A small molecule peptide royal jelly production control method and system

Abstract

The application relates to the technical field of small-molecule peptide royal jelly production control, in particular to a small-molecule peptide royal jelly production control method and system. The method comprises the following steps: applying a preset micro-pH disturbance to an enzymolysis reaction liquid to obtain a pH change curve and extract transient response performance data; acquiring at least one auxiliary response data, cross-verifying the transient response performance data and the auxiliary response data, and obtaining a corrected pH value; monitoring the generation rate of a protein hydrolysate, and obtaining a reverse pH value according to a pre-established enzyme pH-activity relationship model; calculating a pH measurement reliability index based on the transient response performance data, the corrected pH value and the reverse pH value; and adaptively adjusting according to the pH measurement reliability index. The method solves the problems of measurement deviation caused by pH probe aging in small-molecule peptide royal jelly production, and the problems of enzymolysis efficiency reduction, product quality impairment and production cost increase caused by the measurement deviation.

Description

Technical Field

[0001] This application relates to the field of small molecule peptide royal jelly production control technology, and more specifically, to a method and system for controlling the production of small molecule peptide royal jelly. Background Technology

[0002] In the industrial production of small molecule peptide royal jelly, enzymatic hydrolysis is the core process, and its efficiency and product quality directly determine the final product. To ensure stable operation of the enzymatic hydrolysis reaction, the production system is equipped with online monitoring and control equipment, among which the pH probe is a key sensor used to monitor and adjust the pH of the enzymatic hydrolysis reaction solution in real time to maintain the optimal activity environment required by the enzyme preparation. However, pH probes using glass electrodes will gradually age after long-term immersion in complex media, and the response characteristics of the sensitive membrane will progressively deteriorate, manifested as an increase in instantaneous response time and a change in the non-linear correspondence between the output signal and the actual value, resulting in continuous small deviations in online pH measurements.

[0003] Specifically, the pH glass electrode's response speed to hydrogen ion diffusion and potential establishment slows down after prolonged use in the enzymatic hydrolysis reaction solution. Consequently, when the production control system adjusts the pH, the pH probe reading lags behind the actual changes in the enzymatic hydrolysis reaction solution, failing to reflect instantaneous fluctuations in a timely manner; simultaneously, the overall output characteristics, such as electrode slope and asymmetric potential, also experience slight drift. Existing automatic pH calibration programs primarily compensate for static offsets and are insufficient to effectively address the aforementioned dynamic performance degradation. Even after automatic calibration, a persistent and imperceptible slight deviation may still exist between the readings used by the control system and the actual instantaneous pH in the reactor; setting a relatively wide alarm threshold to avoid false alarms makes it difficult for this deviation to trigger an alarm, thus preventing the control logic from accurately sensing and correcting the true pH environment during enzymatic hydrolysis.

[0004] This deviation has a profound impact on enzymatic hydrolysis efficiency. Enzymes are extremely sensitive to pH; even slight and persistent deviations can trigger reversible conformational changes in enzyme molecules, reducing their affinity for protein substrates in royal jelly or slowing the catalytic conversion rate of substrates to small peptides. This leads to a decrease in reaction rate and incomplete conversion, thereby reducing the yield of the target small peptides. Enzymatic hydrolysis deviating from optimal conditions can also easily produce unexpected byproducts or incompletely hydrolyzed large protein fragments, reducing product purity and potentially affecting bioactivity and stability, ultimately weakening the overall quality of the final small peptide royal jelly.

[0005] Under these uncertainties, to ensure consistent batch compliance, the system often adopts a conservative compensation strategy: uniformly increasing the amount of enzyme preparation and extending the enzymatic hydrolysis time based on standard process parameters. While this approach can improve the pass rate, it significantly increases material costs and reduces reactor turnover, compressing production line efficiency and impacting overall economic benefits. Furthermore, the "quantity over quality" and "time over quality" approaches mask fundamental problems at the process control level, resulting in hidden resource waste and limiting further optimization opportunities.

[0006] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0007] This application discloses a method for controlling the production of small molecule peptide royal jelly, which aims to solve the measurement deviation caused by the aging of pH probes in the production of small molecule peptide royal jelly, as well as the resulting problems such as decreased enzymatic hydrolysis efficiency, damaged product quality and increased production costs.

[0008] The technical solution of this application is as follows:

[0009] In a first aspect, this application discloses a method for controlling the production of small molecule peptide royal jelly, comprising:

[0010] During the enzymatic hydrolysis reaction, a preset small pH disturbance is applied to the enzymatic hydrolysis reaction solution, and the dynamic response of the pH sensor is monitored to obtain the pH change curve. Instantaneous response performance data is extracted based on the pH change curve.

[0011] Acquire auxiliary response data from at least one auxiliary sensor, cross-validate the instantaneous response performance data with the auxiliary response data to correct the output of the pH sensor and obtain the corrected pH value;

[0012] The generation rate of protein hydrolysis products in the enzymatic reaction solution is monitored online, and the actual pH environment of the current enzymatic reaction is inferred in reverse based on the pre-established enzyme pH-activity relationship model to obtain the inverse pH value.

[0013] The pH measurement reliability index is calculated based on instantaneous response performance data, corrected pH values, and inversely derived pH values.

[0014] Based on the pH measurement reliability index, the amount of enzyme preparation and the enzymatic hydrolysis reaction time are adaptively adjusted.

[0015] Furthermore, the formation rate of protein hydrolysis products in the enzymatic reaction solution is monitored online, and the actual pH environment of the current enzymatic reaction is inferred from a pre-established enzyme pH-activity relationship model to obtain the inversely derived pH value, including:

[0016] The formation rate of protein hydrolysis products in the enzymatic reaction solution was monitored online, and the change curve of the formation rate was obtained.

[0017] Based on the correspondence between pH change curves and production rate change curves, the instantaneous response slope of the enzyme to small pH perturbations was calculated.

[0018] Based on the instantaneous response slope, the local shape of the enzyme pH-activity relationship model is dynamically corrected to obtain the corrected enzyme pH-activity relationship model.

[0019] Using the modified enzyme pH-activity relationship model, the actual pH environment of the current enzymatic hydrolysis reaction is inferred from the real-time monitored generation rate, and the inversely derived pH value is obtained.

[0020] Furthermore, based on the correspondence between the pH change curve and the production rate change curve, the instantaneous response slope of the enzyme to small pH perturbations is calculated, including:

[0021] Simultaneously monitor changes in at least one non-pH environmental factor, including temperature, substrate concentration, and / or inhibitor concentration;

[0022] By combining the pH change curve with the production rate change curve with data on changes in non-pH environmental factors, the instantaneous response slope of the enzyme to small pH perturbations is calculated, and the activity response caused by pH changes is distinguished from the activity response caused by non-pH environmental factors to obtain the distinguished instantaneous response slope.

[0023] The local shape of the enzyme pH-activity relationship model is dynamically corrected based on the slope of the differentiated instantaneous response.

[0024] Furthermore, based on the differentiated instantaneous response slope, the local shape of the enzyme pH-activity relationship model is dynamically corrected, including:

[0025] The slope of the differentiated instantaneous response is continuously tracked within a preset time window to determine whether it meets the judgment conditions of continuous and consistent deviation and deviation magnitude exceeding a preset deviation threshold.

[0026] When the judgment condition is met, a local correction of the enzyme pH-activity relationship model at the pH point and its neighborhood corresponding to the small pH perturbation is triggered: based on the difference between the differentiated instantaneous response slope and the local slope of the enzyme pH-activity relationship model, the enzyme pH-activity relationship model is updated with a limited amplitude, and the maximum amplitude of each local correction is limited to no more than the preset maximum correction amplitude.

[0027] After completing the local correction, a consistency constraint check is performed. The consistency constraint check includes at least verifying the biological smoothness and continuity of the corrected enzyme pH-activity relationship model. If the condition is not met, the correction result is constrained or rolled back.

[0028] Furthermore, based on the pH measurement reliability index, the amount of enzyme preparation and the enzymatic hydrolysis reaction time are adaptively adjusted, including:

[0029] The pH measurement reliability index was compared with preset high, medium, and low thresholds.

[0030] When the pH measurement reliability index is not lower than the high threshold, the enzymatic hydrolysis is terminated with the shortest termination time according to the preset control target.

[0031] When the pH measurement reliability index is between the high threshold and the medium threshold, the amount of enzyme preparation added and the enzymatic hydrolysis time are increased within the preset limit, and a prompt message is generated at the same time.

[0032] When the pH measurement reliability index is between the medium and low thresholds, inspection information is generated to prompt the pH sensor and auxiliary sensors to be checked.

[0033] When the pH measurement reliability index falls below a low threshold, an alarm is triggered and the system switches to a preset control mode or suspends production, generating an alarm message to remind users to perform offline calibration or replace the sensor.

[0034] Furthermore, continuously tracking the differentiated instantaneous response slope within a preset time window also includes:

[0035] The short-term fluctuation range of the instantaneous response slope after differentiation within a preset time window and the long-term drift rate across multiple preset time windows are continuously calculated.

[0036] Based on short-term fluctuation amplitude and long-term drift rate, the length of the preset time window and the preset deviation threshold used to judge trend deviation are dynamically adjusted.

[0037] Furthermore, before triggering the local correction of the enzyme pH-activity relationship model at the pH point and its neighborhood corresponding to a small pH perturbation when the judgment condition is met, the following steps are also included:

[0038] Continuously monitor the substrate concentration, product concentration, and enzyme activity decay indicators of the enzymatic hydrolysis reaction solution;

[0039] Based on substrate concentration, product concentration, and enzyme activity decay indicators, determine whether the current enzymatic hydrolysis reaction is in a stable phase.

[0040] If it is determined that the enzyme is in an unstable phase, the local modification of the enzyme pH-activity relationship model is suspended.

[0041] If the enzyme is determined to be in a stable phase and the criteria are met, then a local correction of the enzyme pH-activity relationship model for the pH point and its neighborhood corresponding to a small pH perturbation is initiated.

[0042] Furthermore, the substrate concentration, product concentration, and enzyme activity attenuation indicators of the enzymatic hydrolysis reaction solution are continuously monitored, including:

[0043] During the enzymatic hydrolysis process, a small amount of tracer is periodically injected into the enzymatic hydrolysis solution. The tracer does not participate in the enzymatic hydrolysis reaction but has a tracer detection signal.

[0044] Simultaneous monitoring of concentration signals of multiple substrates, multiple products, and tracer detection signals in the enzymatic hydrolysis reaction solution;

[0045] The response drift and background interference of the monitoring signals of multiple substrate concentrations and multiple product concentrations are calibrated in real time based on the changes in the tracer detection signal.

[0046] Decoupling analysis was performed on calibrated concentration signals of multiple substrates and multiple product concentrations to obtain their respective concentration change curves;

[0047] The enzyme activity decay index was calculated based on the concentration change curve.

[0048] Furthermore, real-time calibration is performed on the response drift and background interference of monitoring multiple substrate concentration signals and multiple product concentration signals based on changes in the tracer detection signal, including:

[0049] Continuously monitor the viscosity of the enzymatic hydrolysis solution, the bubble content, and the concentration of the tracer's self-degradation products;

[0050] Based on viscosity, bubble content, and the concentration of tracer degradation products, identify the sources of deviation in the tracer detection signal;

[0051] The tracer detection signal is corrected based on the source of the deviation to obtain the corrected tracer detection signal;

[0052] Based on the corrected tracer detection signal, the response drift and background interference of multiple substrate concentration signals and multiple product concentration signals are calibrated in real time;

[0053] Before decoupling analysis, the calibrated concentration signals of multiple substrates and multiple product signals were preprocessed.

[0054] Multidimensional spectral analysis was performed on the preprocessed concentration signals of various substrates and products to identify anomalous signals that did not match the known substrate and product signals and marked them as unknown interference signals.

[0055] During decoupling analysis, unknown interference signals are treated as independent components and separated to obtain their respective concentration change curves, which are then used to calculate enzyme activity decay indicators.

[0056] Secondly, this application also discloses a small molecule peptide royal jelly production control system, comprising:

[0057] The perturbation application and response monitoring module is used to apply a preset small pH perturbation to the enzymatic hydrolysis reaction solution during the enzymatic hydrolysis reaction process, monitor the dynamic response of the pH sensor to obtain the pH change curve, and extract instantaneous response performance data based on the pH change curve.

[0058] An auxiliary sensor correction module is used to acquire auxiliary response data from at least one auxiliary sensor, cross-validate the instantaneous response performance data with the auxiliary response data, and correct the output of the pH sensor to obtain the corrected pH value.

[0059] The product rate monitoring and pH back-inference module is used to monitor the generation rate of protein hydrolysis products in the enzymatic reaction solution online, and to infer the actual pH environment of the current enzymatic reaction based on the pre-established enzyme pH-activity relationship model to obtain the back-inferred pH value.

[0060] The pH measurement reliability index calculation module is used to calculate the pH measurement reliability index based on instantaneous response performance data, corrected pH value, and inversely derived pH value.

[0061] An adaptive control module is used to adaptively adjust the amount of enzyme preparation and the enzymatic hydrolysis reaction time based on the pH measurement reliability index.

[0062] Beneficial effects:

[0063] This application discloses a method for controlling the production of small molecule peptide royal jelly. By applying minute pH disturbances during the enzymatic hydrolysis reaction and monitoring the dynamic response of the pH sensor, combined with cross-validation using auxiliary sensors, the output deviation of the pH sensor is effectively corrected. Simultaneously, by monitoring the formation rate of protein hydrolysis products online and inferring the true pH environment based on a dynamically corrected enzyme pH-activity relationship model, the accuracy of pH measurement is further improved. A pH measurement reliability index calculated based on multi-source data allows for a comprehensive assessment of the pH sensor's health status, enabling adaptive adjustment of enzyme dosage and hydrolysis reaction time. This method overcomes the measurement inaccuracies caused by pH probe aging in existing technologies, avoiding decreased hydrolysis efficiency, product quality damage, and increased production costs due to minute pH deviations. Through precise control, this application can significantly improve the yield and purity of small molecule peptide royal jelly, reduce enzyme consumption, and shorten reaction time, thereby improving production efficiency and economic benefits. Attached Figure Description

[0064] Figure 1 This is a flowchart illustrating a method for controlling the production of small molecule peptide royal jelly, as provided in this application.

[0065] Figure 2 A flowchart of a small molecule peptide royal jelly production control system provided in this application.

[0066] In the diagram: 1. Disturbance application and response monitoring module; 2. Auxiliary sensor correction module; 3. Product rate monitoring and pH back-calculation module; 4. pH measurement reliability index calculation module; 5. Adaptive control module. Detailed Implementation

[0067] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0068] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0069] Reference Figure 1 This application proposes a method for controlling the production of small molecule peptide royal jelly, comprising:

[0070] S1000: During the enzymatic hydrolysis reaction, a preset small pH disturbance is applied to the enzymatic hydrolysis reaction solution, and the dynamic response of the pH sensor is monitored to obtain the pH change curve. Based on the pH change curve, instantaneous response performance data is extracted.

[0071] S2000: Acquire auxiliary response data from at least one auxiliary sensor, cross-validate the instantaneous response performance data with the auxiliary response data to correct the output of the pH sensor and obtain the corrected pH value;

[0072] S3000: Monitors the generation rate of protein hydrolysis products in the enzymatic reaction solution online, and reverse-infers the actual pH environment of the current enzymatic reaction based on the pre-established enzyme pH-activity relationship model to obtain the reverse-inferred pH value.

[0073] S4000: Calculates the pH measurement reliability index based on instantaneous response performance data, corrected pH value, and inversely derived pH value;

[0074] S5000: Adaptively adjusts the amount of enzyme preparation and the enzymatic reaction time based on the pH measurement reliability index.

[0075] Specifically, in this embodiment:

[0076] Small molecule peptide royal jelly refers to a series of peptides with specific molecular weight ranges obtained by enzymatically hydrolyzing proteins in royal jelly. Its production process requires precise control of the pH environment.

[0077] Enzymatic hydrolysis reaction solution refers to a mixed liquid containing royal jelly proteins, enzyme preparations, and other reaction media, which is the site where the enzymatic hydrolysis reaction takes place.

[0078] A pH sensor is a device used to measure the acidity or alkalinity of an enzymatic reaction solution in real time. It typically employs a glass electrode or other electrochemical principles.

[0079] Minimal pH perturbations refer to the precise, minute changes in pH applied to the reaction solution during an enzymatic hydrolysis reaction. The purpose is to stimulate a dynamic response in the pH sensor and enzyme activity, enabling subsequent performance evaluation. These perturbations typically do not significantly affect the overall progress of the enzymatic hydrolysis reaction, but are sufficient to be detected by highly sensitive sensors.

[0080] A pH change curve is a record of the pH value output by a pH sensor changing over time after a small pH disturbance is applied, reflecting the sensor's dynamic response characteristics to disturbances.

[0081] Instantaneous response performance data is extracted from pH change curves and characterizes the speed and accuracy of pH sensor response to small pH disturbances, such as response time, response slope, and overshoot.

[0082] Auxiliary sensors refer to sensors other than pH sensors used to monitor other parameters of the enzymatic hydrolysis reaction solution, such as conductivity sensors, oxidation-reduction potential (ORP) sensors, and ion-selective electrodes (ISE). Their data can be used to cross-validate the output of the pH sensor.

[0083] Auxiliary response data refers to the measurement data acquired by auxiliary sensors during minor pH disturbances or at specific time points.

[0084] The corrected pH value refers to a more accurate pH value obtained by cross-validating the instantaneous response performance data of the pH sensor with the auxiliary response data of the auxiliary sensor, and then correcting the original output of the pH sensor.

[0085] Protein hydrolysis products refer to small molecule peptides, amino acids, and other substances generated by the decomposition of royal jelly proteins under the action of enzymes.

[0086] The generation rate refers to the amount of protein hydrolysis products generated per unit time, reflecting the speed at which the enzymatic hydrolysis reaction proceeds.

[0087] An enzyme pH-activity relationship model is a mathematical model that describes the activity changes of a specific enzyme preparation under different pH conditions. It is usually represented by a curve with an optimal pH range.

[0088] The reverse-engineered pH value refers to the actual pH value of the current enzymatic hydrolysis reaction solution, which is inferred from the real-time monitored rate of protein hydrolysis product formation and a pre-established enzyme pH-activity relationship model.

[0089] The pH measurement reliability index is a quantitative indicator calculated by combining instantaneous response performance data, corrected pH values, and inversely derived pH values. It is used to assess the reliability of current pH measurements.

[0090] Specifically, during the enzymatic hydrolysis reaction, a preset minute pH disturbance is first applied to the reaction solution, and the pH sensor output is monitored simultaneously to obtain a pH change curve. Then, transient response performance data (e.g., response time and transient response slope) is extracted based on this curve. The disturbance can be periodic (by injecting trace amounts of acid or alkali through a precision metering pump, causing the pH to fluctuate within ±0.05 pH units) or transient (by generating brief local pH changes through electrochemical means). Monitoring can be performed using a high-precision pH meter, recording data multiple times per second to form a time series.

[0091] To improve the accuracy of pH measurements, at least one auxiliary sensor is introduced to acquire auxiliary response data (e.g., a conductivity sensor; alternatively, redox potential or visible-near-infrared spectral signals). The instantaneous response performance data of the pH sensor is cross-compared with the auxiliary response data to detect anomalies such as asynchrony or inconsistent amplitudes. This allows for linear or nonlinear correction of the pH sensor output, yielding a corrected pH value.

[0092] Simultaneously, the formation rate of protein hydrolysis products in the enzymatic reaction solution is monitored online (e.g., by using a near-infrared spectroscopy (NIR) analyzer to track characteristic absorption peaks in real time, or by using online chromatographic analysis to periodically sample and calculate the formation rate). Combined with a pre-established enzyme pH-activity relationship model (by experimentally measuring enzyme activity at different pH levels and fitting the curve), the actual pH environment is inferred in reverse, yielding the inversely calculated pH value (e.g., if the model shows the highest activity at pH 7.0 but the measured formation rate is lower, it suggests that the actual pH may deviate from 7.0).

[0093] After obtaining instantaneous response performance data, corrected pH values, and inversely derived pH values, a pH measurement reliability index is calculated. This index integrates sensor dynamics, cross-sensor consistency, and enzymatic feedback agreement. The calculation method can employ weighted averaging, fuzzy logic, or machine learning, outputting a value from 0 to 100 to measure the reliability of the current pH measurement.

[0094] Adaptive control is implemented based on the pH measurement reliability index: when the index is high, the amount of enzyme preparation and the enzymatic hydrolysis time are precisely set according to the preset optimization target (such as the highest yield or the shortest time); when the index is low, a conservative strategy is adopted (increasing the amount of enzyme preparation or extending the enzymatic hydrolysis time), and an alarm is issued to prompt equipment inspection or calibration.

[0095] This application obtains instantaneous response performance data in real time by applying a preset small pH perturbation during the enzymatic hydrolysis reaction and monitoring the dynamic response of the pH sensor. This makes up for the shortcomings of relying solely on static calibration and can capture changes such as the extension of response time and the reduction of response slope caused by sensor aging.

[0096] Meanwhile, by introducing at least one auxiliary sensor and cross-validating it with the pH sensor data, the pH output is corrected to obtain a corrected pH value that is closer to the true value, which significantly improves the measurement accuracy and robustness and avoids hidden biases caused by the aging or drift of a single sensor.

[0097] Furthermore, online monitoring of the protein hydrolysis product generation rate and reverse inference using an enzyme pH-activity relationship model provide a biological reference independent of physical sensors (e.g., when the expected generation rate corresponding to the optimal pH 7.0 is not reached, the actual pH deviation can be judged), establishing a "biological gold standard" for pH sensing.

[0098] Finally, based on the instantaneous response performance data, the corrected pH value, and the inversely derived pH value, the pH measurement reliability index is calculated, and the amount of enzyme preparation and the enzymatic hydrolysis time are adaptively adjusted accordingly: when the index is high, precise control is achieved to reduce enzyme waste; when the index is low, a conservative strategy is adopted and intervention is suggested to avoid quality or safety risks caused by inaccurate measurement.

[0099] In summary, this application achieves real-time evaluation of pH sensor performance and precise pH correction through minute pH perturbations, cross-validation with auxiliary sensors, back-calculation using an enzyme pH-activity relationship model, and calculation of the pH measurement reliability index. Based on this, it adaptively optimizes enzyme dosage and hydrolysis time. Compared to schemes relying solely on a single pH sensor and lacking dynamic evaluation, this method more accurately senses the real pH environment, significantly improving the stability, economy, and intelligence of small molecule peptide royal jelly production.

[0100] In another embodiment of this application, S3000 is further proposed to include:

[0101] S3100: Online monitoring of the generation rate of protein hydrolysis products in the enzymatic reaction solution, and acquisition of generation rate change curves;

[0102] S3200: Based on the correspondence between the pH change curve and the production rate change curve, calculate the instantaneous response slope of the enzyme to small pH perturbations.

[0103] S3300: Based on the instantaneous response slope, the local shape of the enzyme pH-activity relationship model is dynamically corrected to obtain the corrected enzyme pH-activity relationship model;

[0104] S3400: Using the modified enzyme pH-activity relationship model, based on the real-time monitored generation rate, the actual pH environment of the current enzymatic hydrolysis reaction is inferred in reverse, and the inferred pH value is obtained.

[0105] Specifically, the formation rate of protein hydrolysis products in the enzymatic reaction solution can be monitored online using various online analytical techniques: for example, using near-infrared spectroscopy (NIR), Raman spectroscopy, fluorescence spectroscopy, or high-performance liquid chromatography (HPLC) to obtain the concentration data of small molecule peptides, amino acids, etc. in real time, and calculate the formation rate accordingly; the trajectory over time constitutes the formation rate change curve.

[0106] Furthermore, based on the correspondence between the pH change curve and the production rate change curve, the instantaneous response slope of the enzyme to minor pH perturbations is calculated. Specifically, after applying a minor pH perturbation, a pH sensor records the pH change curve, and simultaneously, the production rate changes accordingly, forming a production rate change curve. Within a short time window following the perturbation, the synchronous changes of the two curves are analyzed to obtain the instantaneous response slope reflecting the enzyme's sensitivity to minor pH fluctuations under current conditions.

[0107] The local shape of the enzyme pH-activity relationship model is dynamically corrected based on the instantaneous response slope. The pre-established enzyme pH-activity relationship model is based on ideal or average conditions, and local drift or deformation may occur during actual operation. The real-time instantaneous response slope is compared with the theoretical slope of the model at the corresponding pH point: if a significant deviation exists, the model parameters or the local curve shape are adjusted according to the direction and magnitude of the deviation in that local region to obtain a corrected enzyme pH-activity relationship model that more closely approximates the actual enzymatic response.

[0108] Subsequently, using the modified enzyme pH-activity relationship model and combined with the real-time monitored production rate, the actual pH environment of the current enzymatic hydrolysis reaction was inferred in reverse, yielding the reverse-calculated pH value. Since the model has been adaptively updated according to the operating conditions, the inversely calculated pH value better reflects the true correspondence between the current enzyme activity state and pH.

[0109] The above technical solutions can significantly improve the accuracy of reverse inference from real pH environments. Dynamic correction ensures that the enzyme pH-activity relationship model continuously matches the actual working conditions, avoiding inference errors introduced by discrepancies between the model and reality. The more accurate reverse-inferred pH value provides a reliable basis for the adaptive adjustment of subsequent enzyme dosage and enzymatic hydrolysis reaction time, thereby enhancing adaptability and robustness to changes in enzyme activity and environmental fluctuations, and promoting precise control, yield improvement, and quality optimization in the production process of small molecule peptide royal jelly.

[0110] In some preferred embodiments, the following specific example illustrates the situation:

[0111] Assuming that in the initial stage of the enzymatic hydrolysis reaction of small molecule peptide royal jelly, the pre-established enzyme pH-activity relationship model shows that the enzyme activity has a specific slope to pH changes near pH 7.0. However, as the reaction proceeds, due to substrate consumption and product accumulation, the enzyme conformation may undergo slight changes, leading to a slight decrease in the sensitivity of its actual activity to pH near pH 7.0, i.e., the actual instantaneous response slope is less than the slope predicted by the model.

[0112] At this point, the system applies a small pH perturbation to the enzymatic hydrolysis solution, for example, briefly raising the pH from 7.0 to 7.05 and then quickly restoring it. During this process, the pH sensor records the pH change curve from 7.0 to 7.05 and back to 7.0. Simultaneously, the system monitors the change in the formation rate of protein hydrolysis products using an online spectrometer and records the rate change curve. Based on these two curves, the system calculates the instantaneous response slope of the enzyme to this small perturbation near pH 7.0. If the calculated instantaneous response slope is consistently lower than the theoretical slope of the preset model at that pH point, it indicates that the preset model has a deviation in that local region.

[0113] Based on this deviation, the system will dynamically correct the local shape of the enzyme pH-activity relationship model in and around pH 7.0. For example, by adjusting model parameters, the slope in this region can be matched with the real-time calculated instantaneous response slope, but the maximum correction magnitude is limited to a preset maximum to ensure the biological rationality of the model. The corrected enzyme pH-activity relationship model will more accurately reflect the actual activity characteristics of the enzyme near pH 7.0. Subsequently, when the system monitors the formation rate of a certain protein hydrolysis product, it can use this corrected model to make back inferences, thereby obtaining a pH value closer to reality, thus providing a more reliable decision-making basis for subsequent enzyme preparation addition and reaction time adjustment.

[0114] In another embodiment of this application, S3200 specifically includes:

[0115] S3210: Simultaneously monitor changes in at least one non-pH environmental factor, including temperature, substrate concentration, and / or inhibitor concentration;

[0116] S3220: Combine the pH change curve and the production rate change curve with the change data of non-pH environmental factors to calculate the instantaneous response slope of the enzyme to small pH perturbations, and distinguish the activity response caused by pH change from the activity response caused by non-pH environmental factors to obtain the distinguished instantaneous response slope.

[0117] S3230: Dynamically correct the local shape of the enzyme pH-activity relationship model based on the slope of the differentiated instantaneous response.

[0118] Specifically, online monitoring of the formation rate of protein hydrolysis products in the enzymatic reaction solution can be achieved by using various online analytical techniques to obtain continuous data and generate a formation rate curve based on the trajectory over time. These techniques include near-infrared spectroscopy, Raman spectroscopy, fluorescence spectroscopy, and high-performance liquid chromatography, which can measure the concentration of small molecule peptides, amino acids, and other products in real time and calculate their formation rate.

[0119] Simultaneous monitoring of changes in at least one non-pH environmental factor refers to obtaining dynamic data on temperature, substrate concentration, and / or inhibitor concentration in real time using appropriate sensors or analytical methods while applying a small pH perturbation and acquiring the pH change curve. For example, temperature is measured in real time using a high-precision temperature sensor; substrate and product concentrations are monitored using online spectroscopic analysis (such as near-infrared spectroscopy, Raman spectroscopy) or chromatographic techniques; and inhibitor concentrations are detected using specific biosensors or chemical analysis methods. This simultaneous monitoring provides data support for subsequently distinguishing the effects of different factors on enzyme activity.

[0120] This method combines pH change curves and production rate change curves with data on changes in non-pH environmental factors to calculate the instantaneous response slope of the enzyme to minor pH perturbations. It distinguishes between the activity response caused by pH changes and those caused by non-pH environmental factors, obtaining the differentiated instantaneous response slope. This involves modeling the association between the three types of data using multivariate data analysis methods. For example, multiple linear regression, partial least squares, or principal component analysis can be used to quantify the individual contributions of pH changes, changes in non-pH environmental factors, and changes in production rate to enzyme activity. This effectively separates enzyme activity changes directly caused by pH perturbations from those caused by non-pH environmental factors, obtaining a more accurate instantaneous response slope reflecting the enzyme's response to pH perturbations.

[0121] In practical applications, dynamically correcting the local shape of the enzyme pH-activity relationship model based on the differentiated instantaneous response slope refers to updating the portion of the pre-established enzyme pH-activity relationship model related to the current pH perturbation point and its neighborhood using the instantaneous response slope that has eliminated non-pH interference. Since this slope only characterizes the influence of pH factors, the model correction is more accurate and can more realistically reflect the pH-activity characteristics of the enzyme under the current reaction conditions.

[0122] This application's solution overcomes the problem of inaccurate instantaneous response slope calculation caused by traditional methods' failure to fully consider multi-factor coupling by simultaneously monitoring and distinguishing non-pH environmental factors. Specifically, when a small pH perturbation is applied, changes in enzyme activity affect the formation rate of protein hydrolysis products; if temperature, substrate concentration, or inhibitor concentration fluctuates simultaneously, their effects will be superimposed on the pH effect. By jointly modeling and analyzing pH change curves and formation rate change curves, and utilizing multivariate statistical analysis or machine learning models, the contributions of various factors can be effectively decoupled, yielding enzyme activity changes caused only by pH perturbations and calculating a more accurate, differentiated instantaneous response slope. The local dynamic correction of the model based on this slope is built on more realistic data, thereby significantly improving the accuracy and reliability of the correction.

[0123] In some preferred embodiments, this application is implemented as follows:

[0124] In the enzymatic hydrolysis of royal jelly small peptides, a high-precision temperature sensor and an online near-infrared spectrometer are deployed in addition to a pH sensor. When a small pH perturbation is applied to the system, the pH sensor records the pH change curve, the temperature sensor records the temperature change curve, and the near-infrared spectrometer monitors the concentration changes of the substrate (e.g., protein) and product (e.g., small peptide) in real time, thereby calculating the generation rate change curve. To calculate the differentiated instantaneous response slope, these simultaneously acquired pH change curves, temperature change curves, and generation rate change curves can be input into a multivariate regression model. This model is pre-trained to identify the independent contributions of pH, temperature, and other potential factors (such as substrate consumption or product accumulation reflected by spectral data) to enzyme activity. For example, a linear model can be constructed: Generation rate change = a * pH change + b * temperature change + c * other factors change + error, where a: pH sensitivity coefficient, b: temperature sensitivity coefficient, and c: other factors sensitivity coefficient. This model allows the calculation of coefficient 'a', which represents the slope of the enzyme's instantaneous response to pH perturbations after excluding the influence of non-pH factors such as temperature. This differentiated instantaneous response slope is then used to dynamically correct the enzyme's pH-activity relationship model, ensuring the accuracy of model updates. For example, if the differentiated instantaneous response slope shows that the enzyme is more sensitive to pH changes at a certain pH point than the model predicts, the model curves for that pH point and its neighborhood are adjusted accordingly to make their slopes more consistent with the actual response. However, the correction magnitude is limited to a preset range to maintain the biological validity of the model.

[0125] In another embodiment of this application, S3230 is further proposed to include:

[0126] S3231: The instantaneous response slope after differentiation is continuously tracked within a preset time window to determine whether it meets the judgment conditions of continuous and consistent deviation and deviation magnitude exceeding the preset deviation threshold.

[0127] S3232: When the judgment condition is met, trigger a local correction of the enzyme pH-activity relationship model at the pH point and its neighborhood corresponding to the small pH perturbation: Based on the difference between the differentiated instantaneous response slope and the local slope of the enzyme pH-activity relationship model, update the enzyme pH-activity relationship model with limited amplitude, and limit the maximum amplitude of each local correction to no more than the preset maximum correction amplitude.

[0128] S3233: After completing the local correction, perform a consistency constraint check. The consistency constraint check includes at least verifying the biological smoothness and continuity of the corrected enzyme pH-activity relationship model. If it is not satisfied, the correction result is constrained or rolled back.

[0129] Specifically, continuously tracking the instantaneous response slope after differentiation within a preset time window means that the system continuously monitors the instantaneous response slope of the enzyme to minute pH disturbances and analyzes its changing trend within a preset time period. The preset time window is set based on the characteristics of the enzymatic reaction, the sensor response speed, and the system's requirements for real-time correction; for example, it can be several seconds to several minutes. Continuous tracking filters out instantaneous noise and short-term fluctuations, ensuring that subsequent corrections are based on true and continuous deviations.

[0130] The determination of whether a deviation is persistent, consistent in direction, and exceeds a preset deviation threshold refers to the system performing statistical and trend analysis on the differentiated instantaneous response slopes within a preset time window based on an algorithm. This analysis determines whether a stable, consistent deviation exists that exceeds the tolerance range. For example, the average and variance of the slope are calculated, or a trend test is performed to confirm a continuous increase or decrease, and whether the difference between this slope and the expected slope of the enzyme pH-activity relationship model exceeds a preset deviation threshold. This preset deviation threshold is determined based on experimental data or expert experience and is used to distinguish between normal fluctuations and true deviations requiring correction.

[0131] In practical applications, when the judgment conditions are met, triggering a local correction of the enzyme pH-activity relationship model at the pH point and its neighborhood corresponding to a small pH perturbation means that once a continuous and significant deviation is confirmed, the system adjusts the parameters or curve shape of the enzyme pH-activity relationship model relative to the current perturbation point and its vicinity. The adjustment is based on the difference between the differentiated instantaneous response slope and the local slope of the model. To avoid over-correction, the magnitude of each local correction is controlled, for example, by limiting the preset maximum correction magnitude to a few percent of the model parameters or a fixed small value.

[0132] Furthermore, after completing the local correction, a consistency constraint check is performed. This involves verifying the biological rationality of the corrected enzyme pH-activity relationship model, including at least checking its biological smoothness and continuity, such as continuity of the first / second derivative and absence of mutation thresholds. For example, it checks for abrupt jumps, unreasonable inflection points, or curve shapes that contradict known enzyme kinetic characteristics. If the constraints are not met, the correction results are constrained (by reducing the correction magnitude) or reverted to the uncorrected model to ensure model stability and biological validity.

[0133] This application's solution, through a combination of "continuous tracking + strict triggering," avoids erroneous corrections caused by instantaneous noise or short-term fluctuations. By comprehensively judging the persistence, directional consistency, and magnitude of deviations, the basis for correction becomes more reliable. Combined with amplitude-limited updates and consistency constraint checks, the correction process is stable, and the results conform to biological laws, enabling the enzyme pH-activity relationship model to more accurately reflect real enzyme activity and providing a solid foundation for subsequent pH measurement reliability index calculations and production control.

[0134] In some preferred embodiments, this application is implemented as follows:

[0135] Assume that during the enzymatic hydrolysis reaction, the system continuously monitors the transient response slope after differentiation. To determine whether the enzyme pH-activity relationship model needs to be modified, the system sets a preset time window of 30 seconds and a preset deviation threshold of 0.05 (e.g., the relative difference between the slope and the expected slope of the model). The system records the transient response slope once per second and analyzes this data within the 30-second time window.

[0136] Specifically, the system calculates the average value and trend of the instantaneous response slope over these 30 seconds. If it finds that the instantaneous response slope shows a downward trend (in a consistent direction) for 20 consecutive seconds (meeting the continuity requirement), and its average deviation reaches 0.06 (exceeding the preset deviation threshold of 0.05), then the judgment condition is met, triggering a local correction of the model.

[0137] At this point, the system calculates a correction based on the difference between the current instantaneous response slope after differentiation and the local slope of the enzyme pH-activity relationship model at that pH point. To prevent overcorrection, the system limits the maximum magnitude of each local correction to no more than 2% of the model parameters. For example, if the calculated correction would cause a 3% change in the model parameters, the system will limit it to within 2%.

[0138] After completing local corrections, the system immediately performs consistency constraint checks. For example, it calculates the second derivative of the corrected model curve to check its smoothness, ensuring there are no sharp inflection points. Simultaneously, it compares the corrected model with known enzyme kinetics to ensure it remains biologically sound. If the check finds unreasonable unevenness in a certain region of the corrected model, the system may constrain the correction result, such as reducing the correction magnitude in that region or reverting directly to the uncorrected model state, recording this anomaly for subsequent manual intervention or further optimization of the correction algorithm. In this way, the robustness and accuracy of the model correction are ensured.

[0139] In another embodiment of this application, S5000 specifically includes:

[0140] S5100: Compare the pH measurement reliability index with preset high, medium and low thresholds;

[0141] S5200: When the pH measurement reliability index is not lower than the high threshold, the enzymatic hydrolysis is terminated with the shortest termination time according to the preset control target.

[0142] S5300: When the pH measurement reliability index is between the high threshold and the medium threshold, the amount of enzyme preparation added and the enzymatic hydrolysis time are increased within the preset limit, and a prompt message is generated at the same time.

[0143] S5400: When the pH measurement reliability index is between the medium and low thresholds, generate inspection information prompting the pH sensor and auxiliary sensors to be checked.

[0144] S5500: When the pH measurement reliability index is lower than the low threshold, an alarm is triggered and the system switches to the preset control mode or suspends production, generating an alarm message to remind users to perform offline calibration or replace the sensor.

[0145] Specifically, the pH measurement reliability index is used to comprehensively evaluate the credibility of pH measurements. It reflects the consistency among three types of information: the instantaneous response performance data extracted from the pH sensor output, the corrected pH value obtained after cross-validation with auxiliary sensors, and the inversely derived pH value based on the enzyme pH-activity relationship model. This index is typically standardized to the 0-1 range, with higher values ​​indicating higher reliability. To achieve finer control, high, medium, and low thresholds are set (e.g., 0.9, 0.7, and 0.5 respectively), and can be iteratively optimized based on production experience and system performance.

[0146] The control strategy based on hierarchical thresholds is as follows:

[0147] (1) When the pH measurement reliability index is not lower than the high threshold, the measurement is deemed highly reliable. The system executes the shortest termination time (shortest termination time: the shortest reaction termination time obtained by optimization under the premise of meeting the target quality index / yield) to end the enzymatic hydrolysis reaction in order to maximize production efficiency.

[0148] (2) When the pH measurement reliability index is between the high threshold and the medium threshold, it is determined that there is a certain degree of uncertainty but it is still acceptable. The system appropriately increases the amount of enzyme preparation and extends the enzymatic hydrolysis time within the preset limit to make up for possible measurement deviations. At the same time, it generates prompt information to remind you to pay attention to the measurement status without immediate manual intervention.

[0149] (3) When the pH measurement reliability index is between the medium threshold and the low threshold: the reliability is judged to be low, the system generates inspection information and prompts to check the status of the pH sensor and auxiliary sensor (including drift, fault and calibration problems).

[0150] (4) When the pH measurement reliability index is lower than the low threshold: it is determined to be seriously unreliable. The system immediately triggers an alarm and switches to the preset control mode or suspends production (e.g., open-loop control based on historical data or empirical parameters). At the same time, an alarm message is generated, strongly requesting offline calibration or replacement of the faulty sensor.

[0151] Through a graded response mechanism centered on the pH measurement reliability index, the system can directly transform the quantitative assessment of "measurement reliability" into differentiated adaptive control: avoiding overly conservative approaches when reliability is high, and avoiding blind trust and miscontrol when reliability is low, thereby balancing production efficiency, process stability, and product quality. On the one hand, the index integrates sensor dynamic performance, measurement results corrected by multiple sensors, and biological feedback inferences, comprehensively characterizing measurement reliability; on the other hand, the graded thresholds provide clear and executable triggering conditions and handling paths for the control strategy, ensuring that the enzymatic hydrolysis process can still operate in a steady state and achieve quality objectives even when uncertainty exists.

[0152] In another embodiment of this application, it is further proposed that the slope of the differentiated instantaneous response be continuously tracked within a preset time window, which also includes:

[0153] S3231-1: Continuously calculate the short-term fluctuation range of the instantaneous response slope after differentiation within a preset time window and the long-term drift rate across multiple preset time windows;

[0154] S3231-2: Based on short-term fluctuation amplitude and long-term drift rate, dynamically adjust the length of the preset time window and the preset deviation threshold used to judge trend deviation.

[0155] Specifically, the short-term fluctuation range of the "distinguished instantaneous response slope" refers to the statistical analysis (such as calculating standard deviation, variance, and mean absolute deviation) of the slope sequence within the current preset time window to quantify its instability and noise level in a short period of time. This is used to assess the reliability of the current data and avoid misjudgments caused by instantaneous noise. The "long-term drift rate across multiple preset time windows" refers to the trend analysis (such as linear regression, exponential smoothing, and moving average) of the slope over a longer time scale (spanning multiple consecutive preset time windows) to identify slow and continuous systematic change trends. This reflects the shift in sensor characteristics or enzyme activity over time and provides a basis for judging the long-term stability of the model.

[0156] In practical applications, the system dynamically adjusts the judgment parameters used to identify trend deviations based on the two types of quantification results mentioned above. Specifically, the length of the preset time window and the preset deviation threshold are adjusted: when short-term fluctuations are large, the preset time window is appropriately extended to smooth noise, and the preset deviation threshold is increased to reduce sensitivity to instantaneous fluctuations; when short-term fluctuations are small but long-term drift rates are significant, the preset time window is appropriately shortened and the preset deviation threshold is decreased to improve the response speed and sensitivity to true trend deviations. This adaptive adjustment matches the judgment conditions with the dynamic characteristics of the current reaction environment, thereby improving the accuracy and timeliness of model correction triggering.

[0157] Therefore, this application, by continuously calculating short-term fluctuation amplitude and long-term drift rate, can simultaneously characterize random noise and systematic changes, thereby distinguishing between instantaneous disturbances and true deviations. Furthermore, with adaptive adjustments to preset time windows and preset deviation thresholds, it avoids the lag or oversensitivity issues caused by fixed parameters: it effectively suppresses noise-triggered erroneous corrections when short-term fluctuations are large, and accelerates the response to true deviations when long-term drifts are significant. This mechanism significantly improves the intelligence and robustness of local corrections in the enzyme pH-activity relationship model.

[0158] In some preferred embodiments, the following specific example illustrates the situation:

[0159] During the enzymatic hydrolysis reaction, the system continuously monitors the instantaneous response slope after differentiation. To quantify short-term fluctuations, a moving standard deviation method is used: within a preset time window containing N data points, the standard deviation of these N slope values ​​is calculated as the short-term fluctuation amplitude. To characterize the long-term drift rate, linear regression analysis is used: a straight line is fitted to the slope sequence spanning M preset time windows (M>1), and its slope is the long-term drift rate. The above two quantitative results are used together to drive subsequent adaptive parameter adjustment.

[0160] When the system detects that short-term fluctuations (such as standard deviation) are consistently higher than the initial set value over a period of time, it indicates that the current data is noisy. At this time, the system increases the preset time window length used to judge trend deviations from T seconds to T + ΔT seconds, and raises the preset deviation threshold from P% to P + ΔP%, in order to suppress false judgments caused by noise by increasing the sampling volume and relaxing the judgment conditions. ΔT refers to the time increment / step size used to adaptively adjust the "preset time window length T".

[0161] Conversely, when short-term fluctuations remain low, but the long-term drift rate (e.g., the absolute value of the linear regression slope) consistently exceeds the preset drift threshold, it indicates a slow and persistent deviation from the true value of the enzyme pH-activity relationship model. In this case, the system shortens the preset time window length from T seconds to T-ΔT seconds and correspondingly lowers the preset deviation threshold to improve sensitivity and response speed to slow drifts, ensuring the model can be corrected in a timely manner.

[0162] In another embodiment of this application, it is further proposed that, before S3232, the method further includes:

[0163] S3232-1: Continuously monitor the substrate concentration, product concentration, and enzyme activity decay indicators of the enzymatic hydrolysis reaction solution;

[0164] S3232-2: Determine whether the current enzymatic hydrolysis reaction is in a stable stage based on substrate concentration, product concentration, and enzyme activity decay indicators;

[0165] S3232-3: If it is determined that the enzyme is in an unstable phase, then the local modification of the enzyme pH-activity relationship model is suspended.

[0166] S3232-4: If it is determined to be in a stable phase and the judgment condition is met, then the local correction of the enzyme pH-activity relationship model at the pH point and its neighborhood corresponding to the small pH perturbation is initiated.

[0167] Specifically, continuous monitoring of substrate concentration, product concentration, and enzyme activity decay indices in the enzymatic hydrolysis reaction solution aims to obtain key parameters characterizing the progress and state of the enzymatic hydrolysis reaction. Substrate and product concentrations directly reflect the degree of reaction progress and catalytic efficiency; enzyme activity decay indices quantify the degree of enzyme inactivation during the reaction (such as enzyme half-life, activity loss rate at a specific time point, or indirect assessment based on tracer methods). These indicators can be obtained in real time through online spectral analysis, chromatographic analysis, or electrochemical sensors.

[0168] The determination of whether the enzymatic hydrolysis reaction is in a stable phase is based on substrate concentration, product concentration, and enzyme activity decay indicators. Macroscopic stability is assessed by analyzing the trends and fluctuations of these indicators within a preset time window. A stable phase is defined as one where the rates of change in substrate and product concentrations are essentially constant, and the enzyme activity decay indicators change gradually. Drastic fluctuations, rapid increases, or decreases are considered instability. The determination can be made using a set of preset thresholds or through dynamic evaluation based on a machine learning model.

[0169] In practical applications, if the enzyme is determined to be in an unstable phase, local modifications to the enzyme pH-activity relationship model are paused to avoid introducing inaccurate corrections when the state is unclear or fluctuates significantly. Conversely, if the enzyme is determined to be in a stable phase and meets the criteria that "the slope of the differentiated instantaneous response satisfies a persistent and consistent deviation with a deviation magnitude exceeding a preset deviation threshold," then local modifications to the enzyme pH-activity relationship model at the pH point and its neighborhood corresponding to minor pH perturbations are initiated. This strategy ensures that modifications are only implemented under stable reaction conditions and with high data reliability, thereby improving the accuracy of the modifications and the robustness of the system.

[0170] This application's approach avoids misjudging changes in apparent activity caused by unstable factors (such as substrate depletion, product inhibition, and rapid enzyme inactivation) as model bias by introducing a stable phase determination before triggering local corrections. Even if the "distinguished instantaneous response slope" meets the deviation determination criteria, the system pauses corrections as long as the reaction is in an unstable phase; corrections are only performed when the reaction is in a stable phase and a deviation that truly requires correction exists. This ensures both the effectiveness and accuracy of the corrections and enhances the reliability and controllability of the enzymatic hydrolysis process of small molecule peptides in royal jelly.

[0171] In some preferred embodiments:

[0172] In another embodiment of this application, S3232-1 specifically includes:

[0173] S3232-11: During the enzymatic hydrolysis reaction, a small amount of tracer is periodically injected into the enzymatic hydrolysis reaction solution. The tracer does not participate in the enzymatic hydrolysis reaction but has a tracer detection signal.

[0174] S3232-12: Simultaneous monitoring of concentration signals of multiple substrates, multiple products, and tracer detection signals in the enzymatic hydrolysis reaction solution;

[0175] S3232-13: Real-time calibration of the response drift and background interference of monitoring multiple substrate concentration signals and multiple product concentration signals based on changes in tracer detection signals;

[0176] S3232-14: Decouple the calibrated concentration signals of multiple substrates and multiple product concentration signals to obtain their respective concentration change curves;

[0177] S3232-15: Calculate the enzyme activity decay index based on the concentration change curve.

[0178] Specifically, a tracer is a substance that does not participate in any chemical reaction in the enzymatic reaction system, does not affect enzyme activity, and is easily detectable. Its role is to provide an independent, traceable signal to characterize non-specific changes unrelated to the target component, such as sensor drift, changes in background absorption or scattering, and changes in the viscosity or turbidity of the reaction solution. Examples include inert dyes, fluorescent substances, or stable isotope labels with specific spectral characteristics. Periodically injecting small amounts of tracer aims to maintain an effective signal throughout the monitoring period while avoiding significant impact on the system. To ensure time alignment and establish accurate correspondence, the system simultaneously acquires multiple substrate concentration signals, multiple product concentration signals, and tracer detection signals. The equipment used can be a multi-wavelength spectrometer, a chromatography-mass spectrometry system, or an integrated biosensor array.

[0179] In practical applications, real-time calibration of the substrate and product concentration signals based on changes in the tracer detection signal to address response drift and background interference refers to using the tracer detection signal as a quantitative reference for non-specific factors to compensate for measurement deviations caused by sensor contamination, optical path changes, or sample matrix effects. By establishing a relationship model between the tracer signal and these non-specific factors, linear or non-linear corrections are applied to the substrate and product signals, thereby eliminating response drift and background interference and ensuring the accuracy of concentration measurements.

[0180] Furthermore, decoupling analysis of calibrated multi-substrate concentration signals and multi-product concentration signals refers to separating the independent contributions of different components from the mixed signal. In enzymatic reaction solutions, different substrates and products often have similar detection characteristics or signal overlap. Principal component analysis, partial least squares, independent component analysis, or spectral decomposition-based algorithms can be used to decompose the mixed signal into independent concentration change curves, thereby accurately obtaining the true concentration of each component in complex systems.

[0181] Once an accurate concentration change curve is obtained, enzyme activity decay indicators can be calculated. For example, by monitoring the decrease in substrate consumption rate, the slowdown in product formation rate, or the accumulation rate of specific intermediates, the stability, degree of inactivation, and changes in catalytic efficiency of the enzyme during the reaction process can be assessed.

[0182] This application's solution overcomes the limitations of traditional methods that rely solely on direct sensor output through a chain mechanism of "tracer—real-time calibration—decoupling analysis": the tracer signal provides an independent and quantifiable reference for non-specific changes, real-time calibration eliminates measurement errors, and decoupling analysis can accurately reconstruct the concentration of each component even under signal overlap conditions. The resulting enzyme activity decay index is more reliable, providing a solid and trustworthy data foundation for determining the stable stage of enzymatic hydrolysis, optimizing local corrections to the enzyme pH-activity relationship model, and subsequent production control, thereby improving the production efficiency and product quality of small molecule peptide royal jelly.

[0183] In some preferred embodiments, the following specific example illustrates the situation:

[0184] Assuming that during the enzymatic hydrolysis of small-molecule peptide royal jelly, it is necessary to continuously monitor the concentration of protein substrate, oligopeptide product, and enzyme activity decay, a small amount of inert fluorescent dye can be periodically injected into the reaction solution as a tracer to overcome the influence of changes in turbidity of the reaction solution and sensor optical path contamination on photometer measurements. This tracer exhibits a stable fluorescence signal at a specific wavelength and does not participate in the enzymatic hydrolysis reaction.

[0185] Specifically, a multi-channel spectrometer is used to simultaneously monitor the UV absorption signal of the protein substrate at 280 nm, the UV absorption signal of the oligopeptide product at 220 nm, and the fluorescence emission signal of the tracer at 520 nm. When the turbidity of the reaction solution increases or the optical path of the spectrometer is contaminated, the fluorescence signal of the tracer will weaken accordingly. The system calculates a calibration factor in real time based on the change in the tracer fluorescence signal and uses this factor to correct the UV absorption signals of the protein substrate and oligopeptide product to eliminate response drift and background interference caused by turbidity or contamination. The corrected substrate and product signals may still overlap; for example, some small peptides may also have weak absorption at 280 nm. In this case, multivariate statistical methods such as partial least squares (PLS) can be used for decoupling analysis. Through a pre-established calibration model, the corrected mixed spectral signal is decomposed into independent protein substrate concentration change curves and oligopeptide product concentration change curves. Finally, based on these accurate concentration change curves, the consumption rate of the protein substrate and the generation rate of the oligopeptide product can be calculated. For example, by comparing the decreasing trend of substrate consumption rates at different time points, enzyme activity decay indicators can be obtained. If a significant decrease in the substrate consumption rate is found over a period of time, it indicates that enzyme activity is declining. These accurate enzyme activity decay indicators will be used to determine whether the enzymatic reaction is in a stable phase, thereby deciding whether to initiate or pause local modifications to the enzyme pH-activity relationship model, ensuring precise control of the production process.

[0186] In another embodiment of this application, S3232-13 further includes:

[0187] S3232-131: Continuously monitor the viscosity of the enzymatic hydrolysis solution, the bubble content, and the concentration of the tracer's self-degradation products;

[0188] S3232-132: Identify the sources of deviation in the tracer detection signal based on viscosity, bubble content, and the concentration of the tracer's own degradation products;

[0189] S3232-133: Correct the tracer detection signal according to the source of deviation to obtain the corrected tracer detection signal;

[0190] S3232-134: Based on the corrected tracer detection signal, calibrate the response drift and background interference of multiple substrate concentration signals and multiple product concentration signals in real time;

[0191] S3232-135: Preprocess the calibrated concentration signals of multiple substrates and multiple product signals before decoupling analysis;

[0192] S3232-136: Perform multidimensional spectral analysis on preprocessed concentration signals of multiple substrates and multiple product signals to identify anomalous signals that do not match known substrate and product signals and mark them as unknown interference signals;

[0193] S3232-137: In decoupling analysis, unknown interference signals are treated as independent components and separated to obtain their respective concentration change curves, which are then used to calculate enzyme activity decay indicators.

[0194] Specifically, continuous monitoring of the viscosity, bubble content, and concentration of tracer degradation products in the enzymatic hydrolysis solution aims to identify environmental and material factors that may affect the accuracy of tracer detection signals. Viscosity changes affect the sensor's hydrodynamic response, bubble content leads to light scattering or ultrasonic signal attenuation, and the concentration of tracer degradation products directly reflects tracer stability. These factors can be acquired in real time using online viscometers, optical or ultrasonic bubble sensors, and specific spectrometers.

[0195] Identifying the sources of deviation in the tracer detection signal based on viscosity, bubble content, and the concentration of tracer degradation products involves analyzing the correlation between these monitored quantities and the tracer detection signal to determine the specific reasons for deviations from the true value. For example, a significant increase in viscosity often corresponds to a systematic drift in the tracer signal; an increase in bubble content can easily cause random fluctuations; and an increase in the concentration of tracer degradation products can cause signal intensity attenuation or the appearance of new interference peaks.

[0196] Furthermore, the tracer detection signal is corrected based on the source of the deviation to obtain a corrected tracer detection signal. The correction can be based on a pre-established calibration model (such as multiple linear regression or neural network model) for compensation, or adaptive filtering techniques can be used to suppress noise and drift, so that the corrected signal more accurately reflects the true concentration or state of the tracer.

[0197] Therefore, based on the corrected tracer detection signal, the response drift and background interference of multiple substrate concentration signals and multiple product concentration signals are calibrated in real time. That is, the reference for calibration is no longer the original tracer signal that may contain biases, but the reliable tracer signal after internal correction, thereby significantly improving the accuracy of substrate and product concentration signal calibration.

[0198] Prior to decoupling analysis, the calibrated concentration signals of various substrates and products are preprocessed. Preprocessing may include signal denoising, baseline correction, data smoothing and normalization to eliminate random noise and systematic errors, improve signal quality, and provide cleaner data input for subsequent decoupling.

[0199] Specifically, multidimensional spectral analysis is performed on preprocessed signals of multiple substrate and product concentrations to identify anomalous signals that do not match known substrate and product signals and mark them as unknown interference signals. Multidimensional spectral analysis can employ techniques such as two-dimensional correlation spectroscopy, principal component analysis (PCA), and independent component analysis (ICA) to decompose and identify patterns in complex mixed signals. These signals are then compared with standard spectra or characteristic patterns of known substrates and products to accurately detect anomalous signals that do not belong to the target components (potentially originating from unknown side reaction products, sensor contamination, or environmental interference).

[0200] Finally, during decoupling analysis, unknown interference signals were treated as independent components and separated to obtain their respective concentration change curves. These concentration change curves were then used to calculate enzyme activity decay indices. In other words, when using methods such as partial least squares (PLS) and multivariate curve resolution for decoupling, non-target signals were no longer simply treated as background noise. Instead, the identified unknown interference signals were incorporated into the model and separated separately to more accurately extract the true concentration change curves of the target substrate and product.

[0201] The proposed solution fundamentally improves the accuracy and reliability of concentration monitoring through "tracer detection signal deviation identification and correction + high-quality preprocessing + unknown interference identification and independent separation": On the one hand, continuous monitoring of viscosity, bubble content, and tracer degradation product concentration can quantify the physical and chemical factors affecting the tracer signal in a timely manner, and correct the tracer signal accordingly to ensure the accuracy of the calibration benchmark; on the other hand, the preprocessing of the calibration signal improves data quality, and multidimensional spectral analysis identifies and separates unknown interference signals, avoiding them from being misjudged as target components or simply treated as background, and significantly reducing decoupling errors.

[0202] Through the above technical solutions, the accuracy and anti-interference ability of substrate and product concentration monitoring are significantly enhanced. Because the inherent bias of the tracer detection signal is identified and corrected, and unknown interference signals are independently separated, the resulting substrate and product concentration change curves are more realistic and reliable. This directly improves the accuracy of enzyme activity decay index calculation, providing a more accurate decision-making basis for the adaptive adjustment of enzyme dosage and enzymatic reaction time. This promotes more refined and stable control of the production process, ultimately optimizing the production efficiency and product quality of small molecule peptide royal jelly.

[0203] In some preferred embodiments, the following specific example illustrates the situation:

[0204] Assume that during the enzymatic hydrolysis of small molecule peptide royal jelly, a near-infrared spectroscopy (NIR) sensor is used to monitor the concentrations of the substrate (protein) and product (small molecule peptide), and a fluorescent tracer is injected to calibrate the drift of the NIR signal and background interference.

[0205] First, to ensure the accuracy of the fluorescent tracer signal, the system continuously monitors the viscosity of the enzymatic reaction solution (e.g., via an online viscometer), the bubble content (e.g., via an ultrasonic sensor), and the concentration of the fluorescent tracer's degradation products (e.g., via UV-Vis spectroscopy at a specific wavelength). When a sudden increase in viscosity is detected, the system identifies this as enhanced fluorescence signal scattering due to increased reaction solution viscosity and compensates for this by adjusting the tracer's fluorescence intensity signal according to a pre-defined correction model. Similarly, if abnormal bubble content is detected, corresponding corrections are made. If the tracer's degradation product concentration exceeds a threshold, a tracer signal attenuation correction is triggered.

[0206] Secondly, these corrected tracer detection signals are used to calibrate in real time the response drift and background interference of the original substrate and product signals detected by the NIR sensor. For example, calibration is performed by establishing a linear or nonlinear relationship between the tracer signal and the NIR signal drift.

[0207] Furthermore, before performing decoupled analysis of substrate and product concentrations, the calibrated NIR signal is preprocessed, for example, by Savitzky-Golay smoothing to remove high-frequency noise and baseline correction to eliminate background shift.

[0208] Subsequently, multidimensional spectral analysis was performed on the preprocessed multi-wavelength NIR spectral data, using methods such as principal component analysis (PCA) combined with residual analysis. By comparing the data with standard spectra of known proteins and peptides, the system can identify anomalous spectral features that do not belong to proteins or peptides and label them as unknown interference signals. For example, if a non-target byproduct is generated during the reaction, it will exhibit unique absorption peaks in the NIR spectrum; these peaks are identified as unknown interference.

[0209] Finally, during decoupling analyses such as multivariate curve resolution (MCR), these labeled unknown interference signals are treated as independent components within the MCR model. This allows the MCR algorithm to more accurately separate proteins, small peptides, and unknown interferences from the mixed spectra, obtaining their respective true concentration change curves. These precise concentration change curves are then used to calculate enzyme activity decay indices, providing reliable data support for precise enzyme dosing and reaction time optimization.

[0210] Reference Figure 2 This application further proposes a small molecule peptide royal jelly production control system, comprising:

[0211] The disturbance application and response monitoring module 1 is used to apply a preset small pH disturbance to the enzymatic hydrolysis reaction solution during the enzymatic hydrolysis reaction process, monitor the dynamic response of the pH sensor to obtain the pH change curve, and extract instantaneous response performance data based on the pH change curve.

[0212] The auxiliary sensor correction module 2 is used to acquire auxiliary response data from at least one auxiliary sensor, cross-validate the instantaneous response performance data with the auxiliary response data, and correct the output of the pH sensor to obtain the corrected pH value.

[0213] Product rate monitoring and pH back-inference module 3 is used to monitor the generation rate of protein hydrolysis products in the enzymatic reaction solution online, and to back-infer the actual pH environment of the current enzymatic reaction based on the pre-established enzyme pH-activity relationship model to obtain the back-inferred pH value.

[0214] pH measurement reliability index calculation module 4 is used to calculate the pH measurement reliability index based on instantaneous response performance data, corrected pH value and reverse-calculated pH value.

[0215] The adaptive control module 5 is used to adaptively adjust the amount of enzyme preparation and the enzymatic reaction time based on the pH measurement reliability index.

[0216] The small molecule peptide royal jelly production control system proposed in this application aims to solve the problem of inaccurate pH measurement caused by the aging of pH sensors in traditional production, as well as the resulting dilemma of low enzymatic hydrolysis efficiency and increased production costs. This system integrates multiple functional modules to achieve precise sensing, real-time evaluation, and intelligent control of the pH environment for the enzymatic hydrolysis reaction.

[0217] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for controlling small molecule peptide royal jelly production, characterized by, include: S1000: During the enzymatic hydrolysis reaction, a preset small pH disturbance is applied to the enzymatic hydrolysis reaction solution, and the dynamic response of the pH sensor is monitored to obtain a pH change curve. Instantaneous response performance data is extracted based on the pH change curve. S2000: Acquire auxiliary response data from at least one auxiliary sensor, cross-validate the instantaneous response performance data with the auxiliary response data to correct the output of the pH sensor and obtain the corrected pH value; S3000: Monitors the generation rate of protein hydrolysis products in the enzymatic reaction solution online, and infers the actual pH environment of the current enzymatic reaction based on the pre-established enzyme pH-activity relationship model to obtain the reverse pH value. S4000: Calculate the pH measurement reliability index based on the instantaneous response performance data, the corrected pH value, and the inversely derived pH value. The calculation method adopts weighted average, fuzzy logic, or machine learning. S5000: Based on the pH measurement reliability index, adaptively adjust the amount of enzyme preparation and the enzymatic hydrolysis reaction time; The instantaneous response performance data is the instantaneous response slope; The S3000 includes: S3100: Online monitoring of the generation rate of protein hydrolysis products in the enzymatic reaction solution, and acquisition of generation rate change curves; S3200: Based on the correspondence between the pH change curve and the generation rate change curve, calculate the instantaneous response slope of the enzyme to small pH perturbations; S3300: Based on the instantaneous response slope, dynamically correct the local shape of the enzyme pH-activity relationship model to obtain the corrected enzyme pH-activity relationship model; S3400: Using the modified enzyme pH-activity relationship model, the actual pH environment of the current enzymatic hydrolysis reaction is inferred from the real-time monitored generation rate to obtain the inferred pH value.

2. The small molecule peptide royal jelly production control method according to claim 1, characterized by, The S3200 includes: S3210: Simultaneously monitor changes in at least one non-pH environmental factor, including temperature, substrate concentration, and / or inhibitor concentration; S3220: Combine the pH change curve and the generation rate change curve with the change data of non-pH environmental factors to calculate the instantaneous response slope of the enzyme to the small pH disturbance, and distinguish the activity response caused by pH change from the activity response caused by non-pH environmental factors to obtain the distinguished instantaneous response slope. S3230: Based on the differentiated instantaneous response slope, dynamically correct the local shape of the enzyme pH-activity relationship model.

3. The small molecule peptide royal jelly production control method according to claim 2, characterized by, S3230 includes: S3231: The slope of the differentiated instantaneous response is continuously tracked within a preset time window to determine whether it meets the judgment conditions of continuous and consistent deviation and deviation magnitude exceeding the preset deviation threshold. S3232: When the determination condition is met, a local correction of the enzyme pH-activity relationship model at the pH point and its neighborhood corresponding to the small pH perturbation is triggered: based on the difference between the differentiated instantaneous response slope and the local slope of the enzyme pH-activity relationship model, the enzyme pH-activity relationship model is updated with a limited amplitude, and the maximum amplitude of each local correction is limited to no more than the preset maximum correction amplitude. S3233: After completing the local correction, perform a consistency constraint check, which includes at least verifying the biological smoothness and continuity of the corrected enzyme pH-activity relationship model. If the condition is not met, constrain or roll back the correction result.

4. The small molecule peptide royal jelly production control method according to claim 1, characterized by, The S5000 includes: S5100: Compare the pH measurement reliability index with preset high, medium and low thresholds; S5200: When the pH measurement reliability index is not lower than the high threshold, the enzymatic hydrolysis is terminated with the shortest termination time according to the preset control target. S5300: When the pH measurement reliability index is between the high threshold and the medium threshold, increase the amount of enzyme preparation and extend the enzymatic hydrolysis time within the preset limit, and generate a prompt message at the same time. S5400: When the pH measurement reliability index is between the medium threshold and the low threshold, generate inspection information prompting the pH sensor and auxiliary sensor to be checked; S5500: When the pH measurement reliability index is lower than the low threshold, an alarm is triggered and the system switches to a preset control mode or suspends production, generating an alarm message to remind users to perform offline calibration or replace the sensor.

5. The small molecule peptide royal jelly production control method according to claim 3, characterized by, The method also includes continuously tracking the slope of the differentiated instantaneous response within a preset time window, and includes: S3231-1: Continuously calculate the short-term fluctuation range of the instantaneous response slope after differentiation within a preset time window and the long-term drift rate across multiple preset time windows; S3231-2: Based on the short-term fluctuation amplitude and the long-term drift rate, dynamically adjust the length of the preset time window and the preset deviation threshold used to determine the trend deviation.

6. The small molecule peptide royal jelly production control method according to claim 3, characterized by, Before S3232, it also includes: S3232-1: Continuously monitor the substrate concentration, product concentration, and enzyme activity decay indicators of the enzymatic hydrolysis reaction solution; S3232-2: Based on the substrate concentration, the product concentration, and the enzyme activity decay index, determine whether the current enzymatic hydrolysis reaction is in a stable stage; S3232-3: If it is determined that the enzyme is in an unstable stage, then the local modification of the enzyme pH-activity relationship model shall be suspended. S3232-4: If it is determined that the enzyme is in a stable phase and the determination condition is met, then the local correction of the enzyme pH-activity relationship model at the pH point and its neighborhood corresponding to the small pH perturbation is initiated.

7. The method for controlling the production of small molecule peptide royal jelly according to claim 6, characterized in that, S3232-1 includes: S3232-11: During the enzymatic hydrolysis reaction, a small amount of tracer is periodically injected into the enzymatic hydrolysis reaction solution, wherein the tracer does not participate in the enzymatic hydrolysis reaction and has a tracer detection signal; S3232-12: Simultaneously monitor the concentration signals of multiple substrates, multiple product concentrations, and the tracer detection signal in the enzymatic hydrolysis reaction solution; S3232-13: Real-time calibration of the response drift and background interference of monitoring multiple substrate concentration signals and multiple product concentration signals based on the changes in the tracer detection signal; S3232-14: Decouple the calibrated concentration signals of multiple substrates and multiple product concentration signals to obtain their respective concentration change curves; S3232-15: Calculate the enzyme activity decay index based on the concentration change curve.

8. The small molecule peptide royal jelly production control method according to claim 7, characterized by, S3232-13 includes: S3232-131: Continuously monitor the viscosity of the enzymatic hydrolysis solution, the bubble content, and the concentration of the tracer's self-degradation products; S3232-132: Identify the source of deviation in the tracer detection signal based on the viscosity, the bubble content, and the concentration of the tracer's own degradation products; S3232-133: Correct the tracer detection signal according to the source of the deviation to obtain the corrected tracer detection signal; S3232-134: Based on the corrected tracer detection signal, calibrate the response drift and background interference of multiple substrate concentration signals and multiple product concentration signals in real time; S3232-135: Preprocess the calibrated concentration signals of multiple substrates and multiple product signals before decoupling analysis; S3232-136: Perform multidimensional spectral analysis on preprocessed concentration signals of multiple substrates and multiple product signals to identify anomalous signals that do not match known substrate and product signals and mark them as unknown interference signals; S3232-137: During decoupling analysis, the unknown interference signal is treated as an independent component and separated to obtain its respective concentration change curve, and the concentration change curve is used to calculate the enzyme activity decay index.

9. A small molecule peptide royal jelly production control system, used to execute the small molecule peptide royal jelly production control method as described in any one of claims 1 to 8, characterized in that, include: The perturbation application and response monitoring module is used to apply a preset small pH perturbation to the enzymatic hydrolysis reaction solution during the enzymatic hydrolysis reaction process, monitor the dynamic response of the pH sensor to obtain the pH change curve, and extract instantaneous response performance data based on the pH change curve. An auxiliary sensor correction module is used to acquire auxiliary response data from at least one auxiliary sensor, cross-validate the instantaneous response performance data with the auxiliary response data, and correct the output of the pH sensor to obtain a corrected pH value. The product rate monitoring and pH back-inference module is used to monitor the generation rate of protein hydrolysis products in the enzymatic reaction solution online, and to infer the actual pH environment of the current enzymatic reaction based on the pre-established enzyme pH-activity relationship model to obtain the back-inferred pH value. The pH measurement reliability index calculation module is used to calculate the pH measurement reliability index based on the instantaneous response performance data, the corrected pH value, and the inversely derived pH value. The calculation method adopts weighted average, fuzzy logic, or machine learning. An adaptive control module is used to adaptively adjust the amount of enzyme preparation and the enzymatic hydrolysis reaction time based on the pH measurement reliability index. The instantaneous response performance data is the instantaneous response slope; The product rate monitoring and pH back-calculation module is also used for: The formation rate of protein hydrolysis products in the enzymatic reaction solution was monitored online, and the change curve of the formation rate was obtained. Based on the correspondence between the pH change curve and the generation rate change curve, the instantaneous response slope of the enzyme to small pH perturbations is calculated. Based on the instantaneous response slope, the local shape of the enzyme pH-activity relationship model is dynamically corrected to obtain the corrected enzyme pH-activity relationship model. Using the modified enzyme pH-activity relationship model, the actual pH environment of the current enzymatic hydrolysis reaction is inferred from the real-time monitored generation rate, and the inferred pH value is obtained.

Images