dominated Sorting Genetic Algorithm) is a multi - objective optimization algorithm based on genetic algorithms. It randomly generates a set of initial parameter combinations (such as temperature, pH value, stirring rate, etc.) as the initial population. Non - dominated sorting: Sort the individuals in the population according to multiple optimization objectives (such as maximizing purity and minimizing energy consumption), and divide the individuals into different levels (Pareto fronts). Crowding degree calculation: Calculate the crowding degree between individuals within the same level to ensure the uniform distribution of solutions and avoid local convergence. Selection, crossover, and mutation: Based on non - dominated sorting and crowding degree, select excellent individuals for crossover and mutation to generate a new generation of population. Iterative optimization: Repeat the above steps until the preset number of iterations or convergence conditions are reached, and finally output the Pareto optimal solution set.
[0034] According to the embodiments of the present invention, it further includes: Collect the temperature gradient time - series data and real - time Raman spectrum data of the enzymatic hydrolysis reactor, extract the Raman peak intensity ratio data of the real - time Raman spectrum data, and use the Raman peak intensity ratio data as the characterization of the sulfonic acid group orientation degree. Generate the time-series trajectory of the orientation degree of sulfonic acid groups under the action of a temperature gradient through molecular dynamics simulation, construct a long short-term memory network, and import the time-series trajectory of the orientation degree of sulfonic acid groups into the long short-term memory network for training; Obtain the time-series trajectory change data of the orientation degree of sulfonic acid groups with the current preset time length and import it into the trained long short-term memory network to predict the change of the orientation degree of sulfonic acid groups within a future preset time period, obtain the prediction result, and judge the change lag of the orientation degree of sulfonic acid groups according to the prediction result; Determine the actual coefficient of variation of the orientation degree of sulfonic acid groups according to the change lag, correct the coefficient of variation of the orientation degree of sulfonic acid groups according to the actual coefficient of variation, and determine the correction time advance of the coefficient of variation according to the change lag, so as to judge whether there is an abnormal folding phenomenon of the molecular chain.
[0035] It should be noted that heparin molecular chains are relatively long and have a complex spatial structure, and there are various interactions between the chemical bonds and groups inside them, such as hydrogen bonds, van der Waals forces, etc. These interactions enable the molecular chain to maintain a relatively stable conformation to a certain extent, restricting the rapid response of sulfonic acid groups to changes in the external environment and changing their orientation. For example, when the temperature gradient changes, the hydrogen bond network inside the molecular chain needs a certain amount of time to readjust, resulting in a lag in the change of the orientation degree of sulfonic acid groups. Generate the time-series trajectory of the orientation degree of sulfonic acid groups under the action of a temperature gradient through molecular dynamics simulation and use a long short-term memory network (LSTM) for training. LSTM has a powerful ability to process sequence data and can learn the complex non-linear relationship between the temperature gradient and the orientation degree of sulfonic acid groups, so as to more accurately predict the change trend of the orientation degree of sulfonic acid groups within a future preset time period. This accurate trend capture ability helps to detect potential abnormalities in advance before the actual orientation degree changes significantly; judge the change lag of the orientation degree of sulfonic acid groups according to the prediction result and determine the actual coefficient of variation accordingly. Traditional calculation methods of the coefficient of variation are often based on the current moment or limited historical data and cannot take into account the possible future change trends, which easily leads to calculation deviations of the coefficient of variation. However, the present invention can correct the coefficient of variation in real time by introducing the prediction result, making it more in line with the actual molecular chain state, thereby improving the prediction accuracy of the abnormal folding phenomenon of the molecular chain. Since it can accurately predict the change trend of the orientation degree of sulfonic acid groups and correct the coefficient of variation, the present invention can discover potential risks in advance before the abnormal folding phenomenon of the molecular chain occurs. For example, when the prediction result shows that the orientation degree of sulfonic acid groups will exceed the normal range in a future period of time, and the correction time advance of the coefficient of variation indicates that immediate intervention measures need to be taken, production personnel can adjust the environmental parameters of the reaction kettle in advance, such as temperature gradient, stirring rate, etc., to avoid the occurrence of abnormal folding of the molecular chain, thereby ensuring the quality and purity of heparin products.
[0036] Figure 4 The block diagram of an optimization system for heparin production process based on environmental analysis according to the present invention is shown.
[0037] In the second aspect of the present invention, an optimization system 4 for heparin production process based on environmental analysis is further provided. The system includes: a memory 41 and a processor 42. The memory includes an optimization method program for heparin production process based on environmental analysis. When the optimization method program for heparin production process based on environmental analysis is executed by the processor, the following steps are implemented: Obtain the historical multi-source environmental monitoring data and historical multi-spectral data of the enzymatic hydrolysis reactor during the heparin production process, and determine the substance generation paths under different environmental conditions during the heparin enzymatic hydrolysis process according to the multi-source environmental monitoring data and the historical multi-spectral data; Determine the production purity of heparin under different environmental conditions according to the substance generation paths, establish a non-linear mapping model between the environmental condition parameters and the heparin production purity based on the random forest algorithm, and calculate the influence weights of each environmental condition parameter on the heparin production purity; Generate an optimized scheme for the enzymatic hydrolysis reaction environment parameters adapted to the environment according to the influence weights; Obtain the molecular weight distribution data of the crude heparin after implementing the optimized scheme for the enzymatic hydrolysis reaction environment parameters. When it is detected that the molecular weight concentration is less than a preset value, collect the vibration spectrum data and molecular conformation change characteristic data of the enzymatic hydrolysis reactor, and adjust the enzymatic hydrolysis reaction parameters according to the vibration spectrum data and the molecular conformation change characteristic data to obtain the adjusted parameters of the reactor; Perform multi-objective optimization matching on the optimized scheme for the enzymatic hydrolysis reaction environment parameters and the adjusted parameters of the reactor to generate an optimized heparin production scheme including the mapping relationship between environmental parameters and process parameters.
[0038] The present invention discloses an optimization method and system for heparin production process based on environmental analysis, aiming to improve the purity and quality of heparin production. The method includes: obtaining the historical environmental monitoring data and multi-spectral data of the enzymatic hydrolysis reactor, and determining the substance generation paths under different environmental conditions; establishing a non-linear mapping model between environmental parameters and production purity based on the random forest algorithm, and calculating the influence weights of each parameter; generating an optimized enzymatic hydrolysis reaction scheme adapted to the environment according to the weights; dynamically adjusting the enzymatic hydrolysis reaction parameters by detecting the molecular weight distribution data of the crude heparin and combining the vibration spectrum and molecular conformation change characteristics; finally, performing multi-objective matching on the environmental optimization scheme and the reaction parameters to generate an optimized production scheme including the mapping relationship between environmental and process parameters. The present invention can effectively improve the production efficiency and quality of heparin and is applicable to industrial production optimization.
[0039] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling, direct coupling, or communication connection between the displayed or discussed components can be through some interfaces. The indirect coupling or communication connection of devices or units can be electrical, mechanical, or other forms.
[0040] The units described above as separate components may or may not be physically separated. The components displayed as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0041] In addition, each functional unit in the embodiments of the present invention can be all integrated in a processing unit, or each unit can be separately used as a unit, or two or more units can be integrated in a unit. The above-mentioned integrated units can be implemented in the form of hardware, or in the form of hardware plus software functional units.
[0042] Those of ordinary skill in the art can understand that all or part of the steps to implement the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps including the above method embodiments. The foregoing storage medium includes: removable storage devices, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks, or optical discs and other various media that can store program codes.
[0043] Alternatively, if the above-mentioned integrated units of the present invention are implemented in the form of software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiments of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: removable storage devices, ROM, RAM, magnetic disks, or optical discs and other various media that can store program codes.
[0044] As described above, it is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims described.
Claims
1. A method for optimizing the heparin production process based on environmental analysis, characterized in that: The following steps are involved: Acquire historical multi-source environmental monitoring data and historical multi-spectral data of the enzymatic hydrolysis reactor in the heparin production process, and determine the substance generation paths under different environmental conditions in the heparin enzymatic hydrolysis process according to the multi-source environmental monitoring data and historical multi-spectral data; Determining the production purity of heparin under different environmental conditions according to the substance generation pathway, establishing a nonlinear mapping model between environmental condition parameters and heparin production purity based on a random forest algorithm, and calculating the influence weight of each environmental condition parameter on the heparin production purity; Generate an environmentally adaptive enzymatic reaction environmental parameter optimization scheme according to the impact weights; Obtaining the molecular weight distribution data of the crude heparin after implementing the enzymatic reaction environment parameter optimization scheme, and when it is detected that the molecular weight concentration is less than a preset value, collecting the vibration spectrum data and molecular conformation change characteristic data of the enzymatic reaction kettle, and adjusting the enzymatic reaction parameters according to the vibration spectrum data and molecular conformation change characteristic data to obtain the reactor adjustment parameters; The enzymatic hydrolysis reaction environmental parameter optimization scheme is matched with the reactor adjustment parameters through multi-objective optimization to generate a heparin production optimization scheme including an environmental parameter-process parameter mapping relationship.
2. The method for optimizing the heparin production process based on environmental analysis according to claim 1, characterized in that: The obtaining of historical multi-source environmental monitoring data and historical multi-spectral data of the enzymatic hydrolysis reactor in the heparin production process, and determining the substance generation paths under different environmental conditions in the heparin enzymatic hydrolysis process according to the multi-source environmental monitoring data and historical multi-spectral data, specifically includes: Obtain historical multi-source environmental monitoring data of historical production batches of heparin from the enzymatic hydrolysis reactor, including time series data of temperature, pH value, dissolved oxygen concentration, ion concentration and enzyme-substrate feed ratio, and simultaneously collect historical multi-spectral data of material changes in the enzymatic hydrolysis reactor in the corresponding historical production batches, including infrared spectral data and continuous scanning data of Raman spectroscopy; Based on a dynamic time warping algorithm, the historical multi-source environmental monitoring data and the historical multi-spectral data are time-series aligned to construct a time-synchronized enzymatic reaction multimodal database; Extracting features from the historical multispectral data based on a continuous projection algorithm, and outputting characteristic wavelengths related to heparin precursors, intermediates, impurities, and target heparin; A partial least squares regression model is established, and the absorbance at the characteristic wavelength is nonlinearly mapped with the heparin potency, sulfation degree and impurity concentration measured in the historical production batches of heparin according to the partial least squares regression model to generate a spectrum-substance concentration relationship model; Input the time series data of historical multi-source environmental monitoring data in the time-synchronized enzymatic reaction multimodal database into the dynamic Bayesian network, and calculate the branch probability of the substance generation path of the enzymatic reaction under different environmental condition parameter combinations in combination with the spectrum-substance concentration relationship network; An environmental condition-substance generation path mapping table under different environmental conditions is generated according to the branch probability of the substance generation path, and the substance generation paths under different environmental conditions in the heparin enzymatic hydrolysis process are obtained.
3. The method for optimizing the heparin production process based on environmental analysis according to claim 1, characterized in that: The production purity of heparin under different environmental conditions is determined according to the substance generation path, a nonlinear mapping model between environmental condition parameters and heparin production purity is established based on the random forest algorithm, and the influence weight of each environmental condition parameter on the heparin production purity is calculated, specifically: Extracting heparin conversion efficiency and impurity content information under different environmental conditions according to the substance generation pathway, and determining the production purity of heparin under different environmental conditions according to the heparin conversion efficiency and impurity content information; Constructing a nonlinear mapping model based on a random forest algorithm, and setting parameter data of the nonlinear mapping model, including the number of decision trees, the maximum depth of the decision trees, and the minimum number of leaf node samples; Importing the production purity and the environmental condition parameters corresponding to each production purity into the nonlinear mapping model for training, inputting the environmental condition parameters and outputting the predicted value of heparin production purity, until a preset number of training iterations is reached, and then stopping the training; A preset number of environmental condition parameters are randomly generated and imported into the nonlinear mapping model to predict the purity of heparin. The contribution of each environmental condition to the purity of heparin production during the prediction process is calculated according to the SHAP value, and the influence weight of each environmental condition on the purity of heparin production is determined according to the contribution.
4. The method for optimizing the heparin production process based on environmental analysis according to claim 1, characterized in that: The environmental parameter optimization scheme for the enzymatic hydrolysis reaction that is environmentally adaptive is generated according to the influence weights, specifically: Determining the optimum environmental condition parameter range for the enzymatic hydrolysis reaction in the heparin production process according to the substance generation pathways under the different environmental conditions; Acquire multi-source environmental monitoring data of the enzymatic hydrolysis reaction in the heparin production process in real time, compare the multi-source environmental monitoring data with the optimal environmental condition parameter range, and calibrate environmental conditions that are not within the optimal environmental condition parameter range as environmental conditions to be regulated; Determine the control priority of each environmental condition to be regulated according to the influence weight, and determine the compensation value of each environmental condition to be regulated according to the degree of deviation between the multi-source environmental monitoring data and the optimal environmental condition parameter interval; An environmentally adaptive enzymatic reaction environmental parameter optimization scheme is constructed according to the control priority and compensation value of each environmental condition to be regulated.
5. The method for optimizing the heparin production process based on environmental analysis according to claim 1, characterized in that: The molecular weight distribution data of crude heparin after implementing the enzymatic reaction environment parameter optimization scheme is obtained. When it is detected that the molecular weight concentration is less than a preset value, the vibration spectrum data and molecular conformation change characteristic data of the enzymatic reaction kettle are collected, and the enzymatic reaction parameters are adjusted according to the vibration spectrum data and molecular conformation change characteristic data to obtain the reactor adjustment parameters, which are specifically: Acquiring mass spectrum data of the crude heparin after implementing the enzymatic reaction environmental parameter optimization scheme, and determining molecular weight distribution data of the crude heparin according to the mass spectrum data; Determine the molecular weight concentration of heparin according to the molecular weight distribution data; if the molecular weight concentration is less than a preset value, obtain the vibration spectrum data of the enzymatic hydrolysis reactor, and simultaneously obtain the real-time Raman spectrum data of the heparin molecular chain during the enzymatic hydrolysis reaction; Extracting molecular conformational change features from the real-time Raman spectroscopy data based on convolutional neural network extraction to generate a dynamic evolution sequence of molecular chain folding degree and sulfonic acid group orientation degree; Decomposing the vibration spectrum data into fundamental frequency component, harmonic component and random vibration component, and calculating the energy proportion and spectrum entropy value of each component; If the amplitude of the fundamental frequency component in the vibration spectrum data exceeds a first preset critical value, the real-time enzymatic hydrolysis reaction process is calibrated as a mechanical resonance interference process, the mechanical stirring rate information of the enzymatic hydrolysis reactor is obtained, the change relationship between the stirring rate and the fundamental frequency component is obtained, the vibration frequency interval causing the resonance interference is determined according to the change relationship, the mechanical stirring rate is adjusted according to the vibration frequency interval, and the first reactor adjustment parameter is obtained; If the coefficient of variation of the orientation degree of the sulfonic acid group in the molecular conformational change characteristics is greater than the second preset critical value, it is determined that the molecular chain is abnormally folded, and the direction of the temperature gradient change in the reactor is adjusted according to the offset of the COS bond stretching vibration peak in the Raman spectrum data to obtain the second reactor adjustment parameter; The enzymatic hydrolysis reaction parameters are adjusted according to the first reactor adjustment parameters and the second reactor adjustment parameters to obtain the reactor adjustment parameters.
6. The method for optimizing the heparin production process based on environmental analysis according to claim 1, characterized in that: The enzymatic hydrolysis reaction environmental parameter optimization scheme is matched with the reactor adjustment parameters through multi-objective optimization to generate a heparin production optimization scheme including an environmental parameter-process parameter mapping relationship, specifically: Taking the maximization of the purity of heparin production as the optimization goal, taking the environmental condition parameters in the optimization scheme of the environmental parameters of the enzymatic hydrolysis reaction as the decision variables, taking the dynamic adjustment parameters as the constraints, and determining the feasible domain of the environmental parameters and the feasible domain of the dynamic adjustment parameters according to the decision variables and the constraints; Based on the NSGA algorithm, a parameter search and matching is performed on the feasible domain of the environmental parameters and the feasible domain of the dynamic adjustment parameters, and an optimal parameter combination that meets the optimization goal is output; Extracting the environmental parameters and reactor parameters in the optimal parameter combination to construct a mapping relationship, determining the reactor parameters corresponding to the maximum purity of heparin production under the set environmental parameters, and constructing an environmental parameter-process parameter mapping relationship; The enzymatic hydrolysis reaction in the heparin production process is optimized according to the environmental parameter-process parameter mapping relationship to obtain a heparin production optimization plan.
7. A heparin production process optimization system based on environmental analysis, characterized in that: The heparin production process optimization system based on environmental analysis includes a storage device and a processor. The storage device includes a heparin production process optimization method program based on environmental analysis. When the heparin production process optimization method program based on environmental analysis is executed by the processor, the following steps are implemented: Acquire historical multi-source environmental monitoring data and historical multi-spectral data of the enzymatic hydrolysis reactor in the heparin production process, and determine the substance generation paths under different environmental conditions in the heparin enzymatic hydrolysis process according to the multi-source environmental monitoring data and historical multi-spectral data; Determining the production purity of heparin under different environmental conditions according to the substance generation pathway, establishing a nonlinear mapping model between environmental condition parameters and heparin production purity based on a random forest algorithm, and calculating the influence weight of each environmental condition parameter on the heparin production purity; Generate an environmentally adaptive enzymatic reaction environmental parameter optimization scheme according to the impact weights; Obtaining the molecular weight distribution data of the crude heparin after implementing the enzymatic reaction environment parameter optimization scheme, and when it is detected that the molecular weight concentration is less than a preset value, collecting the vibration spectrum data and molecular conformation change characteristic data of the enzymatic reaction kettle, and adjusting the enzymatic reaction parameters according to the vibration spectrum data and molecular conformation change characteristic data to obtain the reactor adjustment parameters; The enzymatic hydrolysis reaction environmental parameter optimization scheme is matched with the reactor adjustment parameters through multi-objective optimization to generate a heparin production optimization scheme including an environmental parameter-process parameter mapping relationship.
Citation Information
Cited By
Industrial exception automatic processing system based on operation instruction matching
CN120598217A
Production process parameter dynamic adjustment method and device based on AI visual inspection
CN120612016A
Comprehensive two-dimensional mass spectrum data non-targeted screening method and device, medium and computer equipment
CN121191637A
Parameter cooperative control system based on yeast extract enzymolysis reaction
CN121320086A