A macroporous resin column efficiency monitoring method and system for porcine brain peptide purification

Abstract

The application discloses a macroporous resin column efficiency monitoring method and system for pig brain peptide purification, and belongs to the technical field of biopharmaceutical engineering and natural product extraction and separation. The method comprises the following steps: monitoring the drift of the interception performance of an upstream ultrafiltration membrane to obtain a feed load correction factor; calculating a micro-pore channel blockage coefficient based on the cumulative lipid adsorption amount to represent the internal implicit adsorption recession of the resin; calculating a bed layer compaction factor based on a dynamic flow resistance coefficient to determine the bed layer compaction and flow field abnormality; and fusing the feed load correction factor, the micro-pore channel blockage coefficient and the bed layer compaction factor to construct a comprehensive column efficiency recession index to correct the real-time separation degree. The system adopts a multi-dimensional threshold judgment strategy to distinguish the recession mechanism, and adaptively triggers deep regeneration, backwashing or upstream membrane cleaning according to the judgment result, so that a cross-section cooperative feedback mechanism is formed, the service life of the resin is prolonged, and the product quality is stable.

Description

Technical Field

[0001] This invention belongs to the field of biopharmaceutical engineering and natural product extraction and separation technology, specifically relating to a macroporous resin column efficiency monitoring method and system for the purification of porcine brain peptides. Background Technology

[0002] Pig brain, as an important biological resource, is rich in bioactive peptides with neuroregulatory effects. However, pig brain tissue is dense, high in lipids, and rich in endogenous proteases. If the raw material is not subjected to timely low-temperature treatment after collection, it is prone to rapid protein degradation and the generation of a large number of impurity peptides, leading to the destruction of the target peptide structure and the loss of bioactive components. This results in batch-to-batch quality fluctuations and increases the difficulty of industrial quality control. Current laboratory research often uses high-performance liquid chromatography (HPLC) to obtain high-purity products, but its high equipment and operating costs and low throughput make it difficult to meet the needs of large-scale production. Therefore, industrial production tends to use scalable column chromatography processes, among which macroporous adsorption resins are widely used due to their large adsorption capacity, high mechanical strength, and cost advantages.

[0003] However, scaling up chromatography processes from the laboratory to industrial-grade macroporous resin columns is not a simple volumetric scaling-up; it is often accompanied by significant scale-up effects. To adapt to plant space and operating conditions, industrial columns often require adjustments to column bed geometry, such as reducing the column height-to-diameter ratio, which leads to decreased resolution. Simultaneously, large-diameter columns are more prone to non-uniform flow fields such as wall flow and channeling, and under long-term pressure differentials and fluid scouring, packing compaction occurs, resulting in reduced mass transfer efficiency and decreased column stability. Existing processes mostly employ open-loop control methods with fixed flow rates or gradients, lacking effective characterization and feedback of changes in the column's internal state, thus leading to significant fluctuations in column efficiency and difficulty in ensuring batch-to-batch consistency.

[0004] More importantly, porcine brain extract is a high-fat feed system. Even after defatting and ultrafiltration, trace amounts of lipids and protein colloids may still remain in the feed. During long-term operation, these impurities can gradually adsorb and contaminate the resin pores or cover the resin surface, leading to cumulative and insidious microscopic column efficiency decline. In the early stages of contamination, macroscopic operating parameters such as column pressure drop and flow rate may still be within the set range, making it difficult to detect abnormalities in time. However, the resin specific surface area and effective pore size have already decreased, causing adverse changes in the diffusion and adsorption kinetics of small molecule brain peptides within the pores, which in turn leads to a continuous decline in yield and purity without being easily detected. On the other hand, upstream ultrafiltration membrane fouling may cause drift in equivalent pore size and cutoff boundary, causing the feed molecular weight distribution to deviate from the design window and exacerbating resin column load fluctuations. Furthermore, if the regeneration process relies solely on a fixed cleaning procedure, it is difficult to objectively evaluate the degree of lipid residue removal, resulting in the resin entering the next batch in an insufficiently regenerated state, accelerating performance degradation.

[0005] Therefore, existing technologies still lack a method to quantitatively monitor and determine the column efficiency changes and upstream membrane module status during the operation of the chromatography column, taking into account the characteristics of high fat content and easy residue in pig brain materials, and the characteristics of macroporous adsorption resins being sensitive to flow fields, easy to compact, and easy to contaminate pores, and to achieve closed-loop control. As a result, it is difficult to stably obtain high-yield, high-purity, and batch-to-batch consistent brain peptide products under continuous or long-term industrial operation conditions. Summary of the Invention

[0006] This invention aims to at least solve one of the technical problems existing in the prior art. To this end, this invention proposes a macroporous resin column efficiency monitoring method for the purification of porcine brain peptides, characterized by comprising the following steps:

[0007] S1: The state of the crude porcine brain extract entering the ultrafiltration membrane module is regulated, and the transmembrane pressure difference, permeate flux and lipid content of the feed batch are collected. The membrane fouling resistance coefficient is calculated and the equivalent molecular weight cutoff drift value of the ultrafiltration membrane is estimated. When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane exceeds the drift threshold, the linkage control is executed and the feed load correction factor is generated.

[0008] S2: Calculate the cumulative lipid adsorption amount based on resin parameters and operating parameters, and calculate the micropore blockage coefficient based on the cumulative lipid adsorption amount; when the operating flow rate is stable, use the micropore blockage coefficient to quantify the resin's occult adsorption decay.

[0009] S3: Collect bed pressure drop and outlet velocity data to calculate the dynamic flow resistance coefficient, and compare the dynamic flow resistance coefficient with the reference flow resistance coefficient of fresh packing to obtain the bed compaction factor; when the bed compaction factor fluctuates abnormally, determine whether it is wall flow or channel flow, and combine the micro-pore blockage coefficient to distinguish between physical compaction and lipid contamination blockage.

[0010] S4: Collect target brain peptide chromatographic peak data to calculate real-time resolution and tailing factor, and construct a comprehensive column efficiency decline index by feed load correction factor, micropore blockage coefficient and bed compaction factor, and use the comprehensive column efficiency decline index to correct the real-time resolution.

[0011] S5: Compare the comprehensive column efficiency decline index with the failure threshold. Based on the comparison results, trigger the deep regeneration procedure, backwash resuspension and replacement of packing material, and perform cleaning of the upstream ultrafiltration membrane module or switch to the standby ultrafiltration membrane module.

[0012] The generation of the feed load correction factor includes the following sub-steps:

[0013] S11: Perform feed status windowing adjustment, which includes at least frequency conversion adjustment of the circulating pump speed to maintain constant membrane flow rate, and control the feed liquid temperature fluctuation within a preset small range through heat exchanger.

[0014] S12: Collect transmembrane pressure difference, permeate flux and lipid content of raw material batch, calculate membrane fouling resistance coefficient; perform trend differentiation on membrane fouling resistance coefficient to obtain resistance change rate, and obtain the equivalent molecular weight cutoff drift value of ultrafiltration membrane through exponential function mapping.

[0015] S13: When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane exceeds the drift threshold, linkage control is performed; after linkage control, the additional impurity load is calculated based on the drift amplitude and the concentration distribution of impurity proteins greater than 3000 Da in the feed, and the additional impurity load is normalized to obtain the feed load correction factor.

[0016] The calculation of the micropore blockage coefficient based on the cumulative lipid adsorption amount includes the following sub-steps:

[0017] S21: Integrate the chromatography run time to obtain the lipid adsorption increment for the current cycle, and sum them up to obtain the cumulative lipid adsorption amount;

[0018] S22: The cumulative lipid adsorption amount is mapped to the micropore blockage coefficient using an S-shaped growth nonlinear function. The micropore blockage coefficient increases rapidly after the cumulative lipid adsorption amount reaches the sensitivity threshold.

[0019] S23: When the rate of change of chromatography operating flow rate is lower than the threshold of the rate of change of chromatography operating flow rate and the rate of change of column bed pressure drop is lower than the threshold of the rate of change of column bed pressure drop, the degradation of the microscopic hidden adsorption capacity inside the resin is quantified by the microscopic channel blockage coefficient.

[0020] The process of obtaining the bed compaction factor includes the following sub-steps:

[0021] S31: Collect pressure drop and outlet velocity data of the column bed and calculate the dynamic flow resistance coefficient;

[0022] S32: Compare the dynamic flow resistance coefficient with the reference flow resistance coefficient of fresh packing to obtain the bed compaction factor; when the bed compaction factor fluctuates abnormally, determine whether it is wall flow or channel flow.

[0023] S33: Construct a fault classification matrix: Physical compaction is determined when the bed compaction factor is greater than the compaction threshold and the micropore blockage coefficient is less than the blockage threshold; lipid deposition caking and blockage are determined when the bed compaction factor and the micropore blockage coefficient increase simultaneously.

[0024] The construction of the comprehensive column efficiency decline index includes the following sub-steps:

[0025] S41: Lock the target brain peptide chromatographic peak with retention time within a preset range, calculate the real-time resolution, and calculate the tailing factor from the peak width and half-peak width.

[0026] S42: Multiply the real-time separation degree by the reciprocal of the feed load correction factor and divide by the weighted sum of the bed compaction factor and the micro-channel blockage coefficient to obtain the corrected separation degree;

[0027] S43: The modified separation degree and the tailing factor are normalized and weighted to obtain the comprehensive column efficiency decline index.

[0028] The triggering of the deep regeneration procedure includes the following sub-steps:

[0029] S51: When the overall column efficiency decline index is below the failure threshold, maintain the chromatography process; when the overall column efficiency decline index is above the failure threshold, proceed to the fault diagnosis branch.

[0030] S52: When resin micro-lipid poisoning is determined, a deep regeneration procedure is triggered; the deep regeneration procedure includes at least a high concentration of ethanol or a regeneration reagent containing alkaline additives, and the regeneration soaking time is extended linearly according to the micro-pore blockage coefficient.

[0031] S53: When the physical structure of the resin column is determined to be in failure, stop feeding and perform pulse backwashing until the bed compaction factor returns to the reference range.

[0032] S54: When the comprehensive column efficiency decline index is higher than the failure threshold but the feed load correction factor continues to exceed the feed load correction factor warning line, an upstream ultrafiltration membrane module cleaning command is sent.

[0033] Establishing a synergistic feedforward control mechanism between the equivalent molecular weight cutoff drift value of the ultrafiltration membrane and the deep regeneration process includes the following sub-steps:

[0034] S61: When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane indicates a positive shift in the cutoff boundary and the overall column efficiency decline index is within the normal range, latent contamination is determined, and preventive enhanced regeneration is performed after the current chromatography batch ends.

[0035] S62: The intensity of preventive enhanced regeneration is positively correlated with the equivalent molecular weight cutoff drift of the ultrafiltration membrane. The implementation of preventive enhanced regeneration includes increasing the ethanol concentration from the baseline concentration and reducing the cleaning flow rate of the organic solvent.

[0036] Establishing a synergistic rheological diagnostic mechanism between microscopic pore blockage coefficient and bed compaction factor includes the following sub-steps:

[0037] S71: Perform impulse response tests during the intervals between chromatography column operations and calculate the elastic recovery rate of the bed compaction factor from its peak value to the baseline value;

[0038] S72: Construct the resin aging and hardening index. The resin aging and hardening index is positively correlated with the micropore blockage coefficient and negatively correlated with the elastic recovery rate.

[0039] S73: When the bed compaction factor increases and the resin aging and hardening index exceeds the aging threshold, irreversible lipid hardening and compaction are determined, and an alarm for filler scrapping and replacement is triggered.

[0040] Establishing a synergistic adsorption kinetic hysteresis compensation mechanism between the micropore blockage coefficient and the tailing factor includes the following sub-steps:

[0041] S81: Calculate the covariance between the micro-channel blockage coefficient and the tailing factor. When a significant positive correlation trend is observed, determine the mass transfer-impeded aging stage.

[0042] S82: Activate the fractionation endpoint drift compensation algorithm. The fractionation endpoint drift compensation algorithm calculates the delay time based on the micro-channel blockage coefficient.

[0043] S83: Increase the delay time based on the original collection endpoint, and increase the UV absorbance threshold of the fractionation cut point according to the tailing factor.

[0044] A macroporous resin column efficiency monitoring system for porcine brain peptide purification includes:

[0045] Upstream Steady-State and Drift Monitoring Module: This module regulates the state of the crude porcine brain extract entering the ultrafiltration membrane module, and collects transmembrane pressure difference, permeate flux, and lipid content of each batch of feedstock. It calculates the membrane fouling resistance coefficient and estimates the equivalent molecular weight cutoff drift value of the ultrafiltration membrane. When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane exceeds the drift threshold, it performs linkage control and generates a feed load correction factor.

[0046] Micropore decay modeling module: Calculates the cumulative lipid adsorption amount based on resin parameters and operating parameters, and calculates the micropore blockage coefficient based on the cumulative lipid adsorption amount; when the operating flow rate is stable, the micropore blockage coefficient is used to quantify the resin's hidden adsorption decay.

[0047] Flow field and compaction diagnosis module: Collects column bed pressure drop and outlet flow velocity to calculate dynamic flow resistance coefficient, and compares the dynamic flow resistance coefficient with the reference flow resistance coefficient of fresh packing to obtain the bed compaction factor; when the bed compaction factor fluctuates abnormally, it determines wall flow or channel flow, and combines the micro-pore blockage coefficient to distinguish between physical compaction and lipid contamination blockage.

[0048] Separation feature analysis and correction module: Collect target brain peptide chromatographic peak data to calculate real-time resolution and tailing factor, and construct a comprehensive column efficiency decline index by feed load correction factor, micropore blockage coefficient and bed compaction factor, and use the comprehensive column efficiency decline index to correct the real-time resolution.

[0049] Multidimensional Judgment and Adaptive Decision Module: Compares the comprehensive column efficiency decline index with the failure threshold, and triggers the deep regeneration program, backwashing and resuspension and replacement of packing material according to the comparison results, as well as performing upstream ultrafiltration membrane module cleaning or switching to the standby ultrafiltration membrane module.

[0050] Compared with the prior art, the beneficial effects of the present invention are:

[0051] This invention establishes a mapping model based on cumulative lipid adsorption and a dynamic flow resistance monitoring mechanism based on a variant of Darcy's law to obtain the micropore blockage coefficient and bed compaction factor, respectively. Based on these, a fault classification matrix is ​​constructed to distinguish between microscopic lipid poisoning and macroscopic physical compaction. This reduces the probability of blind backflushing or ineffective regeneration when column pressure increases, and allows for the identification of micropore blockage trends before significant abnormalities in bed pressure drop. This enables the matching of repair strategies such as pulsed backflushing and deep chemical regeneration, reducing unplanned resin packing losses and extending resin lifespan.

[0052] This invention monitors the equivalent molecular weight cutoff drift of the ultrafiltration membrane and triggers preventative enhanced regeneration during the contamination latency period to inhibit the migration of lipid impurities into the depths of the resin micropores. Simultaneously, based on the correlation between the comprehensive column efficiency decline index and the feed load correction factor, it triggers a cleaning command for the upstream ultrafiltration membrane module, thereby reducing false failures of the resin caused by abnormal feed loads. This cross-process linkage mechanism improves the production line's adaptability to batch-to-batch variations in raw materials and maintains column efficiency stability under long-term operating conditions.

[0053] This invention employs a multi-dimensional control strategy targeting the resin's freshness, aging, and end-of-life stages: during operation, a comprehensive column efficiency degradation index is used to quantify separation performance in real time; in the mid-to-late aging stages, a synergistic adsorption kinetic hysteresis compensation mechanism dynamically adjusts the UV absorbance thresholds at the fractionation endpoint and cut-off point based on the microscopic pore blockage coefficient to reduce yield loss caused by retention time drift; in the late aging stage, a synergistic rheological diagnostic mechanism identifies irreversible hardening and caking and triggers an end-of-life warning. This reduces product purity fluctuations and improves overall yield and production efficiency. Attached Figure Description

[0054] Figure 1 The main flowchart of the method for monitoring and controlling the column efficiency of macroporous resins;

[0055] Figure 2 The flowchart for fault diagnosis and adaptive decision-making logic;

[0056] Figure 3 This is a schematic diagram of the system's technological process and control principles;

[0057] Figure 4 This is a block diagram of the system's functional modules. Detailed Implementation

[0058] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0059] Example 1

[0060] This invention proposes a macroporous resin column efficiency monitoring method for the purification of porcine brain peptides, comprising:

[0061] S1: The state of the crude porcine brain extract entering the ultrafiltration membrane module is regulated, and the transmembrane pressure difference, permeate flux and lipid content of the feed batch are collected. The membrane fouling resistance coefficient is calculated and the equivalent molecular weight cutoff drift value of the ultrafiltration membrane is estimated. When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane exceeds the drift threshold, the linkage control is executed and the feed load correction factor is generated.

[0062] S2: Calculate the cumulative lipid adsorption amount based on resin parameters and operating parameters, and calculate the micropore blockage coefficient based on the cumulative lipid adsorption amount; when the operating flow rate is stable, use the micropore blockage coefficient to quantify the resin's occult adsorption decay.

[0063] S3: Collect bed pressure drop and outlet velocity data to calculate the dynamic flow resistance coefficient, and compare the dynamic flow resistance coefficient with the reference flow resistance coefficient of fresh packing to obtain the bed compaction factor; when the bed compaction factor fluctuates abnormally, determine whether it is wall flow or channel flow, and combine the micro-pore blockage coefficient to distinguish between physical compaction and lipid contamination blockage.

[0064] S4: Collect target brain peptide chromatographic peak data to calculate real-time resolution and tailing factor, and construct a comprehensive column efficiency decline index by feed load correction factor, micropore blockage coefficient and bed compaction factor, and use the comprehensive column efficiency decline index to correct the real-time resolution.

[0065] S5: Compare the comprehensive column efficiency decline index with the failure threshold. Based on the comparison results, trigger the deep regeneration procedure, backwash resuspension and replacement of packing material, and perform cleaning of the upstream ultrafiltration membrane module or switch to the standby ultrafiltration membrane module.

[0066] like Figure 3 As shown, the process flow in this embodiment relies on the upstream ultrafiltration unit to perform the core pretreatment steps, and achieves closed-loop monitoring and feedback through the drift and correction module in the top-level control system.

[0067] The generation of the feed load correction factor includes the following sub-steps:

[0068] S11: Perform feed status windowing adjustment, which includes at least frequency conversion adjustment of the circulating pump speed to maintain constant membrane flow rate, and control the feed liquid temperature fluctuation within a preset small range through heat exchanger.

[0069] The first paragraph mainly concerns temperature control and flow rate regulation of rheological properties. Combined with... Figure 3 The hardware configuration of the "upstream ultrafiltration unit" takes into account that the crude extract of pig brain is a non-Newtonian fluid, and even after degreasing, there may still be a certain proportion of residual lipid components in the feed solution, and its viscosity... Temperature The membrane is highly sensitive to changes in shear rate. To reduce transmembrane pressure differential and shear field fluctuations caused by rheological property fluctuations, thereby reducing apparent retention fluctuations due to reversible deformation of the membrane structure, the system incorporates a design in the ultrafiltration circulation loop. Figure 3 The illustrated tubular heat exchanger performs temperature control. Although the enzymatic hydrolysis process temperature is typically... However, to ensure the stability of membrane separation, this system controls the temperature of the feed solution entering the ultrafiltration membrane using a PID controller. Within this range, this avoids changes in liquid viscosity due to temperature fluctuations. Dramatic changes. Meanwhile, the system is configured with... Figure 3 The variable frequency drive circulation pump shown performs constant flow rate control, and the flow sensor is located at the membrane module. Figure 3 The Chinese logo is Real-time feedback of membrane surface flow rate The controller dynamically adjusts the pump speed and frequency based on feedback values. This embodiment sets the tangential flow velocity at the membrane surface. Keep constant as This results in a strong shear disturbance condition within the flow channel, which is related to the generalized Reynolds number. Evaluate and maintain exist Within this range, concentration polarization can be effectively suppressed to reduce the thickness fluctuation of the concentration polarization layer.

[0070] The second section mainly concerns colloid pre-removal and membrane module selection. This involves the process of feed liquid entering... Figure 3 Before the polysulfone ultrafiltration membrane module shown, the system first arranges for the feed liquid to pass through... The microfiltration membrane module, which is integrated into Figure 3 The microfiltration membrane used in the raw material tank outlet pipeline is selected because it can effectively trap suspended particles and microorganisms in the crude pig brain extract, thereby controlling the feed turbidity to the preset threshold in this embodiment. The following measures aim to reduce the impact of particulate clogging and colloidal fouling on the flux and retention performance of the ultrafiltration membrane. The feed solution then enters the filtration membrane with a molecular weight cutoff of [missing information]. The polysulfone ultrafiltration membrane module was selected based on the polysulfone material's resistance to enzymatic hydrolysis and its stable retention accuracy. The retention range can effectively retain large molecular weight proteins and allow molecules smaller than 1000 to be retained. The target brain peptides are absorbed.

[0071] S12: Collect transmembrane pressure difference, permeate flux and lipid content of raw material batch, calculate membrane fouling resistance coefficient; perform trend differentiation on membrane fouling resistance coefficient to obtain resistance change rate, and obtain the equivalent molecular weight cutoff drift value of ultrafiltration membrane through exponential function mapping.

[0072] The first section mainly involves data acquisition and resistance coefficient calculation. During system operation, the ultrafiltration operating pressure is first controlled at [specific pressure range] according to the brain peptide ultrafiltration process standard. , and then rely on Figure 3 The sensor group configured in the ultrafiltration membrane module includes a pressure sensor. With flow sensor To balance data accuracy and controller load Frequency acquisition of transmembrane pressure difference With flux Simultaneously adopt The system processes the sampled signal using a sliding window filter at each sampling point to suppress noise caused by electromagnetic interference in the industrial environment. The system reads lipid content data from the raw material batch information and combines it with real-time apparent viscosity. ,exist Figure 3 The "Drift and Correction" logic module shown calculates the membrane fouling resistance coefficient based on Darcy's law. The calculation formula is: ,in The membrane inherent resistance is obtained from the calibration of the water flux. For real-time acquisition of transmembrane pressure difference, To transmit flux in real time, This represents the viscosity of the liquid at the current temperature and shear rate.

[0073] The second section mainly involves the estimation of drift values ​​and model mapping. The controller addresses the aforementioned membrane fouling resistance coefficient. By performing time differentiation, the rate of change of resistance with time can be obtained. .like The slow, linear growth is interpreted by the system as normal filter cake layer thickening; if... The membrane exhibits a sharp, non-linear increase, and considering the high lipid content of the current feed, the system determines that the membrane pores are at risk of experiencing an equivalent molecular weight cutoff shift and a decrease in apparent retention capacity due to lipid swelling or pressure compression. The system then retrieves the offline-calibrated resistance change rate – permeate side target range. The component leakage rate-equivalent MWCO mapping model outputs the equivalent molecular weight cutoff drift value of the ultrafiltration membrane. For example, the original plan to retain The membrane's equivalent retention capacity under current operating conditions may drift to [value missing]. .

[0074] S13: When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane exceeds the drift threshold, linkage control is performed; after linkage control, the additional impurity load is calculated based on the drift amplitude and the concentration distribution of impurity proteins greater than 3000 Da in the feed, and the additional impurity load is normalized to obtain the feed load correction factor.

[0075] The first section mainly concerns the triggering and execution sequence of the coordinated control. The system sets the drift threshold to... That is, the equivalent retention is greater than Triggered on time. When a drift value is detected. When the standard is exceeded, Figure 3 The drift and correction in the membrane are prioritized to send commands to the upstream ultrafiltration unit to execute linkage control. The control priority and specific operation are performed in the following order: First, adjust the transmembrane pressure difference, and automatically reduce the transmembrane pressure difference. Set value to The following are the most effective and fastest methods to suppress trapping drift; secondly, changing the circulation flow rate and adjusting... Figure 3 The variable frequency drive speed of the circulating pump in the system changes the circulating flow rate back to its original value. To enhance membrane shear force and reduce lipid adsorption, a bypass reflux is activated by opening the bypass reflux valve to return a portion of the concentrate to the ultrafiltration feed buffer tank for remixing, thereby reducing the local lipid concentration peak on the membrane surface and further inhibiting the drift of the cutoff boundary towards macromolecules.

[0076] The second section mainly involves the calculation and application of the correction factor. After the coordinated control is implemented, the system maintains stable operation under the current conditions. If the drift value remains at Right now If extraneous proteins are leaking into the downstream resin system, a feed load correction factor generation step is executed. The system calls the molecular weight distribution database of porcine brain enzymatic hydrolysis products, where the target peptide's main peak is concentrated in... Read the current feed material that is greater than Concentration distribution of extraneous proteins, combined The drift amplitude was used to calculate the additional contaminant load entering the downstream resin system. Subsequently, an empirical formula fitted from historical operating data was used to generate a feed load correction factor. In this embodiment, due to the high lipid content in the feed, a corrected result was calculated. The feed load correction factor pass Figure 3The feedforward loop shown transmits the data to the downstream control module to calibrate the expected theoretical plate number of the downstream resin column, for example, by changing the expected theoretical plate number of the resin column from... Block calibration This ensures that the separation effect of the resin column is not affected by the increase in feed impurities.

[0077] Through the above steps, the system has evolved from only monitoring flux to... The decline was expanded to include simultaneous monitoring of ultrafiltration membrane retention performance drift. Characterization was used to establish a closed-loop control system consisting of steady-state control, drift monitoring, coordinated regulation, and correction factor feedback. In a high-fat, complex-molecular-weight porcine brain crude extract system, this method can accurately quantify downstream impurity load fluctuations caused by changes in membrane retention capacity, and adjust the feed load correction factor accordingly. A feedforward compensation signal is provided to the resin adsorption unit. After actual operation testing, linkage control, and correction, the concentration of [something] in the permeate is greater than [something]. Impurity content reduced The pollution load of downstream resin columns is reduced. Effectively reduced molecular weight greater than This reduces the risk of impurities entering the finished product section, while simultaneously ensuring the stability of the resin column efficiency.

[0078] The calculation of the micropore blockage coefficient based on the cumulative lipid adsorption amount includes the following sub-steps:

[0079] S21: Integrate the chromatography run time to obtain the lipid adsorption increment for the current cycle, and sum them up to obtain the cumulative lipid adsorption amount;

[0080] The first paragraph mainly involves real-time data acquisition and discrete integral calculations. (Combined with...) Figure 3 The hardware layout of the "core chromatography unit" shown indicates that the system is equipped with a [missing information - likely a component or feature] upstream of the chromatography column feed inlet. Figure 3 The instrument is an online near-infrared spectrometer labeled NIR. This device is compatible with monitoring porcine brain peptide feed solutions and is resistant to background interference, with a detection range covering [missing information]. Used for real-time monitoring of residual lipid concentration in the feed entering the chromatography column. If online detection is not available on-site, the system can be based on each A linear interpolation model is established using offline detection data from a single sampling to achieve real-time estimation of residual lipid concentration in the feed. Simultaneously, Figure 3 The flow sensor shown provides real-time feedback on the flow rate during chromatographic operation. In this embodiment, the flow rate is controlled at... That is, per hour Double bed volumetric flow rate, suitable for industrial production The standard range. Figure 3The micropore decay modeling module performs discrete integral calculations on the single-batch runtime to calculate the lipid adsorption increment, i.e., the lipid input dose, for the current cycle. Set the runtime for a single batch. for Sampling period for To balance data accuracy and controller load, the calculation formula is as follows: .in The total number of samples per batch Second-rate, This represents the empirical adsorption coefficient of the resin for lipids. In this embodiment... Values This value was obtained through offline material balance calibration. Specifically, three sets of simulated feed solutions with different lipid concentrations were passed through a fresh resin column. The total lipid mass injected and the lipid mass after resin adsorption were measured and fitted to estimate the actual proportion of lipids entering the resin micropores. This calculation process fully considers the dynamic impact of feed flow rate fluctuations and batch-to-batch differences in lipid residue on the total adsorption amount.

[0081] The second section mainly concerns the total accumulation and renewal throughout the entire life cycle of the resin. Macroporous resins typically require continuous use for dozens to hundreds of batches; this embodiment assumes continuous use of the resin. Replace after each batch. The system establishes a full lifecycle database of the resin, recording the total residual accumulated lipids at the end of each batch. Let the total residual accumulated lipids at the end of the previous batch be... The real-time cumulative lipid adsorption amount after the end of this batch The updated formula is .in In this embodiment, the lipid elution efficiency is that of a conventional regeneration process. This indicates that conventional regeneration can be eluted. Lipids are adsorbed on the surface but cannot be eluted from lipids deep within micropores. Conventional regeneration processes employ... ethanol solution and Flow rate elution elution efficiency This parameter was obtained by measuring the residual lipid content in the resin before and after regeneration. It accurately reflects the residual cumulative lipid contact dose that the resin filler has experienced from its fresh state to the present moment, providing core input parameters for the subsequent calculation of the micropore blockage coefficient.

[0082] S22: The cumulative lipid adsorption amount is mapped to the micropore blockage coefficient using an S-shaped growth nonlinear function. The micropore blockage coefficient increases rapidly after the cumulative lipid adsorption amount reaches the sensitivity threshold.

[0083] The first section mainly involves the construction and parameter calibration of the sigmoid nonlinear mapping model. Based on... Figure 3 The computational logic of the microscopic pore decay modeling module considers that lipid accumulation on resin exhibits a phase transition process of "surface coverage - pore filling - pore sealing." To accurately simulate this process, the system uses a Logistic function with an S-shaped growth trend to construct a mapping model, representing the cumulative lipid adsorption amount. Convert to dimensionless microscopic channel blockage coefficient The model formula is set as follows: ,in The sensitivity threshold, This is the sensitivity coefficient. In this embodiment... Set as The critical lipid accumulation level characterizes the point at which the resin's adsorption performance begins to decline significantly. Set as This characterizes the decay rate of pore blockage. and The results were obtained through accelerated aging tests of the resin under standard operating conditions and joint calibration using data from multiple batches of operation. Specifically, fresh resin was subjected to accelerated aging tests using feed solutions of varying lipid concentrations to determine the dynamic adsorption capacity decay. Time corresponding As a sensitivity threshold And based on different corresponding The sensitivity coefficient is obtained by fitting the Logistic function to the value. This model can accurately map the nonlinear relationship between cumulative lipid adsorption and the degree of pore blockage.

[0084] The second paragraph mainly involves the analysis of the physical meaning of the coefficient changes. Microscopic channel blockage coefficient. The change has a clear physical meaning: when the cumulative lipid adsorption amount Far below the sensitivity threshold That is, less than At this time, lipids mainly cover the resin surface in the form of a monolayer and do not enter the interior of the micropores. The growth is slow and the value is close to The resin's adsorption performance showed virtually no decline; when Approaching and reaching The sensitivity threshold range is... At this time, lipids begin to accumulate in multiple layers and undergo hydrophobic cross-linking inside the resin micropores, leading to a sharp narrowing of the effective mass transfer channels and an increase in mass transfer resistance. It shows a rapid upward trend and from Rise to ;when Much higher That is, greater than At that time, the resin micropores were completely blocked by lipids. Approaching The resin essentially loses its mass transfer and adsorption capacity. This change process accurately simulates the physical process by which lipid impurities, from quantitative to qualitative change, lead to the obstruction of microscopic mass transfer in the resin.

[0085] S23: When the rate of change of chromatography operating flow rate is lower than the threshold of the rate of change of chromatography operating flow rate and the rate of change of column bed pressure drop is lower than the threshold of the rate of change of column bed pressure drop, the degradation of the microscopic hidden adsorption capacity inside the resin is quantified by the microscopic channel blockage coefficient.

[0086] The first paragraph mainly concerns the criteria for determining macroscopic steady state. The blockage of resin micropores is highly concealed and easily confused with the decrease in adsorption efficiency caused by fluctuations in macroscopic operating conditions; therefore, it is necessary to first determine the macroscopic steady state. The system utilizes... Figure 3 Pressure and flow sensors are installed at the inlet and outlet of the chromatography column to monitor the rate of change of the chromatography flow rate in real time. Rate of change of pressure drop in column bed Set a preset micro threshold as well as When detected and and continue At this point, the system determines that the current chromatography system is in a stable hydrodynamic state where "macroscopic operating conditions have not changed drastically." The core purpose of this determination is to eliminate the influence of macroscopic flow rate fluctuations, column bed physical compaction, and other factors on adsorption performance, ensuring that a decrease in adsorption efficiency at this time is more likely related to impaired microscopic mass transfer in the resin or shielding of adsorption sites, rather than physical flow channel blockage or equipment malfunction.

[0087] The second section mainly concerns the quantification and output of microscopic decay. Under the premise of satisfying the aforementioned macroscopic steady state, the system utilizes the microscopic channel blockage coefficient calculated in step S22. The degradation of the microscopic, invisible adsorption capacity within the resin is quantified. The system outputs an equivalent capacity reduction factor based on the calibration model. The calculation formula is: ,in These are the fitting coefficients obtained from multiple batches of resin aging experiments. For example, when the macroscopic pressure is normal but... Value climbed to At that time, the calculation yielded The system output warning indicates that although the fluid channels are clear, the effective dynamic adsorption capacity of the resin has been reduced by approximately [percentage missing]. The equivalent capacity reduction factor Synchronous transmission to Figure 3The "Comprehensive Evaluation" module shown (corresponding to subsequent step S4) works in conjunction with the feed load correction factor and bed compaction factor to construct a comprehensive column efficiency decline index. This quantification process fully realizes the technical requirement of quantifying the decline in the microscopic hidden adsorption capacity inside the resin by utilizing the micropore blockage coefficient when the chromatography flow rate does not change drastically.

[0088] Through the above embodiments, this invention precisely solves the technical problem of "hidden inactivation" of macroporous resin during the purification of porcine brain peptides. In existing technologies, resin inactivation monitoring often relies solely on column bed pressure alarms. However, when a pressure increase is observed, the resin has already undergone severe physical caking, making performance recovery impossible and requiring replacement of the packing material. In contrast, this embodiment calculates the cumulative lipid adsorption amount through discrete integral calculation. The microscopic channel blockage coefficient is obtained by combining the Logistic nonlinear mapping model. It can detect abnormalities in column bed pressure in advance. The monitoring method detected the tendency of resin micropore blockage in each production batch, providing an early warning of hidden deactivation. Actual operation of this production line has verified that the resin regeneration efficiency has been improved after adopting the monitoring method described in this embodiment. Extended service life That is, from Batch extension to Batch size; simultaneously, by proactively avoiding the use of "defective" resin, the yield of the target brain peptide product improved. Purity fluctuations are controlled within Within this range, the risk of loss of precious brain peptide products due to blindly continuing to use "sick" resins has been effectively reduced, and production costs have been lowered.

[0089] The process of obtaining the bed compaction factor includes the following sub-steps:

[0090] S31: Collect pressure drop and outlet velocity data of the column bed and calculate the dynamic flow resistance coefficient;

[0091] The first section mainly involves the acquisition of fluid dynamics parameters and velocity conversion. Based on... Figure 3 The hardware configuration of the "core chromatography unit" includes high-precision pressure sensors installed at both the inlet and outlet of the chromatography column. These sensors have an accuracy of [insert accuracy here]. And compatible With a pressure range and resistance to feed contamination, it is used for real-time acquisition of column bed pressure drop. At the same time, utilizing Figure 3 The electromagnetic flowmeter shown monitors volumetric flow rate. This flow meter is resistant to interference from viscous liquids. (This is combined with the cross-sectional area of ​​the chromatography column.) In this embodiment, the diameter of the chromatography column is [missing information]. Therefore, the cross-sectional area is Calculate real-time linear velocity The formula is This embodiment employs a dual-track flow rate system, meaning the bed volumetric flow rate is used in the process description and controlled to be [value missing]. Linear velocity is used in the formula calculation. The conversion formula between the two is: .in The volume of the column bed is, in this embodiment, Substituting into the calculation, the linear velocity is obtained. Approximately Considering the non-Newtonian fluid properties of the pig brain enzymatic hydrolysis permeate, its viscosity is significantly affected by temperature and shear rate. The system synchronously reads... Figure 3 Real-time feed liquid temperature at the outlet of the heat exchanger of the "upstream ultrafiltration unit" And stabilized at The apparent viscosity of the liquid under the current operating conditions is calculated based on a pre-calibrated temperature-shear rate-apparent viscosity model. In this embodiment, it is stable at The calibration model was obtained by measuring the apparent viscosity of the liquid at different temperatures and shear rates using a rotational rheometer and fitting a quadratic regression model.

[0092] The second section mainly involves the calculation of flow resistance and noise reduction for interference during stripping operations. This is done to eliminate the direct interference of operating flow rate fluctuations and changes in feed viscosity on pressure readings. Figure 3 The flow field and compaction diagnostic module in the system is based on a variant of Darcy's law and calculates the dynamic flow resistance coefficient. The calculation formula is: The aforementioned This coefficient approximates the equivalent permeation resistance of the resin bed and can effectively degrade operating conditions such as flow rate. and viscosity The dynamic influence reflects the physical changes in the pore structure and permeability of the resin bed. To suppress the high-frequency noise caused by pump pulsation, the system uses the calculated... The sequence uses a length of The moving average filtering process is applied to each sampling point. The window length is selected based on a balance between noise suppression and capturing parameter change trends, and the sampling period is consistent with that in Example 2. Filtering can reduce noise interference caused by pump pulsation. The above ensures The values ​​are stable and reliable.

[0093] S32: Compare the dynamic flow resistance coefficient with the reference flow resistance coefficient of fresh packing to obtain the bed compaction factor; when the bed compaction factor fluctuates abnormally, determine whether it is wall flow or channel flow.

[0094] The first section mainly involves the construction and benchmark comparison of the bed compaction factor. After the chromatography column is filled with fresh packing material or backwashed and resuspended, the system performs a standard cleaning procedure, which involves using deionized water... Flow rate cleaning And continuously collect data under this standard operating condition. The flow resistance coefficient is calculated, and the average value is taken as the reference flow resistance coefficient for fresh packing. In this embodiment for During subsequent operation, the flow field and compaction diagnostic module will calculate the dynamic flow resistance coefficient in real time. and By comparison, the dimensionless bed compaction factor was obtained. The calculation formula is: .when exist When the porosity is within the range, it indicates that the bed porosity remains at its initial optimal state; when When this occurs, it indicates an increase in the equivalent permeability resistance of the bed, suggesting a decrease in porosity or localized physical compaction; when If the backwash intensity is too high, it indicates that the bed is loose, which may be caused by excessive backwashing. The backwash intensity should be reduced appropriately.

[0095] The second section primarily concerns the determination of fluctuations in wall flow or channel flow. The system is set to a time length of [duration missing]. A moving window is used to effectively capture flow field fluctuations and avoid misjudgments due to instantaneous interference, and the bed compaction factor within the window is calculated in real time. Standard deviation The calculation formula is: ,in The number of samples within the window. Second-rate, For the window Average value. System preset fluctuation threshold. Timely judgment criteria When detected And the duration exceeds When this occurs, the system determines that there is a non-uniform flow field within the chromatography column, i.e., wall flow or channeling. If the overall pressure drop of the column bed is low, i.e., less than [a certain value], [further information is needed]. Furthermore, if the fluctuations are gradual, it is identified as wall flow; if the pressure drop in the column bed drops instantaneously and oscillates at high frequency, it is identified as channel flow. This method, based on time-domain volatility and duration, can effectively distinguish between steady-state compaction and unstable flow fields.

[0096] S33: Construct a fault classification matrix: Physical compaction is determined when the bed compaction factor is greater than the compaction threshold and the micropore blockage coefficient is less than the blockage threshold; lipid deposition caking and blockage are determined when the bed compaction factor and the micropore blockage coefficient increase simultaneously.

[0097] The first paragraph mainly deals with the classification and determination of physical compaction and chemical blockage. The system in... Figure 3 Under the top-level control framework shown, a physical-chemical coupled fault classification matrix is ​​constructed, and the physical parameter bed compaction factor calculated in this embodiment is used. The chemical parameter micropore blockage coefficient calculated by the micropore decay modeling module in Example 2 is compared with that in Example 2. Perform joint discrimination, among which First, determine the dominant type of physical compaction: when detected... That is, exceeding the compaction threshold, but If the clogging threshold is not reached, the system determines the fault type as physical compaction-dominated. This indicates that the resin particle micropores are relatively clean, and the increased resistance is mainly due to the overly tight particle arrangement leading to decreased bed permeability. The appropriate strategy at this time is to prioritize physical backwashing, i.e., start... Figure 3 The backwash pump at the bottom of the "core chromatography unit" uses deionized water to... The flow rate is used for pulse backwashing, each flush... ,interval Total duration Or reduce the chromatography flow rate to Achieve low-flow-rate bed loosening. Next, determine the dominant type of micropore chemical blockage: when detected... Within the normal range or slightly elevated but At that time, the system determined the fault type to be hidden chemical inactivation. This indicates that the macroscopic flow channels are unobstructed, but the resin micropores have been occupied by lipids. The treatment strategy is to perform the deep chemical regeneration described in Example 2, i.e., using... ethanol solution added by Flow rate soaking Furthermore, the dynamic adsorption capacity is reduced and corrected.

[0098] The second section mainly concerns the identification and treatment of lipid deposition and blockage. When lipid deposition and blockage are detected... and It shows a synchronous upward trend, that is, within a unit of time rate of change greater than and rate of change greater than Ultimately achieved and At that time, the system determined the fault type to be caking and blockage caused by lipid deposition. This indicates that the resin particles were not only physically compacted, but the gaps between the particles and the internal micropores were also filled and cemented by viscous lipids. The treatment strategy is to first perform high-intensity chemical regeneration, using... ethanol solution added by Flow rate soaking To dissolve the lipids, physical backwashing and resuspension are performed to avoid direct backwashing that could cause resin particle breakage. The judgment results and handling strategies of this fault classification matrix will be synchronously transmitted to subsequent steps S4 and S5, providing key physical dimensions for the construction of the comprehensive column efficiency decline index and adaptive decision-making. This fully realizes all the technical requirements for calculating the dynamic flow resistance coefficient, constructing the bed compaction factor, judging flow field anomalies, and distinguishing between physical compaction and chemical blockage.

[0099] Through the above implementation methods, this invention precisely overcomes the technical blind spot in existing industrial chromatography processes where relying solely on column pressure elevation as a single indicator cannot distinguish the root cause of a fault, perfectly implementing steps S31-S33. In the industrial-scale amplification of porcine brain peptide purification, this embodiment constructs a dynamic flow resistance coefficient... This effectively eliminates the dynamic interference between flow rate and feed viscosity, reflecting changes in the physical structure of the bed; furthermore, it utilizes the bed compaction factor... With microscopic channel blockage coefficient The constructed two-dimensional fault classification matrix achieves precise decoupling between physical compaction and chemical blockage, solving the pain points of blind backwashing or ineffective regeneration in existing technologies. Verified through actual operation on this production line, this physical-chemical coupling diagnostic mechanism can effectively address issues before the column bed pressure drop becomes significantly abnormal (i.e., less than [a certain value]). Before, in advance Each production batch identifies the risk of micropore clogging and accurately distinguishes between conditions requiring backflushing and those requiring chemical regeneration; unplanned downtime of resin columns is reduced by approximately The resin filler loss rate is reduced. Improved hydraulic stability of chromatography column The purity fluctuation of the target brain peptide product is controlled within Within this range, the stability and production efficiency of industrial chromatography processes have been significantly improved.

[0100] The construction of the comprehensive column efficiency decline index includes the following sub-steps:

[0101] S41: Lock the target brain peptide chromatographic peak with retention time within a preset range, calculate the real-time resolution, and calculate the tailing factor from the peak width and half-peak width.

[0102] The first section mainly involves the intelligent identification and locking of target brain peptide chromatographic peaks. Based on... Figure 3 The hardware configuration of the "Detection and Fractionation Unit" and the system connection to the chromatography workstation. Figure 3 The online ultraviolet detector shown has a detection accuracy of [missing information]. Preferred in Absorbance data during the elution process were collected at a wavelength that is compatible with the UV absorption characteristics of brain peptides and is resistant to interference from impurities. Because the composition of porcine brain enzymatic hydrolysis products is complex and retention time drift can occur due to column aging, Figure 3The separation feature parsing and correction module in the process uses adaptive window logic to lock the target brain peptide. The specific process is as follows: First, take the nearest... The average target peak center volume of each normal production batch is used as the baseline peak center volume. In this embodiment, Then set The dynamic search window, i.e., the search scope is To accommodate retention time drift; then, a dual constraint is used to prevent false locking, namely, only peak area ratios greater than or equal to are identified within the window. And the peak height is greater than or equal to the adjacent miscellaneous peaks A chromatographic peak that is twice the size of the target peak; if no peak matching the criteria is found within the window, expand the window to [a larger value]. If the peak is still not found, a spectrum anomaly alarm will be triggered. Within this range, the system automatically searches for the elution volume corresponding to the highest peak as the retention time of the target peak for the current batch. It also simultaneously identifies the nearest critical impurity peaks before and after the main peak.

[0103] The second section mainly involves the calculation of real-time separation and tailing factor. The system reads the retention times of the target peak and the nearest critical impurity peak. , and the width of the peak bottom , The width of the peak base is defined as the peak height. Peak width at the specified location, and real-time separation degree calculated. The formula is Meanwhile, to assess the uniformity of adsorption sites in the resin channels, the system was strictly tested according to pharmacopoeia specifications. Calculating the tailing factor using the peak height method Take the peak height of the target peak. Down to Peak width at time And measure the horizontal distance from the peak front (i.e., the rising edge) to the peak midline. The calculation formula is: In this embodiment, the standard separation degree of fresh resin Set as The product purity must be greater than or equal to Requirements; Standard trailing factor Set as This is within the acceptable range for industrial applications. (Calculated in real time) and This serves as the base input data for subsequent model corrections.

[0104] S42: Multiply the real-time separation degree by the reciprocal of the feed load correction factor and divide by the weighted sum of the bed compaction factor and the micro-channel blockage coefficient to obtain the corrected separation degree;

[0105] The first paragraph mainly involves the definition of the correction factor and the establishment of the correction logic. Figure 3 In the top-level control logic shown, this step achieves cross-unit data fusion. Due to fluctuations in feed impurity concentration during actual production, and the unavoidable physical compaction and chemical aging of the resin bed, the original separation degree is used directly. It is difficult to accurately reflect the remaining life of the resin. Therefore, the system introduces three key correction factors from different hardware units: feed load correction factor... From Figure 3 The upstream ultrafiltration unit has a value range of [value range missing]. , A smaller value indicates a higher feed impurity load; bed compaction factor From Figure 3 The flow field diagnosis of the core tomography unit has the following value range: Used to characterize the effect of physical flow channel narrowing and non-ideal flow field on peak broadening; micro-channel blockage coefficient From Figure 3 Microscopic modeling of the core tomographic unit, with values ​​ranging from: It is used to characterize the weakening effect of increased chemical mass transfer resistance on separation efficiency.

[0106] The second section mainly involves the specific calculation of the corrected separation degree. The separation feature analysis and correction module constructs a correction model based on physicochemical coupling to calculate the dimensionless corrected separation degree. The formula model strictly adopts self-consistent boundary conditions: ;

[0107] in, The term is used to eliminate column efficiency underestimation caused by excessive feed impurity load, restoring the resolution to the standard feed level; the denominator contains This is the resin state resistance term, used to quantify the weakening effect of physical compaction and chemical blockage on separation. When the resin is in an ideal state... , And it is a standard feed. When, the denominator is Correcting the separation equal to the original separation degree The boundary conditions are perfectly self-consistent. and These are the weighting coefficients for the effects of physical compaction and chemical blockage on separation degree, respectively. This embodiment is based on... Batch historical operation data regression analysis, set , This indicates that the impact of micropore chemical blockage on separation efficiency has a greater weight than that of physical compaction. This corrected separation efficiency... It accurately reflects the theoretical separation capability determined solely by the chemical activity of the resin under standard feed and standard hydrodynamic conditions.

[0108] S43: The modified separation degree and the tailing factor are normalized and weighted to obtain the comprehensive column efficiency decline index.

[0109] The first section primarily concerns the normalization of a single indicator. To merge the separation index and kurtosis index, which differ in scale, into a unified decision index, the system first normalizes the parameters. For corrected separation, a separation decay term is defined. The calculation formula is: The reference resolution The difference between the value and the standard separation Consistency, that is For the tailing factor, a peak shape deterioration term is defined. The calculation formula is: , The function is to find the minimum value, where the maximum allowed tailing factor is... Set as When the calculation result is greater than or equal to Time to take To prevent exponential overflow.

[0110] The second section mainly concerns the generation and classification of the comprehensive index. The system performs a weighted summation of the two indicators mentioned above to obtain a dimensionless comprehensive column efficiency decline index. The calculation formula is: In this embodiment, based on the industrial requirement that purity takes precedence over yield, a separation degree weight is set. Tail factor weight The final result Values ​​distributed in arrive The grading criteria are as follows: when When the column efficiency is deemed excellent; when When a slight decline in column efficiency is detected, the adaptive control procedure in Example 5 is triggered; when When the column efficiency is severely degraded, a warning is triggered for deep regeneration or packing replacement.

[0111] Through the above implementation methods, this invention constructs an evaluation system that can improve the accuracy of identifying the root causes of column efficiency decline, perfectly implementing the complete steps S41-S43. Traditional column efficiency evaluation only focuses on the original separation. Often, when there is an increase in impurities in the feed material, the resin is mistakenly judged to be ineffective, with a false judgment rate greater than or equal to [missing information]. Or, if the feed material is relatively clean, the micro-aging of the resin may be masked, resulting in a false negative rate greater than or equal to [missing information]. This embodiment introduces... The correction effectively reduces the interference from feed load fluctuations, lowering the misjudgment rate to a minimum. The following; by introducing and The weighted correction incorporates the specific impacts of resin physical compaction and chemical blockage on separation capacity into the evaluation model, enabling precise differentiation of the root causes of column efficiency decline, with an identification accuracy greater than or equal to [percentage missing]. The final generated comprehensive column efficiency decline index It is no longer simply chromatographic parameters, but a quantitative input indicator that integrates upstream feed fluctuations, midstream physical field changes, and downstream chemical field evolution. This, in conjunction with Examples 1 to 3, forms a complete closed loop of "feed control - status monitoring - column efficiency evaluation - decision control." Actual operation of this production line has verified that adopting the evaluation system of this example improves the accuracy of resin life management. Product purity fluctuations from Controlled Within this period, the yield increased. This provides reliable quantitative input support for precise lifetime management of the brain peptide purification process and adaptive decision-making in the subsequent step S5.

[0112] The triggering of the deep regeneration procedure includes the following sub-steps:

[0113] S51: When the overall column efficiency decline index is below the failure threshold, maintain the chromatography process; when the overall column efficiency decline index is above the failure threshold, proceed to the fault diagnosis branch.

[0114] The first section mainly involves the initialization and steady-state determination of the main process. For example... Figure 2 As shown at the beginning of the process, the control program starts from the "Start" node. The system monitors the running status of the chromatography column in real time. When the "Batch Chromatography End" signal is confirmed, the "Data Input" interface is immediately activated, and the real-time running data generated in the previous steps is retrieved from the database. This embodiment sets a failure threshold. for The system is based on Execute in real time for a periodic period Figure 2 The core diamond-shaped decision box in the middle This determination is the dividing line between steady-state production and abnormal handling. The normal branch is when a problem is detected... (Right now Figure 2 When the "Yes" path is selected, it indicates that the resolution and peak shape of the current chromatography column meet the quality standards, and the system determines that the column efficiency is normal. At this time, the control system issues a command to maintain the current process parameters, keeping the flow rate, sample loading volume, and cutoff threshold unchanged, and directly guides the program to the next batch of separation tasks until the entire production plan flows to the end node. The fault branch occurs when a fault is detected. (Right now Figure 2When the system selects the "No" path (in the command line), it indicates that the performance of the chromatography column has significantly deteriorated, and continued production will face the risk of decreased yield or excessive purity. The system immediately cuts off the normal production loop, and the instruction program enters the fault diagnosis branch. Based on decision tree logic, this branch first retrieves the fluid dynamics parameter bed compaction factor Kc as the first diversion criterion, and accurately identifies the dominant root cause of the column efficiency decline through a multi-level series of diamond-shaped decision boxes.

[0115] S52: When resin micro-lipid poisoning is determined, a deep regeneration procedure is triggered; the deep regeneration procedure includes at least a high concentration of ethanol or a regeneration reagent containing alkaline additives, and the regeneration soaking time is extended linearly according to the micro-pore blockage coefficient.

[0116] The first section mainly involves troubleshooting and determining the mechanism of chemical faults. Upon entering the fault diagnosis branch, if the previous determination of physical compaction was negative (i.e....), (Not exceeding the standard), the system further retrieves the chemical dimension parameter, micropore blockage coefficient. ,implement Figure 2 The central diamond-shaped decision box "Kc is normal and" >Threshold? When the judgment result is "yes," it indicates that the macroscopic flow channels of the bed are still unobstructed, but the interior of the resin micropores has been occupied by accumulated lipid impurities, leading to increased mass transfer resistance. Based on this, the system outputs the diagnostic result "Judgment: Microscopic lipid poisoning or occult inactivation." For this chemical fault, the system automatically triggers the "deep regeneration program." "Linearly extend soaking time." This procedure uses a regeneration reagent composed of high-concentration ethanol and alkaline additives. The preferred formulation in this embodiment is... Adding ethanol aqueous solution The aim is to use the strong lipophilicity of high-concentration organic solvents to dissolve hydrophobic lipids, and to use the saponification effect of alkaline solutions to convert oils into soluble fatty acid salts, thereby synergistically removing stubborn contaminants in micropores.

[0117] The second section mainly concerns the linear adaptive control of regeneration time. To ensure thorough regeneration while avoiding solvent waste and time loss due to over-washing, the system strictly adheres to... Figure 2 "According to" An intelligent control strategy of "linearly extending soaking time" is employed. A base regeneration time is set. for Time extension factor for The formula for calculating regeneration time is: For example, when Climb to At that time, the system will automatically extend the soaking time to After regeneration is complete, the system executes the standard rinsing procedure and transfers the process to the "restore production" node, ultimately "ending" the current control cycle.

[0118] S53: When the physical structure of the resin column is determined to be in failure, stop feeding and perform pulse backwashing until the bed compaction factor returns to the reference range.

[0119] The first section primarily concerns the identification and hydraulic repair of purely physical faults. The system first executes... Figure 2 The diamond-shaped decision box on the left reads "Kc > threshold and..." "Normal?" When the judgment result is "yes", that is, the bed compaction factor... but If the reading is within the normal range, it indicates that the resin channels are clean, and the fault stems only from excessively tight particle packing in the bed or uneven flow channels in certain areas. The system outputs a diagnostic result of "Judgment: Physical compaction or channeling". At this point, the system executes... Figure 2 The corresponding execution box is "Pause feed and pulse backwash until Kc returns to baseline". This program starts the backwash pump, using deionized water... High-velocity flow impacts the bed from bottom to top, and employs operational... stop The pulsed mode utilizes hydraulic shear force and bed relaxation effect to break the mechanical interlocking between particles. Backwashing continues until real-time monitoring detects... It fell back and stabilized at The baseline range. After the repair is completed, the process moves to "Resume Production".

[0120] The second section mainly concerns the identification and robust repair of severe complex faults. If all of the above single fault determinations are "no", the system continues execution. Figure 2 The lower middle diamond-shaped decision box shows "Kc abnormally high and..." "Abnormally high?" When detected... and All significantly exceeded the standard (i.e. Figure 2 When the path is "yes" (in the system's diagnostic code), it indicates that the resin particles are not only physically compacted, but the interparticle spaces and micropores are also cemented by viscous lipids. The system outputs the diagnostic result as "Judgment: Lipid Deposition and Cementing". For this serious fault, simple backwashing is ineffective; the system then executes... Figure 2 The "powerful chemical cleaning and resuspension" strategy is as follows: First, a strong alkaline solvent is pumped in to soak and dissolve the cementitious material. After the bed is loosened, mechanical stirring or high-flow-rate backwashing is performed to achieve "resuspension" and restructure the bed. After the repair is completed, "production is resumed".

[0121] S54: When the comprehensive column efficiency decline index is higher than the failure threshold but the feed load correction factor continues to exceed the feed load correction factor warning line, an upstream ultrafiltration membrane module cleaning command is sent.

[0122] The first section mainly involves cross-process source tracing diagnosis and linkage feedback. When all preceding judgments regarding the resin body in the fault diagnosis branch are "no", the system retrieves the upstream parameter "feed load correction factor" (i.e. Figure 2 In ),implement Figure 2 The diamond-shaped judgment box in the lower right corner asks, "Resin parameters are normal, but K remains consistently low?". If the resin itself is found to have... and All are within the normal range, but The line remains below the warning line (i.e. Figure 2 The “Yes” path in the system indicates that the column efficiency decline of the chromatography column is not due to resin failure, but rather to the overload of impurities in the upstream feed. The system outputs the diagnostic result as “Judgment: Upstream ultrafiltration membrane aging”.

[0123] The second section mainly concerns the closed-loop execution and fallback mechanism of the cleaning command. After confirming that the problem is caused by upstream load, the system executes... Figure 2 The corresponding execution box is "Send upstream ultrafiltration module cleaning instruction cross-section feedback". Upon receiving this instruction, the pretreatment unit immediately pauses feeding and performs CIP online cleaning on the ultrafiltration membrane module to restore membrane flux and retention performance, eliminating potential downstream hazards. If... Figure 2 All automatic diagnostic paths, including the upstream aging detection, returned "No", meaning... The values ​​are normal, but If the levels still exceed the limits, the system determines that an unknown and complex fault has occurred, triggering a "manual intervention inspection" signal in the fallback phase, prompting operators to conduct on-site troubleshooting. Whether through automatic cleaning or manual inspection, after all repair actions are completed, the system converges to the "restore production" node and ultimately flows to the "end" node, completing the closed-loop intelligent control of the entire lifecycle.

[0124] Through the above embodiments, this invention establishes an intelligent chromatography production system with self-healing capabilities. Traditional chromatography control is often isolated between upstream and downstream processes, and the regeneration procedure is fixed and rigid. This step not only achieves differentiated and precise repair for physical compaction and chemical poisoning, significantly reducing reagent consumption and labor costs; it also innovatively introduces a cross-unit feedback mechanism in step S54, which can identify and resolve pseudo-resin failure caused by upstream ultrafiltration membrane aging. Actual production data shows that this adaptive control mechanism extends the effective service life of macroporous resin. The consumption of regenerated solvent has been reduced. And control the unplanned downtime rate of the entire production line to within The following significantly improves the overall operational efficiency of the brain peptide automated production line.

[0125] Establishing a synergistic feedforward control mechanism between the equivalent molecular weight cutoff drift value of the ultrafiltration membrane and the deep regeneration process includes the following sub-steps:

[0126] S61: When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane indicates a positive shift in the cutoff boundary and the overall column efficiency decline index is within the normal range, latent contamination is determined, and preventive enhanced regeneration is performed after the current chromatography batch ends.

[0127] Specifically, the first section mainly involves real-time logic determination of cross-process parameters. The control system is a Siemens S7-1500 PLC, which supports real-time acquisition and logic determination of cross-process parameters. The sampling period is consistent with Examples 1 to 4. The system acquires two core parameters in real time: the equivalent molecular weight cutoff drift of the ultrafiltration membrane calculated in the preceding ultrafiltration steady-state control step. And the comprehensive column efficiency decline index calculated in step S43 of Example 4. When detected It is a positive value and exceeds the preset drift threshold. That is, greater than And the duration of this state is greater than or equal to To avoid misjudgment due to transient fluctuations, this indicates a positive shift in the upstream ultrafiltration membrane's retention boundary, meaning that molecules that should have been retained were greater than or equal to... Larger molecular weight lipid impurities are leaking into downstream chromatography processes. At this time, if the overall column efficiency degradation index of the downstream chromatography column is simultaneously detected... It is still within the normal range, i.e., less than the failure threshold. In this embodiment Normal range This indicates that these additional lipids that have just entered the resin bed have not yet caused a decrease in macroscopic column efficiency and have not caused irreversible resin fouling. The system will then... Exceeding the threshold and continuing If the failure threshold is not triggered, it is determined to be resin latent contamination.

[0128] The second section mainly concerns the issuance of preventative operational instructions and emergency response. Once latent contamination is determined, the system will not interrupt the current chromatography batch. Instead, it will associate the alarm signal with the current production batch number, simultaneously executing a control system pop-up notification and recording the batch in the production log. This interruption will continue until the batch's chromatography elution process is completed, the material is discharged, and before entering the regeneration stage. The system automatically switches programs, replacing the regular regeneration program with a preventative enhanced regeneration program. The parameters for the regular regeneration program are: Ethanol Mixed reagents, to Flow rate soaking This strategy utilizes the production batch intervals to promptly remove any excess lipid load that has just been adsorbed onto the resin surface or superficial layers, preventing it from diffusing into the deeper micropores of the resin and undergoing irreversible hydrophobic cross-linking. If the current batch is an emergency production batch and preventative regeneration cannot be performed immediately, the system will escalate the alarm signal, prompting operators to shorten the subsequent run time for that batch, and prioritizing preventative regeneration for the next batch. After preventative regeneration is completed, the system monitors in real time. Need to be reduced to the following, Need to be reduced to After confirming that the pollution risk has been completely eliminated, normal production parameters will be restored.

[0129] S62: The intensity of preventive enhanced regeneration is positively correlated with the equivalent molecular weight cutoff drift of the ultrafiltration membrane. The implementation of preventive enhanced regeneration includes increasing the ethanol concentration from the baseline concentration and reducing the cleaning flow rate of the organic solvent.

[0130] The first paragraph mainly concerns the synergistic regulation of regeneration reagent concentration. The intensity of preventative enhanced regeneration and... The correlation is positive. First, the ethanol concentration in the regeneration reagent is adjusted. The composition of the regeneration reagent is consistent with conventional regeneration, and ethanol is added... Mix reagents, adjusting only the ethanol concentration. Set a baseline ethanol concentration. for Concentration adjustment coefficient for The basis is Each increase The ethanol concentration needs to be increased. To ensure effective dissolution; upper limit of ethanol concentration for To avoid exceeding equipment limits and the risk of azeotropic reaction, the ethanol concentration in the regeneration reagent should be... The calculation formula is For example, when Offset to At that time, the system automatically increases the ethanol concentration to ;like Shift to extreme anomalies such as The calculated value exceeds Then according to Execution. Ethanol concentration adjustment was achieved using a Milton Roy M920 online reagent preparation system, with an adjustment accuracy of [insert accuracy here]. .

[0131] The second section primarily concerns the coordinated adjustment of the cleaning flow rate. While enhancing the dissolving power of the regeneration reagent, the system simultaneously reduces the cleaning flow rate of the regeneration reagent to increase the contact time between the reagent and the resin. A baseline regeneration flow rate is set. for Flow rate adjustment coefficient for The basis is Each increase Flow rate decreases Lower limit of flow rate for This is to prevent uneven reagent distribution caused by excessively low flow rates. Regeneration reagent washing flow rate. The calculation formula is , This is a function to find the maximum value. For example, when... Offset to At that time, the flow rate decreased to ;like Offset to The calculated value is lower than Then according to Execution. Through this synergistic adjustment strategy of concentration increase and deceleration, a synergistic feedforward linkage regulation mechanism between ultrafiltration and chromatography is formed. Practical operation has verified that this mechanism can reduce the risk of lipid contamination accumulation caused by upstream ultrafiltration membrane performance drift. The above improves lipid removal rate. This mechanism, in conjunction with the control strategies of Examples 1 to 5, forms a complete closed loop of ultrafiltration steady-state control, resin condition monitoring, column efficiency evaluation, feedforward early warning, and preventive regeneration, ensuring the long-term stable operation of the production line.

[0132] The establishment of a synergistic rheological diagnostic mechanism for micropore blockage coefficient and bed compaction factor includes the following sub-steps: S71: Perform impulse response test during the interval of the chromatography column operation and calculate the elastic recovery rate of the bed compaction factor from the peak value to the reference value;

[0133] The first section primarily concerns dynamic testing methods for the operational gap of the chromatography column. The control system utilizes the time interval between the end of elution and the start of regeneration of the chromatography column. During the intervals between operations, with the feed solution completely emptied from the chromatography column and the column bed in a stable state, a flow impulse response test is performed. The feed pump used for the test must support rapid speed adjustment, with a speed adjustment accuracy of [insert accuracy here]. Speed ​​adjustment time less than or equal to The specific operation is as follows: control the feed pump at... Instantly increase flow rate to Double the reference flow rate and maintain To apply a stable hydraulic compression load, the reference flow velocity is the same as in the previous Example 3. ; then in The internal flow velocity suddenly drops to The system enters the column bed relaxation stage at a flow rate twice the reference velocity. During this process, the system... The sampling period is maintained in real time, following the method described in Example 3. The calculation logic calculates and records the bed compaction factor. The change curve. During the relaxation phase, From the peak scope Gradually falling back to the baseline value scope .

[0134] The second section primarily concerns the quantification of the elastic recovery rate. The system is based on the Kelvin-Voigt viscoelastic model for the relaxation phase... The variation curve is fitted, and the goodness of fit must meet the following requirements. To ensure reliable results, the Kelvin-Voigt model fitting formula is as follows: ,in Relaxation time, in units of ; The relaxation time constant is expressed in units of 1 / 2 t. Elastic recovery rate is a core parameter characterizing the viscoelastic properties of resins. Defined as relaxation phase The rate of fall is calculated using the following formula: The unit is Fresh resin has good elasticity. Usually in Within the range; while excessive lipids were adsorbed, i.e., the lipid content was greater than or equal to Furthermore, aged resin exhibits significant viscous hysteresis. It will significantly decrease to the following.

[0135] S72: Construct the resin aging and hardening index. The resin aging and hardening index is positively correlated with the micropore blockage coefficient and negatively correlated with the elastic recovery rate.

[0136] The first section primarily involves the collaborative modeling of multidimensional parameters. Neither a simple decrease in physical elasticity nor simple chemical micropore blockage is sufficient to comprehensively describe the aging degree of the resin; therefore, a comprehensive resin aging and hardening index is systematically constructed. The core of this index reflects the synergistic effect of chemical blockage and decreased physical elasticity. This index is similar to the micropore blockage coefficient defined in Example 2. Positive correlation, A larger value indicates a greater amount of lipid filling within the micropores and a more significant resin skeleton hardening effect; this is also related to the elastic recovery rate calculated in this embodiment. Negative correlation The smaller the value, the more severe the adhesive bonding between resin particles and the weaker the elasticity. The core value is to quantify the irreversibility of resin aging, providing a basis for the determination of step S73.

[0137] The second section mainly involves the specific calculation of the aging hardening index. Key system parameters include: standard elastic recovery rate. for Consistent with the range of fresh resin; weighting coefficient for , characterization Influence weights; weight coefficients for , characterization The influence weight of the resin aging and hardening index. The calculation formula uses a normalized weighted model to ensure that it is arrive Dimensionless values ​​between:

[0138] ;

[0139] Boundary condition verification: when the resin is in a fresh state... , hour, The logic is self-consistent; when the resin is severely aged... , hour, Approaching It is in a state of severe hardening and aging; conversely, when At that time, the resin was only slightly aged and still had repair value.

[0140] S73: When the bed compaction factor increases and the resin aging and hardening index exceeds the aging threshold, irreversible lipid hardening and compaction are determined, and an alarm for filler scrapping and replacement is triggered.

[0141] The first paragraph primarily concerns the final determination of the nature of the fault. The system monitors the bed compaction factor in real time. With resin aging and hardening index When detected Greater than the compaction threshold scope And continued When this occurs, it indicates that macroscopic compaction has occurred in the bed, and the system enters the decision branch: if Less than the aging threshold scope The system was determined to be reversible physical compaction, and only triggered a backwashing and resuspension procedure at a flow rate of [missing information]. The time is ;like Greater than or equal to the aging threshold And simultaneously satisfy If the resin hardens and hardens, it is determined to be irreversible lipid hardening and hardening, and the resin cannot be repaired by backwashing or chemical regeneration.

[0142] The second section mainly concerns the triggering of the scrapping and replacement process. Once irreversible lipid hardening and caking are determined, forced use will result in a column pressure increase greater than or equal to [a certain value]. Product yield decreased by more than or equal to Therefore, the control system immediately executes a locking operation, closing the feed valve and regeneration valve of the chromatography column, disconnecting the column bed from the upstream and downstream sections, stopping feeding and regeneration operations, and sending a packing waste and replacement alarm to the distributed control system. After the alarm is sent, the operator must... After the packing is replaced, a flow impulse response test is performed. If... Restore to , , If the verification is successful, the resin can be reused. This mechanism avoids pointless regeneration attempts on completely degraded resin and reduces the consumption of regeneration reagents. In addition, operators are guided to replace the packing material in a timely manner to ensure the overall operating efficiency of the production line and the safety and quality of the products.

[0143] Establishing a synergistic adsorption kinetic hysteresis compensation mechanism between the micropore blockage coefficient and the tailing factor includes the following sub-steps:

[0144] S81: Calculate the covariance between the micro-channel blockage coefficient and the tailing factor. When a significant positive correlation trend is observed, determine the mass transfer-impeded aging stage.

[0145] The control system takes the chromatographic data at the end of each batch of chromatography runs as input and establishes a length of [length missing]. Batch sliding data window, preferred Batch. The system reads the number of records in the window in real time. Batch micropore blockage coefficient and trailing factor ,in Calculated from Example 2, The results were obtained by fitting the chromatographic peaks from this batch. The system calculates the covariance of these two parameter sequences. and Pearson correlation coefficient The judgment criteria employ a triple logic of correlation, significance, and deterioration trend: when a correlation coefficient is detected... Greater than or equal to significance level Less than And within the window The slope of the linear fit is greater than batch When, it indicates The rise and The increase in [value] shows a significant positive correlation and is in a continuously deteriorating trend. This positive correlation characterizes the synchronous change of increased mass transfer resistance within the resin and intensified peak tailing, indicating that the diffusion path is restricted due to the reduction in the effective pore size of the micropores. Therefore, when the above conditions are met simultaneously, the system determines that the chromatography column has entered the mass transfer-impeded aging stage and triggers subsequent compensation and correction steps.

[0146] S82: Activate the fractionation endpoint drift compensation algorithm. The fractionation endpoint drift compensation algorithm calculates the delay time based on the micro-channel blockage coefficient.

[0147] Once the system is determined to have entered the mass transfer-impeded aging stage, the fractionation endpoint drift compensation algorithm is activated. The algorithm is based on the increment of the micro-channel blockage coefficient. Calculate the time delay at the end of the fractionation. ,in The baseline clogging factor for fresh resin is given, and its range is [value missing]. Set the maximum allowed delay time. for Time compensation coefficient for Delay time amount The calculation formula uses a saturation model:

[0148] ;

[0149] For example, when the benchmark And detected Climb to At that time, the calculation yielded Approximately If the value does not exceed the upper limit, the calculated value will be used. The system will... Stacked up to the original standard elution and collection endpoint time Above, a new dynamic time window hard upper limit is formed. This compensation measure ensures that the main peak of the target product, which is delayed due to diffusion obstruction, still has the opportunity to be collected, avoiding yield losses caused by premature cutoff.

[0150] S83: Increase the delay time based on the original collection endpoint, and increase the UV absorbance threshold of the fractionation cut point according to the tailing factor.

[0151] While extending the upper limit of collection time, to prevent the intensification of tailing and the mixing of low-purity components, the system is configured according to... Dynamically increase the UV absorbance threshold at the fractionation cut point In this embodiment Defined as the stop-collection threshold, i.e., when the outlet UV absorbance drops to this value and remains thereafter. Stop collecting when the time is right. Set a baseline absorbance threshold for fresh resin. for Nonlinear adjustment coefficient for upper limit of threshold for Standard trailing factor for Corrected UV absorbance threshold The calculation formula is:

[0152] ;

[0153] For example when Deterioration to hour, From Up to approximately Fractionation control employs a dual-criteria strategy: the primary criterion is the real-time absorbance. And continue Immediately stop the collection of the target component; the hard upper limit criterion is that if the time hard upper limit is reached... The time is still not satisfied If the condition is not met, collection will be forcibly stopped and the system will switch to the waste discharge pipeline. This mechanism achieves a dynamic optimal balance between yield and purity in the later stages of resin aging through a strategy of maintaining purity with a threshold and preventing over-collection over a certain period of time.

[0154] Example 2

[0155] See Figure 4 A macroporous resin column efficiency monitoring and control system for the purification of porcine brain peptides, comprising:

[0156] Upstream Steady-State and Drift Monitoring Module: This module regulates the state of the crude porcine brain extract entering the ultrafiltration membrane module, and collects transmembrane pressure difference, permeate flux, and lipid content of each batch of feedstock. It calculates the membrane fouling resistance coefficient and estimates the equivalent molecular weight cutoff drift value of the ultrafiltration membrane. When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane exceeds the drift threshold, it performs linkage control and generates a feed load correction factor.

[0157] Micropore decay modeling module: Calculates the cumulative lipid adsorption amount based on resin parameters and operating parameters, and calculates the micropore blockage coefficient based on the cumulative lipid adsorption amount; when the operating flow rate is stable, the micropore blockage coefficient is used to quantify the resin's hidden adsorption decay.

[0158] Flow field and compaction diagnosis module: Collects column bed pressure drop and outlet flow velocity to calculate dynamic flow resistance coefficient, and compares the dynamic flow resistance coefficient with the reference flow resistance coefficient of fresh packing to obtain the bed compaction factor; when the bed compaction factor fluctuates abnormally, it determines wall flow or channel flow, and combines the micro-pore blockage coefficient to distinguish between physical compaction and lipid contamination blockage.

[0159] Separation feature analysis and correction module: Collect target brain peptide chromatographic peak data to calculate real-time resolution and tailing factor, and construct a comprehensive column efficiency decline index by feed load correction factor, micropore blockage coefficient and bed compaction factor, and use the comprehensive column efficiency decline index to correct the real-time resolution.

[0160] Multidimensional Judgment and Adaptive Decision Module: Compares the comprehensive column efficiency decline index with the failure threshold, and triggers the deep regeneration program, backwashing and resuspension and replacement of packing material according to the comparison results, as well as performing upstream ultrafiltration membrane module cleaning or switching to the standby ultrafiltration membrane module.

[0161] In addition, the parts of the technical solutions provided in the embodiments of this application that are consistent with the implementation principles of the corresponding technical solutions in the prior art have not been described in detail, so as to avoid excessive elaboration.

[0162] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0163] The preset parameters or preset thresholds mentioned above are all set by those skilled in the art based on actual conditions or obtained through large-scale data simulation.

[0164] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A method for monitoring the column efficiency of macroporous resin for the purification of porcine brain peptides, characterized in that, include: S1: Collect transmembrane pressure difference, permeate flux and lipid content of feed batch, calculate membrane fouling resistance coefficient, and estimate the equivalent molecular weight cutoff drift value of ultrafiltration membrane; when the equivalent molecular weight cutoff drift value of ultrafiltration membrane exceeds the drift threshold, perform linkage control and generate feed load correction factor; S2: Obtain resin parameters and operating parameters, calculate the cumulative lipid adsorption amount, and calculate the micropore blockage coefficient based on the cumulative lipid adsorption amount; When the operating flow rate is stable, the degradation of resin's occult adsorption is quantified by the micropore blockage coefficient. S3: Collect the pressure drop and outlet velocity of the column bed to calculate the dynamic flow resistance coefficient, and compare the dynamic flow resistance coefficient with the reference flow resistance coefficient of fresh packing to obtain the bed compaction factor; when the bed compaction factor fluctuates abnormally, determine whether it is wall flow or channel flow, and combine the micro-pore blockage coefficient to distinguish between physical compaction and lipid contamination blockage. S4: Collect target brain peptide chromatographic peak data to calculate real-time resolution and tailing factor, and construct a comprehensive column efficiency decline index by feed load correction factor, micropore blockage coefficient and bed compaction factor, and use the comprehensive column efficiency decline index to correct the real-time resolution. S5: Compare the comprehensive column efficiency decline index with the failure threshold. Based on the comparison results, trigger the deep regeneration procedure, backwash resuspension and replacement of packing material, and perform cleaning of the upstream ultrafiltration membrane module or switch to the standby ultrafiltration membrane module.

2. The method for monitoring the column efficiency of macroporous resin for purifying porcine brain peptides according to claim 1, characterized in that, The generated feed load correction factor includes: S11: Perform feed status windowing adjustment, which includes at least frequency conversion adjustment of the circulating pump speed to maintain constant membrane flow rate, and control the feed liquid temperature fluctuation within a preset small range through heat exchanger. S12: Collect transmembrane pressure difference, permeate flux and lipid content of raw material batch, calculate membrane fouling resistance coefficient; perform trend differentiation on membrane fouling resistance coefficient to obtain resistance change rate, and obtain the equivalent molecular weight cutoff drift value of ultrafiltration membrane through exponential function mapping. S13: When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane exceeds the drift threshold, linkage control is performed; after linkage control, the additional impurity load is calculated based on the drift amplitude and the concentration distribution of impurity proteins greater than 3000 Da in the feed, and the additional impurity load is normalized to obtain the feed load correction factor.

3. The method for monitoring the column efficiency of macroporous resin for purifying porcine brain peptides according to claim 2, characterized in that, The calculation of the micropore blockage coefficient based on the cumulative lipid adsorption amount includes: S21: Integrate the chromatography run time to obtain the lipid adsorption increment for the current cycle, and sum them up to obtain the cumulative lipid adsorption amount; S22: The cumulative lipid adsorption amount is mapped to the micropore blockage coefficient using an S-shaped growth nonlinear function. The micropore blockage coefficient increases rapidly after the cumulative lipid adsorption amount reaches the sensitivity threshold. S23: When the rate of change of chromatography operating flow rate is lower than the threshold of the rate of change of chromatography operating flow rate and the rate of change of column bed pressure drop is lower than the threshold of the rate of change of column bed pressure drop, the degradation of the microscopic hidden adsorption capacity inside the resin is quantified by the microscopic channel blockage coefficient.

4. The method for monitoring the column efficiency of macroporous resin for purifying porcine brain peptides according to claim 3, characterized in that, The obtained bed compaction factor includes: S31: Collect pressure drop and outlet velocity data of the column bed and calculate the dynamic flow resistance coefficient; S32: Compare the dynamic flow resistance coefficient with the reference flow resistance coefficient of fresh packing to obtain the bed compaction factor; when the bed compaction factor fluctuates abnormally, determine whether it is wall flow or channel flow. S33: Physical compaction is determined when the bed compaction factor is greater than the compaction threshold and the micropore blockage coefficient is less than the blockage threshold; lipid deposition caking and blockage are determined when the bed compaction factor and the micropore blockage coefficient increase simultaneously.

5. The method for monitoring the column efficiency of macroporous resin for purifying porcine brain peptides according to claim 4, characterized in that, The construction of the comprehensive column efficiency decline index includes: S41: Lock the target brain peptide chromatographic peak with retention time within a preset range, calculate the real-time resolution, and calculate the tailing factor from the peak width and half-peak width. S42: Multiply the real-time separation degree by the reciprocal of the feed load correction factor and divide by the weighted sum of the bed compaction factor and the micro-channel blockage coefficient to obtain the corrected separation degree; S43: The modified separation degree and the tailing factor are normalized and weighted to obtain the comprehensive column efficiency decline index.

6. The method for monitoring the column efficiency of macroporous resin for purifying porcine brain peptides according to claim 5, characterized in that, The triggering deep regeneration procedure includes: S51: When the overall column efficiency decline index is below the failure threshold, maintain the chromatography process; when the overall column efficiency decline index is above the failure threshold, proceed to the fault diagnosis branch. S52: When resin micro-lipid poisoning is determined, a deep regeneration procedure is triggered; the deep regeneration procedure includes at least a high concentration of ethanol or a regeneration reagent containing alkaline additives, and the regeneration soaking time is extended linearly according to the micro-pore blockage coefficient. S53: When the physical structure of the resin column is determined to be in failure, stop feeding and perform pulse backwashing until the bed compaction factor returns to the reference range. S54: When the comprehensive column efficiency decline index is higher than the failure threshold but the feed load correction factor continues to exceed the feed load correction factor warning line, an upstream ultrafiltration membrane module cleaning command is sent.

7. The method for monitoring the column efficiency of macroporous resin for purifying porcine brain peptides according to claim 6, characterized in that, Establishing a synergistic feedforward control mechanism between the equivalent molecular weight cutoff drift value of the ultrafiltration membrane and the deep regeneration process includes: S61: When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane indicates a positive shift in the cutoff boundary and the overall column efficiency decline index is within the normal range, latent contamination is determined, and preventive enhanced regeneration is performed after the current chromatography batch ends. S62: The intensity of preventive enhanced regeneration is positively correlated with the equivalent molecular weight cutoff drift of the ultrafiltration membrane. The implementation of preventive enhanced regeneration includes increasing the ethanol concentration from the baseline concentration and reducing the cleaning flow rate of the organic solvent.

8. The method for monitoring the column efficiency of macroporous resin for purifying porcine brain peptides according to claim 7, characterized in that, Establishing a synergistic rheological diagnostic mechanism between microscopic pore blockage coefficient and bed compaction factor includes: S71: Perform impulse response tests during the intervals between chromatography column operations and calculate the elastic recovery rate of the bed compaction factor from its peak value to the baseline value; S72: Construct the resin aging and hardening index. The resin aging and hardening index is positively correlated with the micropore blockage coefficient and negatively correlated with the elastic recovery rate. S73: When the bed compaction factor increases and the resin aging and hardening index exceeds the aging threshold, irreversible lipid hardening and compaction are determined, and an alarm for filler scrapping and replacement is triggered.

9. The method for monitoring the column efficiency of macroporous resin for purifying porcine brain peptides according to claim 8, characterized in that, Establishing a synergistic adsorption kinetic hysteresis compensation mechanism between the micropore blockage coefficient and the tailing factor includes: S81: Calculate the covariance between the micro-channel blockage coefficient and the tailing factor. When a significant positive correlation trend is observed, determine the mass transfer-impeded aging stage. S82: Activate the fractionation endpoint drift compensation algorithm. The fractionation endpoint drift compensation algorithm calculates the delay time based on the micro-channel blockage coefficient. S83: Increase the delay time based on the original collection endpoint, and increase the UV absorbance threshold of the fractionation cut point according to the tailing factor.

10. A macroporous resin column efficiency monitoring system for porcine brain peptide purification, used to perform the macroporous resin column efficiency monitoring method for porcine brain peptide purification as described in any one of claims 1-9, characterized in that, The system includes: Upstream Steady-State and Drift Monitoring Module: This module regulates the state of the crude porcine brain extract entering the ultrafiltration membrane module, and collects transmembrane pressure difference, permeate flux, and lipid content of each batch of feedstock. It calculates the membrane fouling resistance coefficient and estimates the equivalent molecular weight cutoff drift value of the ultrafiltration membrane. When the equivalent molecular weight cutoff drift value of the ultrafiltration membrane exceeds the drift threshold, it performs linkage control and generates a feed load correction factor. Micropore decay modeling module: Calculates the cumulative lipid adsorption amount based on resin parameters and operating parameters, and calculates the micropore blockage coefficient based on the cumulative lipid adsorption amount; when the operating flow rate is stable, the micropore blockage coefficient is used to quantify the resin's hidden adsorption decay. Flow field and compaction diagnosis module: Collects column bed pressure drop and outlet flow velocity to calculate dynamic flow resistance coefficient, and compares the dynamic flow resistance coefficient with the reference flow resistance coefficient of fresh packing to obtain the bed compaction factor; when the bed compaction factor fluctuates abnormally, it determines wall flow or channel flow, and combines the micro-pore blockage coefficient to distinguish between physical compaction and lipid contamination blockage. Separation feature analysis and correction module: Collect target brain peptide chromatographic peak data to calculate real-time resolution and tailing factor, and construct a comprehensive column efficiency decline index by feed load correction factor, micropore blockage coefficient and bed compaction factor, and use the comprehensive column efficiency decline index to correct the real-time resolution. Multidimensional Judgment and Adaptive Decision Module: Compares the comprehensive column efficiency decline index with the failure threshold, and triggers the deep regeneration program, backwashing and resuspension and replacement of packing material according to the comparison results, as well as performing upstream ultrafiltration membrane module cleaning or switching to the standby ultrafiltration membrane module.

Images