Coix seed melanin decomposition component extraction method based on membrane separation

Abstract

The invention relates to the technical field of natural product separation, and discloses a coix seed melanin decomposition component extraction method based on membrane separation. The method comprises the following steps: acquiring preprocessing parameters of a coix seed raw material and optical characteristic data of an initial suspension; dividing a light transmittance stable stage and a light transmittance jump stage according to the variation trend of the optical characteristic data along with the processing duration; screening a target separation level based on a two-stage suspension particle size distribution difference coefficient; under the target separation level, calculating a separation efficiency index by combining the transmembrane flux attenuation characteristic of the membrane component, the membrane pore size distribution and the operation pressure parameter; determining a cycle tolerance factor based on the residue component concentration and the membrane contamination rate; screening a target membrane component according to the circulating tolerance factor, and constructing a multi-stage membrane separation sequence in combination with penetrating fluid component data; and executing a gradient extraction process. The method realizes efficient extraction of melanin decomposition components of coix seeds, improves the separation efficiency and the cycle stability, and is suitable for separation and purification of active components of natural products.

Description

Technical Field

[0001] This invention relates to the field of natural product separation technology, specifically to a method for extracting melanin decomposition components from Coix seed based on membrane separation. Background Technology

[0002] Job's tears, a traditional food and medicine ingredient, contains melanin-decomposing components with significant application potential in pharmaceuticals and cosmetics. In recent years, with the increasing demand for the development of natural active ingredients, the efficient extraction of melanin-decomposing components from Job's tears has become a research hotspot. Currently, extraction methods for these components mainly include traditional processes such as solvent extraction, ultrasound-assisted extraction, and microwave extraction. Solvent extraction relies on large amounts of organic solvents, which not only easily leads to the inactivation of active ingredients but also presents challenges such as difficulty in subsequent separation and purification and solvent residue. While ultrasound and microwave-assisted extraction can improve the extraction rate, they often suffer from high energy consumption and equipment maintenance costs, and the extracted products contain high levels of impurities, making it difficult to meet the requirements for high-purity applications. Membrane separation technology, due to its advantages such as no phase change, low energy consumption, and simple operation, is increasingly being applied to the separation and purification of natural product components. However, the application of existing membrane separation technologies in the extraction of melanin decomposition components from Coix seed has significant limitations. Most methods do not consider the correlation between raw material pretreatment parameters and the optical characteristics of the suspension, and only use fixed membrane separation layers for extraction, resulting in insufficient separation targeting. The transmittance of the suspension changes dynamically over time during processing, and existing technologies lack a systematic analysis of this change pattern, making it impossible to adjust the separation strategy according to the stage characteristics of transmittance, which easily leads to the loss of target components or low separation efficiency. The selection of membrane modules often relies on empirical judgment, failing to scientifically screen them based on key parameters such as differences in suspension particle size distribution and transmembrane flux decay. This results in rapid membrane fouling rates and poor recyclability, increasing the cost of industrial applications. Furthermore, the construction of existing multi-stage membrane separation sequences lacks gradient design, making it difficult to achieve precise fractional extraction of melanin decomposition components of different molecular weights, thus hindering the efficient utilization of Coix seed resources. Therefore, developing an extraction method that can dynamically adjust the separation strategy and optimize membrane module selection based on raw material characteristics is crucial to improving the extraction efficiency of melanin decomposition components from Coix seed. Summary of the Invention

[0003] The purpose of this invention is to provide a method for extracting melanin decomposition components from Coix seed based on membrane separation, so as to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides a method for extracting melanin decomposition components from Coix seed based on membrane separation, the method comprising: Pretreatment parameters and optical characteristic data of the initial suspension of Coix seed raw material were obtained. Based on the trend of the optical characteristic data of the initial suspension with treatment time, a transmittance stabilization stage and a transmittance jump stage were defined. Target separation levels were screened based on the difference coefficient of particle size distribution in the suspension between the transmittance stabilization and jump stages. Under the target separation level, the separation efficiency index of each membrane module was calculated based on the transmembrane flux attenuation characteristics of the membrane module in the transmittance stabilization and jump stages, combined with membrane pore size distribution data and operating pressure parameters. The cycle tolerance factor of each membrane module was determined based on the residual component concentration data and membrane fouling rate of the membrane module in the transmittance jump stage, combined with the separation efficiency index. Target membrane modules were screened based on the cycle tolerance factor, and a multi-stage membrane separation sequence was constructed by combining the permeate component data corresponding to the target separation level in the transmittance jump stage. A gradient extraction process for the melanin decomposition components of Coix seed was executed based on the multi-stage membrane separation sequence.

[0005] Preferably, the preprocessing parameters include the average particle size and crushing temperature; the optical characteristic data include real-time transmittance and dynamic light scattering data; and the transmembrane flux decay characteristics include the initial flux decay rate and the steady-state flux maintenance duration.

[0006] Preferably, the method for obtaining the separation efficiency index includes: At any target separation level, membrane modules with cross-membrane flux records during the transmittance stabilization phase are designated as candidate membrane modules, and membrane modules with permeate composition data during the transmittance jump phase are designated as associated membrane modules. Any candidate membrane module is selected as the target candidate membrane module, and any associated membrane module is selected as the target associated membrane module. Based on the initial flux decay rate, steady-state flux maintenance duration, and membrane pore size distribution matching degree of the target candidate membrane module and the target associated membrane module, a flux synergy factor relative to the target associated membrane module is obtained. The pressure fluctuation similarity between the operating pressure parameters of the target candidate membrane module during the transmittance stabilization phase and the operating pressure parameters of the target associated membrane module during the transmittance jump phase is calculated to obtain the pressure characteristic factor corresponding to each pressure monitoring point. Using the flux synergy factor as a weight, the pressure characteristic factors corresponding to each pressure monitoring point are weighted and fused to obtain the membrane flux contribution of the target candidate membrane module relative to the target associated membrane module. The average of the membrane flux contribution of the target candidate membrane module relative to all associated membrane modules is used as the separation efficiency index of the target candidate membrane module.

[0007] Preferably, the method for obtaining the flux synergy factor includes: An arbitrary operating time point is selected as the baseline node; based on the parameter differences between the target candidate membrane module and the target associated membrane module in the dimensions of average particle size and crushing temperature, the membrane material compatibility coefficient is determined; the ratio of the operating pressure parameter of the target associated membrane module at the baseline node to the operating pressure parameter of the target candidate membrane module at the baseline node is used as the first synergy coefficient; the ratio of the membrane material compatibility coefficient to the steady-state flux maintenance time of the target associated membrane module at the baseline node is used as the second synergy coefficient; the product of the first synergy coefficient and the second synergy coefficient is calculated to obtain the flux synergy factor of the target candidate membrane module relative to the target associated membrane module at the baseline node.

[0008] Preferably, the method for obtaining the cycle tolerance factor includes: Based on the average concentration of residual components and the membrane fouling rate at each operating time point, the fouling accumulation index of the target candidate membrane module is calculated. The product of the separation efficiency index and the fouling accumulation index of the target candidate membrane module is normalized and used as the cycle tolerance factor of the target candidate membrane module.

[0009] Preferably, the method for obtaining the pollution accumulation index includes: The characteristic correlation value between the average concentration of residual components and the membrane fouling rate of the target candidate membrane module at each operating time point is calculated. The average of the characteristic correlation values ​​corresponding to all operating time points during the transmittance jump stage is used as the fouling accumulation index of the target candidate membrane module.

[0010] Preferably, the construction of the multi-stage membrane separation sequence includes: Candidate membrane modules with a cycle tolerance factor greater than a preset tolerance threshold are selected as the main separation units in the multi-stage membrane separation sequence. The fouling accumulation index of each associated membrane module is obtained, and the product of the flux synergy factor between the main separation unit and the associated membrane module and the fouling accumulation index of the associated membrane module is used as the coupling matching degree of the associated membrane module relative to the main separation unit. For any main separation unit, a membrane assembly sequence is constructed in descending order of the coupling matching degree of the associated membrane module relative to the main separation unit, and the coupling matching degree of the associated membrane modules at the same level in the sequence is equal. The membrane assembly sequences of all main separation units constitute the multi-stage membrane separation sequence.

[0011] Preferably, the gradient extraction process includes: In the multi-stage membrane separation sequence, the layer containing the first shared associated membrane module is designated as the primary separation layer; after the primary separation layer, the layer containing the associated membrane module with the highest coupling matching degree is designated as the secondary separation layer; the membrane module between the main separation unit and the primary separation layer is subjected to a first pressure gradient mode, the membrane module between the primary separation layer and the secondary separation layer is subjected to a second pressure gradient mode, and the membrane module between the secondary separation layer and the final membrane module is subjected to a third pressure gradient mode to perform transmembrane separation.

[0012] Preferably, the division of the transmittance stabilization stage and the transmittance jump stage includes: The real-time transmittance values ​​at each time point are obtained to form a transmittance sequence. The first-order difference value of the transmittance sequence is calculated, and the time points with positive first-order difference values ​​are selected as transmittance increase nodes. The transmittance increase nodes are arranged in ascending order according to their corresponding first-order difference values ​​to obtain a difference sequence. The second-order difference value of the difference sequence is calculated, and the transmittance increase node corresponding to the maximum value of the second-order difference value is set as the jump characteristic node. The time period before the jump characteristic node is the transmittance stable stage, and the time period after the jump characteristic node is the transmittance jump stage.

[0013] Preferably, the target separation hierarchy includes: The difference between the average particle size of the suspension in the stable transmittance stage and the average particle size of the suspension in the jump transmittance stage is calculated and normalized to obtain the particle size distribution difference coefficient. If the particle size distribution difference coefficient is greater than the preset difference threshold, the separation level corresponding to the processing stage is taken as the target separation level.

[0014] Compared with the prior art, the beneficial effects of the present invention are: This membrane separation-based method for extracting melanin decomposition components from Coix seed provides a systematic solution for the efficient extraction of natural active ingredients through multi-dimensional parameter analysis and dynamic optimization design. In the extraction process, by acquiring the pretreatment parameters of the Coix seed raw material and the optical characteristic data of the initial suspension, the physicochemical properties of the suspension can be accurately captured, laying the foundation for the subsequent separation stage division. Based on the trend of optical characteristic data changes with processing time, the method divides the process into a stable transmittance stage and a transmittance jump stage, clearly identifying the dynamic changes in the particle state in the suspension. This avoids the fuzzy judgment of the separation stage in traditional methods, making the separation operation more closely reflect the actual state of the material. Screening target separation levels based on the difference coefficient of suspension particle size distribution between the stable and abrupt transmittance stages allows for targeted focusing on particle ranges enriched with target components, reducing interference from non-target particles and making separation operations more targeted. Within the target separation level, the separation efficiency index is calculated by combining the transmembrane flux decay characteristics of the membrane module at different transmittance stages, membrane pore size distribution data, and operating pressure parameters. This enables a quantitative evaluation of the membrane module's separation performance, overcoming the limitations of traditional experience-based membrane module selection and ensuring a good match between the selected membrane module and separation requirements. Determining the cycle tolerance factor by analyzing the residual component concentration data and membrane fouling rate during the transmittance transition phase of the membrane module allows for a comprehensive assessment of the membrane module's long-term stability, reducing frequent replacements due to membrane fouling and extending the effective service life of the membrane module. Screening target membrane modules based on the cycle tolerance factor and constructing multi-stage membrane separation sequences enables gradient extraction of target components across different molecular weight ranges. This avoids the limitations of single-stage membrane separation for complex components and improves the purity of the separated target components. The entire methodology, through dynamic parameter analysis and hierarchical optimization design, enables the extraction process to adapt to different pretreatment states of Coix seed raw materials, reducing the stringent requirements for the uniformity of raw material pretreatment. Simultaneously, the construction of a multi-stage membrane separation sequence achieves stepwise enrichment of target components, reducing component loss during separation and making the extraction efficiency more aligned with actual production needs. Furthermore, by scientifically quantifying the performance parameters of membrane modules, this method improves the applicability of membrane separation technology in the extraction of natural products, providing a valuable approach for the separation and purification of similar active ingredients. Attached Figure Description

[0015] Figure 1 This is a time-series diagram of the membrane separation-based method for extracting melanin decomposition components from Coix seed according to the present invention. Figure 2 A flowchart of the method for obtaining the separation efficiency index; Figure 3 A flowchart of the method for obtaining cyclic tolerance factors; Figure 4 A flowchart illustrating the method for obtaining the pollution accumulation index. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Please see Figure 1 This invention provides a method for extracting melanin decomposition components from Coix seed based on membrane separation, the method comprising: By utilizing the transmittance variation characteristics of suspensions, the membrane separation layers and component selection can be optimized to achieve efficient gradient extraction of components. Pretreatment parameters and optical characteristic data of the initial suspension of Coix seed raw material were obtained. Based on the trend of optical characteristic data changes with processing time, a transmittance stabilization stage and a transmittance jump stage were defined. Target separation levels were screened based on the difference coefficient of suspension particle size distribution between the stabilization and jump stages. Within the target separation level, the separation efficiency index of each membrane module was calculated based on the transmembrane flux attenuation characteristics of the membrane module during the stabilization and jump stages, combined with membrane pore size distribution data and operating pressure parameters. The cycle tolerance factor of each membrane module was determined based on the residual component concentration data and membrane fouling rate during the transmittance jump stage, combined with the separation efficiency index. Target membrane modules were screened based on the cycle tolerance factor, and a multi-stage membrane separation sequence was constructed by combining the permeate component data corresponding to the target separation level during the transmittance jump stage. A gradient extraction process for Coix seed melanin decomposition components based on the multi-stage membrane separation sequence was executed.

[0018] Example 1: See Figure 2 This embodiment involves the data acquisition and separation efficiency index calculation process. Preprocessing parameters include the average particle size and crushing temperature. The average particle size is measured using a laser particle size analyzer on the raw material, and the crushing temperature is recorded in real-time using a temperature sensor during the raw material processing. Optical characteristic data includes real-time transmittance values ​​and dynamic light scattering data. The real-time transmittance value is continuously monitored using a UV-Vis spectrophotometer at a specific wavelength to track changes in the transmittance of the suspension during processing. Dynamic light scattering data is obtained by dynamically capturing the motion trajectory and particle size distribution of the suspended particles using a dynamic light scattering instrument. Transmembrane flux decay characteristics include the initial flux decay rate and the steady-state flux maintenance duration. The initial flux decay rate is calculated based on the ratio of the initial flux value of the membrane module to the flux decrease per unit time. The steady-state flux maintenance duration is determined by recording the duration after the membrane flux reaches a stable state.

[0019] At the target separation level, membrane modules with cross-membrane flux records during the transmittance stabilization phase are selected as candidate membrane modules, while those with permeate composition data during the transmittance jump phase are selected as associated membrane modules. One candidate membrane module is chosen as the target candidate membrane module, and one associated membrane module is chosen as the target associated membrane module. The pore size distribution matching degree between the target candidate membrane module and the target associated membrane module is calculated based on the difference in their standard deviations of pore size; a smaller difference indicates a higher matching degree. Based on the initial flux decay rate, steady-state flux maintenance time, and pore size distribution matching degree of the target candidate membrane module and the target associated membrane module, a flux synergy factor is calculated. The calculation process for the flux synergy factor involves multiplying the pore size distribution matching degree by the steady-state flux maintenance time by the reciprocal of the initial flux decay rate to obtain the final synergy factor value.

[0020] The similarity of pressure fluctuations between the operating pressure parameters of the target candidate membrane module during the transmittance stabilization phase and those of the target associated membrane module during the transmittance jump phase is calculated using the correlation coefficient of the pressure time series. Each pressure monitoring point corresponds to a pressure characteristic factor value. Using a flux synergy factor as the weight, the pressure characteristic factors corresponding to each pressure monitoring point are weighted and fused. The fusion method is to sum the products of the flux synergy factor and the pressure characteristic factor values. The result serves as the membrane flux contribution of the target candidate membrane module relative to the target associated membrane module. The membrane flux contribution values ​​of the target candidate membrane module relative to all associated membrane modules are calculated one by one and averaged. The average value is the separation efficiency index of the target candidate membrane module.

[0021] The calculation of the separation efficiency index involves the dynamic adjustment and optimization of multiple parameters. The initial flux decay rate of the target candidate membrane module reflects the performance change trend of the membrane module in the initial stage, while the steady-state flux maintenance time reflects the durability of the membrane module in the stable operation stage. Membrane pore size distribution matching degree is used to evaluate the compatibility between different membrane modules; a higher matching degree indicates a better synergistic effect when the membrane modules are used in combination. Pressure fluctuation similarity is used to measure whether the changing trends of operating pressure in different stages are consistent; a higher similarity indicates a more stable impact of pressure changes on membrane flux.

[0022] In the calculation of the flux synergy factor, the ratio of membrane pore size distribution matching degree to steady-state flux maintenance time reflects the synergistic performance of the membrane module under specific conditions, while the reciprocal of the initial flux decay rate is used to adjust the weight of the synergy factor. The weighted fusion process of pressure characteristic factors ensures a reasonable allocation of the contribution of different pressure monitoring points to membrane flux, and the final membrane flux contribution accurately reflects the overall efficiency of the membrane module in multi-stage separation.

[0023] The calculation process of averaging the membrane flux contribution of the target candidate membrane module relative to all associated membrane modules ensures the objectivity and representativeness of the separation efficiency index. As an indicator quantifying the overall efficiency of a membrane module in multi-stage separation, the separation efficiency index provides data support for subsequent membrane module screening and combination. The calculation process of this index fully considers the dynamic performance changes of the membrane module and the interaction of operating parameters, ensuring the stability and efficiency of the separation process.

[0024] Throughout the implementation process, data acquisition and calculation were based on actual measurements and dynamic monitoring, ensuring the accuracy and reliability of the parameters. Recording preprocessing parameters and monitoring optical characteristic data provided fundamental data for the subsequent screening and assembly of membrane modules. The extraction and analysis of transmembrane flux attenuation characteristics provided crucial evidence for membrane module performance evaluation. The calculation of flux synergy factors and the weighted fusion of pressure characteristic factors ensured the rationality and accuracy of membrane flux contribution. The final calculation of the separation efficiency index provided key parameters for constructing multi-stage membrane separation sequences and executing gradient extraction procedures.

[0025] This implementation method ensures separation efficiency and component purity by dynamically monitoring the optical properties of the suspension and membrane performance parameters. Accurate recording of pretreatment parameters and dynamic monitoring of optical characteristic data provide reliable data support for subsequent membrane module screening and assembly. Extraction and analysis of transmembrane flux attenuation characteristics provide crucial information for membrane module performance evaluation and optimization. The calculation of flux synergy factors and the weighted fusion of pressure characteristic factors ensure the reasonable allocation and accurate calculation of membrane flux contributions. The final determination of the separation efficiency index provides key parameters for the construction of multi-stage membrane separation sequences and the efficient execution of gradient extraction processes.

[0026] Example 2: See Figure 3 This embodiment relates to the process of obtaining flux synergy factor and cycle tolerance factor. An arbitrary operation time node is selected as the baseline node, determined based on uniform sampling along the processing time axis. The parameter difference between the target candidate membrane module and the target associated membrane module in the dimension of average fragmentation particle size is calculated as the ratio of the absolute difference between their average fragmentation particle sizes to the average fragmentation particle size. The parameter difference in the dimension of fragmentation temperature is calculated as the ratio of the absolute difference between their temperature values ​​to the average temperature. The membrane material compatibility coefficient is calculated through a combination of parameter differences, specifically expressed as:

[0027] in: Indicates the compatibility coefficient of membrane materials. The parameter representing the degree of difference in the dimension of the mean particle size of the crushed particles. This represents the degree of variability in parameters related to crushing temperature. This coefficient is used to quantify the adaptability of membrane modules under feed processing conditions; a higher value indicates better compatibility.

[0028] The ratio of the operating pressure parameter of the target associated membrane module at the baseline node to that of the target candidate membrane module at the baseline node is used as the first synergy coefficient. The operating pressure parameter is directly read from the pressure sensor data. The ratio of the membrane material compatibility coefficient to the steady-state flux maintenance time of the target associated membrane module at the baseline node is used as the second synergy coefficient. Multiplying the first and second synergy coefficients yields the flux synergy factor of the target candidate membrane module relative to the target associated membrane module at the baseline node. This factor reflects the coordinated performance of the membrane module at a specific node. The calculation process of the flux synergy factor comprehensively considers operating pressure, membrane material compatibility, and steady-state flux maintenance capability, and can accurately characterize the synergistic effect of the membrane module in the dynamic separation process.

[0029] The average concentration of residual components at each operating time point of the target candidate membrane module was calculated by averaging the concentrations of residual components in the suspension determined by liquid chromatography. The membrane fouling rate was derived by regression analysis based on the membrane flux decline rate and the cumulative residual value. The fouling accumulation index was calculated by summing and averaging the products of the average concentration of residual components and the membrane fouling rate at each operating time point of the target candidate membrane module. The separation efficiency index and the fouling accumulation index of the target candidate membrane module were multiplied, and the results were normalized by subtracting the minimum value and dividing by the range to obtain the cycle tolerance factor of the target candidate membrane module. The cycle tolerance factor is used to assess the tolerance of the membrane module in repeated use; a higher value indicates that the performance of the membrane module degrades more slowly during long-term operation.

[0030] In the calculation of the flux synergy factor, the introduction of the membrane material compatibility coefficient quantifies the differences in adaptability of different membrane modules under feed processing conditions. The calculation of the differences in average particle size and crushing temperature ensures the objectivity of parameter differences, avoiding bias caused by a single indicator. The first synergy coefficient reflects the pressure coordination of the membrane module at the baseline node through the ratio of operating pressure, while the second synergy coefficient reflects the long-term performance of the membrane module through the ratio of steady-state flux maintenance time to the membrane material compatibility coefficient. The product of these two coefficients serves as the flux synergy factor, comprehensively reflecting the overall coordination capability of the membrane module during dynamic separation.

[0031] The fouling accumulation index is calculated based on dynamic monitoring data of the average concentration of residual components and the membrane fouling rate. The average concentration of residual components is determined periodically by liquid chromatography to ensure the accuracy and representativeness of the data. The membrane fouling rate is derived from regression analysis of the membrane flux decline rate and the cumulative residual value, reflecting the fouling accumulation trend of the membrane module during operation. The fouling accumulation index, which is the product of the two and then averaged, can effectively characterize the fouling load of the membrane module under specific operating conditions.

[0032] The cycle tolerance factor is calculated by combining the separation efficiency index and the fouling accumulation index, obtained through normalization of their product. The separation efficiency index reflects the overall efficiency of the membrane module in multi-stage separation, while the fouling accumulation index characterizes the fouling load of the membrane module. The combination of these two factors allows the cycle tolerance factor to simultaneously reflect the separation performance and long-term operational tolerance of the membrane module. Normalization ensures the comparability of cycle tolerance factors among different membrane modules, providing an objective basis for subsequent membrane module selection and combination.

[0033] Throughout the implementation process, data acquisition and calculation were based on actual measurements and dynamic monitoring, ensuring the accuracy and reliability of the parameters. The selection of the baseline node was based on uniform sampling along the processing time axis, avoiding deviations caused by human factors. The calculation of the membrane material compatibility coefficient comprehensively considered the differences in the average particle size and crushing temperature, ensuring the comprehensiveness and objectivity of the coefficient. The calculation process of the flux synergy factor fully considered operating pressure, membrane material compatibility, and steady-state flux maintenance capability, accurately reflecting the dynamic coordination performance of the membrane module. The calculation of the fouling accumulation index was based on dynamic monitoring data of residual component concentration and membrane fouling rate, ensuring the accuracy and representativeness of the index. The calculation of the cycle tolerance factor combined the separation efficiency index and the fouling accumulation index, and normalization ensured the comparability and objectivity of the factors.

[0034] This implementation method ensures the accurate calculation of flux synergy factors and cycle tolerance factors by dynamically monitoring the operating parameters and performance changes of the membrane modules. The introduction of a membrane material compatibility coefficient quantifies the differences in adaptability of different membrane modules under feed processing conditions, providing an important basis for the calculation of the flux synergy factor. The calculation of the fouling accumulation index is based on dynamic monitoring data of residual component concentration and membrane fouling rate, ensuring the accuracy and representativeness of the index. The calculation of the cycle tolerance factor combines the separation efficiency index and the fouling accumulation index, and normalization ensures the comparability and objectivity of the factors. The entire implementation process fully considers the feasibility of actual operation and the reliability of data, ensuring the practicality and effectiveness of the method.

[0035] Example 3: See Figure 4This embodiment involves the specific calculation process of the pollution accumulation index and the construction process of the multi-stage membrane separation sequence. The average concentration of residual components in the target candidate membrane module at each operating time point is obtained by determining the concentration of residual components in the suspension using liquid chromatography. The arithmetic mean is calculated after sampling three times at each point. The membrane fouling rate is calculated based on the transmembrane flux change rate: the flux decrease of the target candidate membrane module per unit time is recorded, and the slope of its time-varying curve is fitted using linear regression. The characteristic correlation value is calculated as the product of the average residual component concentration and the membrane fouling rate. The characteristic correlation values ​​corresponding to all operating time points during the transmittance jump stage are summed and divided by the total number of nodes to calculate the pollution accumulation index. The formula for this index is as follows:

[0036] in: Indicates the pollution accumulation index. Indicates the first Average concentration of residue components at each time point (unit: mg / L). Indicates the first Membrane fouling rate at each time point (unit: L / (m²·h·min)). This represents the total number of operation time points during the light transmittance transition phase.

[0037] When screening multi-stage membrane separation sequences, candidate membrane modules with a cycle tolerance factor greater than a preset tolerance threshold are selected as the main separation unit. The preset tolerance threshold is set to a fixed value (e.g., 0.85) based on historical operating data, and this value does not change with experimental conditions. The fouling accumulation index of all associated membrane modules is obtained and calculated using the same formula. The coupling matching degree of the associated membrane module relative to the main separation unit is calculated according to a specific rule: the flux synergy factor between the main separation unit and the associated membrane module (using the calculation result from Example 2) is multiplied by the fouling accumulation index of the associated membrane module. This product reflects the synergistic antifouling ability of the two during the separation process.

[0038] For any master separation unit, associated membrane modules are arranged in descending order of coupling matching degree to construct a membrane assembly sequence. When different associated membrane modules have equal coupling matching degree, they are assigned to the same level in the sequence and exist side by side. The sequence arrangement principle is: the associated membrane module with the highest coupling matching degree is closest to the master separation unit, and the module with the second highest coupling matching degree is located at the next level. All master separation units independently construct their own membrane assembly sequences, and the sequences are physically connected through shared associated membrane modules, ultimately forming a multi-level membrane separation sequence with a tree-like topology.

[0039] During the determination of residual component concentration, the suspension sample was pretreated through a 0.45 μm filter membrane before being injected into the liquid chromatograph. The mobile phase was methanol-water (70:30), and the detection wavelength was set to 280 nm. The membrane fouling rate regression model was established as follows: Operating time was plotted on the x-axis, and the measured transmembrane flux was plotted on the y-axis. The least squares method was used to fit a linear equation, and the absolute value of the slope was taken as the fouling rate value. Data was collected at fixed intervals (e.g., every 5 minutes) during the operating time, with the total number of data points determined by the actual duration of the transmittance transition phase.

[0040] The calculation of coupling matching degree relies on the effective transfer of flux synergy factors. The flux synergy factors of the main separation unit and associated membrane modules need to be compared at the same operating time point. If the nodes do not overlap, the nearest neighbor interpolation method is used to align the data sequences. The coupling matching degree of associated membrane modules at the same level has equivalent separation function under the condition of equal numerical values, and they can be connected in parallel or used alternately in actual operation. The topological connection principle of the multi-stage membrane separation sequence is as follows: the main separation unit is located at the first level of the separation system, the associated membrane modules at the same level are in a parallel position on the separation path, and the secondary level modules extend sequentially.

[0041] During sequence construction, a mapping table needs to be established to record four parameters: main separation unit number, associated membrane module number, coupling matching degree value, and hierarchical position. The position numbering principle is as follows: the main separation unit is numbered 0 at each level, and the number increments by 1 for each additional level; modules at the same level share the position number but are assigned independent serial numbers. For example, in the sequence of main unit A, if the coupling matching degree of associated membrane modules B and C is both 0.92, B and C coexist at level 1, and are labeled as level 1-1 and level 1-2 respectively. All sequence data is stored in a matrix structure, with row vectors corresponding to main separation units and column vectors arranging the associated membrane module indices according to their hierarchical positions.

[0042] This implementation process emphasizes a dynamic data synchronization mechanism. The residual concentration and membrane fouling rate required for calculating the pollution accumulation index must be collected at the same time point to ensure time alignment. The reuse of flux synergy factors in multi-stage sequence construction requires verification of the timeliness of the original data; expired data automatically triggers a recalculation process. Sequence connection logic is implemented through programmatic configuration: when a new associated membrane module is added, its coupling matching degree with all main separation units is automatically calculated, and its inclusion in the sequence is determined based on a threshold.

[0043] A closed-loop verification mechanism was established throughout the implementation process. The pollution accumulation index calculation results were linked to the actual cleaning frequency of the membrane modules, and any data anomalies were traced back to the residue concentration detection process. The structural rationality of the multi-level sequence was verified through pressure gradient testing: when the transmembrane pressure difference distribution deviated from the preset range, the hierarchical order of the associated membrane modules was automatically adjusted. The screening results of the main separation unit and the energy consumption data of subsequent extraction processes formed a feedback loop, used to optimize the preset tolerance threshold settings.

[0044] At the operational level, attention must be paid to the consistency of reagent specifications. All methanol used in liquid chromatography is of chromatographic purity, and the water is ultrapure water (resistivity ≥18 MΩ·cm). The dynamic light scattering instrument is calibrated with standard latex particles before each use, and the particle size detection error is controlled within ±2 nm. The flux recording device uses a high-precision flow sensor (range 0-100 L / h, accuracy ±0.5% FS), and the pressure sensor range covers 0-1 MPa. The membrane fouling regression model must contain at least 10 consecutive data points, and the correlation coefficient must be higher than 0.95 to be considered a valid slope.

[0045] Example 4: This example details the execution of the gradient extraction process, using a set of actual membrane module configurations as an example. In a multi-stage membrane separation sequence, the layer containing the first shared associated membrane module is designated as the primary separation layer. This layer is determined by scanning the connection relationships of all associated membrane modules in the membrane assembly sequence and identifying the first membrane module jointly connected by two or more main separation units. For example, when the sequence of main separation unit M1 contains associated membrane modules A→B→C, and the sequence of main separation unit M2 is A→D→E, membrane module A is the first shared associated membrane module, and its layer is marked as the primary separation layer.

[0046] The layer containing the membrane module with the highest coupling matching degree after the primary separation layer is designated as the secondary separation layer. The coupling matching degree ranking method is as follows: extract the coupling matching degree values ​​of all associated membrane modules after the primary separation layer, and select the membrane module corresponding to the maximum value. For example, in the above example, in the sequence of primary separation unit M1, the coupling matching degree of membrane module B is 0.92, and that of membrane module C is 0.85; in the sequence of primary separation unit M2, the coupling matching degree of membrane module D is 0.89, and that of membrane module E is 0.81. Then, membrane module B (0.92) has the highest coupling matching degree, and its layer is designated as the secondary separation layer. The implementation parameters for the pressure gradient mode are shown in Table 1.

[0047] Table 1: The implementation parameters for the pressure gradient mode are as follows.

[0048] Pressure gradient mode Scope of action Pressure change mode Critical value setting (MPa) First Mode Main separation unit → Primary separation layer linearly increasing 0.8 Second Mode Primary separation layer → Secondary separation layer constant pressure 0.6 Third Mode Secondary separation layer → terminal component Decreasing stepwise (0.1 decrease each time). 0.3 The first pressure gradient mode is implemented between the main separation unit and the primary separation layer. During operation, the initial pressure is 0.1 MPa, increasing linearly at a rate of 0.05 MPa per minute until a critical value of 0.8 MPa is reached. Pressure sensors monitor the transmembrane pressure difference in real time, automatically pausing pressure increases when pressure fluctuations exceed ±5%, and resuming only after stabilization. For example, the pressure regulation from the main separation unit M1 to membrane module A starts at 0.1 MPa, uniformly increasing to 0.8 MPa over 14 minutes, with the actual pressure value recorded every 30 seconds during this period.

[0049] The second pressure gradient mode maintains a constant pressure between the primary and secondary separation layers. The system automatically reads the outlet pressure of the primary separation layer membrane module as a reference and stabilizes it at 0.6 MPa via a proportional control valve. When associated membrane modules exist in parallel (e.g., membrane modules B and D both serve as secondary separation layer modules), each branch is equipped with an independent pressure balancing valve to ensure that the pressure difference between parallel paths does not exceed 0.02 MPa. The constant pressure maintenance phase continues until the permeate accumulation in this layer reaches a set volume (e.g., 10 L), after which it transitions to the next mode.

[0050] The third pressure gradient mode performs a step-by-step pressure reduction from the secondary separation layer to the final membrane module. The initial pressure is set to the secondary separation layer operating pressure (0.6 MPa), and the pressure is reduced by 0.1 MPa after every 200 mL of permeate collection, until the final safe pressure of 0.3 MPa is reached. Permeate conductivity is monitored simultaneously during the pressure reduction process. If the rate of change in conductivity exceeds 5% / min, the pressure reduction is paused, and the process continues after the parameters stabilize. For example, the pressure from membrane module B to the final membrane module F starts at 0.6 MPa and is reduced to 0.3 MPa in three steps, with each reduction approximately 7 minutes apart.

[0051] The physical connections of the membrane modules follow a hierarchical topology. The main separation unit connects to the primary separation layer module via stainless steel flanges, with all connecting pipes having a uniform inner diameter of 25mm. A T-junction splitter is used between the primary and secondary separation layers, with each branch equipped with a flow meter (range 0-10L / min) and a pressure buffer tank (500mL). The outlets of the terminal modules converge into a common collection tank, which is equipped with a level sensor and a temperature probe. All connections use sanitary quick-connect clamps, and the sealing material is EPDM rubber.

[0052] The dynamic adjustment mechanism for operating parameters includes the following key points: In the first mode of pressure increase, if the real-time flux decrease rate of a membrane module exceeds 20% of the initial value, the system automatically inserts a 5-minute pressure holding phase. In the second mode of constant pressure, if the flow difference between parallel branches persists for more than 3 minutes exceeding 15%, a pressure fine-tuning command is triggered (adjustment range ±0.03 MPa). In the third mode, after each step-down pressure reduction, the system detects the change in the fouling accumulation index of the upstream membrane module; if the increase exceeds 0.1, the duration of the current pressure phase is extended by 10%.

[0053] The data acquisition scheme during implementation included: a pressure sensor recording pipeline pressure every 10 seconds (accuracy ±0.01 MPa), and a flow meter calculating the cumulative permeate flow rate every minute (resolution 0.1 L). Permeate samples were collected periodically at the outlet of each membrane module (15-minute intervals), and pH (±0.01 units) and turbidity (NTU units) were immediately measured. Transmembrane flux data were obtained by weighing the mass change of the collection tank using an electronic balance (accuracy ±0.1 g), with the sampling frequency synchronized with the pressure recording.

[0054] The anomaly handling procedure is designed as follows: If no permeate outflow is detected when the pressure rises to 0.5 MPa in the first mode, the system automatically switches to the standby main separation unit. In the constant pressure phase of the second mode, if pressure fluctuations exceed the set range of ±8% three consecutive times, the associated membrane module is considered clogged, and a backflushing procedure is initiated. In the third mode, if the flux recovers by more than 10% after two step-down pressure reductions, the process is immediately stopped, and the membrane module's sealing is checked. All abnormal events are recorded with their timestamps, parameter deviations, and handling measures, forming an operation log for subsequent analysis.

[0055] In this implementation example, the equipment status monitoring interface displays color-coded indicators for three pressure gradient zones in real time: the first mode is displayed as an orange gradient fill, the second mode as green horizontal stripes, and the third mode as blue vertical stripes. Operators can intuitively observe the current pressure range position and remaining duration. The historical data playback function supports viewing the pressure-flow curve overlay graph of each stage along a timeline, with the curve colors consistent with the pressure gradient mode indicator colors.

[0056] The maintenance procedure requires the following operations after each gradient extraction: disassemble the end-stage assembly to check the wear of the sealing rings, measure the clear water flux recovery rate of the primary separation layer membrane module, and record the maximum operating pressure tolerance value of the secondary separation layer module. All data, including but not limited to: total operation time, percentage of time for each pressure gradient stage, abnormal event code, and cumulative permeate production, are entered into the membrane module performance database. The database records serve as the original basis for updating the coupling matching degree calculation model.

[0057] Example 5: This example involves a precise division of transmittance change stages and a screening mechanism for target separation levels. Real-time transmittance values ​​at each time point are acquired through an optical signal receiver. This device records the transmittance data of the suspension every 5 seconds at a wavelength of 520nm, forming a time-transmittance sequence. During sequence processing, the transmittance values ​​of adjacent time points are differentially analyzed to generate a first-order difference sequence. All time points with positive first-order difference values ​​are classified as transmittance increase nodes. This classification operation must eliminate instrument noise interference: a valid node is confirmed only when three consecutive difference values ​​are positive and their absolute values ​​are greater than 0.5%.

[0058] The nodes showing an increase in transmittance are arranged in ascending order of their corresponding first-order difference values ​​to form a difference sequence. The values ​​in this difference sequence undergo a second-order difference operation to obtain a second-order difference sequence, which involves calculating the difference between adjacent first-order difference values. The position corresponding to the maximum value in the second-order difference sequence is marked as a transition characteristic node; if multiple maximum values ​​exist, the earliest one is selected. The time period before the transition characteristic node is defined as the transmittance stabilization phase, specifically from the sequence start node to the previous valid node. The time period after the transition characteristic node is defined as the transmittance transition phase, ranging from the transition characteristic node to the sequence end node. The definition of the time periods must be synchronized with the suspension processing log to ensure that the phase divisions perfectly match the equipment operation periods.

[0059] The particle size distribution data of the suspension was acquired using a dynamic light scattering device. During the transmittance stabilization phase, the particle size measurements at all valid detection time points were taken as the arithmetic mean. Each detection point was sampled 10 times per minute, and the mean was calculated after removing the maximum and minimum values. During the transmittance jump phase, the same data cleaning rules were used to calculate the mean particle size. The particle size distribution difference coefficient was calculated as follows: the absolute value of the mean transmittance during the stabilization phase was subtracted from the mean transmittance during the jump phase, divided by the maximum allowable particle size (200 nm), and the result was rounded to three decimal places. The preset difference threshold was 0.15, determined based on historical data statistical analysis. If the calculated difference coefficient exceeded the preset threshold, the membrane separation equipment level corresponding to the current processing stage was marked as the target separation level.

[0060] The data acquisition equipment is kept in a constant temperature environment (25±0.5℃) during operation. The transmittance detection device must be calibrated with a standard reference solution before each use. The transmittance value of the standard solution is preset to 90.0%, and the calibration procedure is automatically triggered when the measured deviation exceeds ±0.3%. The sampling cell capacity of the dynamic light scattering instrument is fixed at 5mL, and ultrasonic oscillation for 30 seconds is performed before each measurement to eliminate the influence of air bubbles. Particle size data recording includes three dimensions: median particle size, mode particle size, and root mean square particle size. The median particle size value is used uniformly when calculating the coefficient of difference.

[0061] The anomaly handling mechanism of the stage division algorithm includes: when the maximum value of the second-order difference is lower than the preset sensitivity (0.8), it automatically switches to the third-order difference to relocate the transition node; when the transmittance stabilization stage lasts for less than 15% of the total process, it is judged as an invalid division and optical data is re-acquired; if there are fewer than five valid data points in a certain period during the particle size mean calculation, the data in that period will not participate in the difference coefficient calculation. All judgment operations have independent timestamp records, which are stored together with the original data in the operation database.

[0062] The process for determining the target separation level incorporates a multi-level verification mechanism. When the difference coefficient exceeds the threshold for the first time, the system automatically retrieves the difference coefficient records from the previous three historical operations for that level. If two or more of the three records also exceed the threshold, the target separation level is confirmed to be valid. Level location information is mapped using the physical location encoding of the membrane separation equipment: the centrifugation section is labeled L1, the microfiltration section L2, the ultrafiltration section L3, and the nanofiltration section L4. Each level allocation result generates equipment configuration instructions, controlling the valves to switch to the corresponding membrane module channel.

[0063] Quality control requirements during implementation: Optical inspection equipment undergoes daily zero-point drift testing; if the drift value exceeds ±0.1%, it must be suspended from use; the particle size detection unit is verified with standard particles every four hours, and the median particle size error is controlled within ±3nm; all difference calculations retain the original decimal places, and rounding is only performed when the final difference coefficient is output. The operation record table for the target separation level contains a complete parameter chain including division time, transition node position, mean value of the stable stage, mean value of the transition stage, calculation coefficients, and judgment results. The parameter storage format is a dual backup of binary encoding and text.

[0064] The system is configured with auxiliary monitoring views: a time axis dynamically displays an overlay of the transmittance curve and the first-order difference curve, with abrupt change characteristic nodes marked with a red cursor; the right panel synchronously displays the current particle size distribution box plot, with box plots of the stable stage and the abrupt change stage compared side by side. Historical operation data can be loaded as a semi-transparent background curve to help operators identify abnormal fluctuations.

[0065] When the same layer is consecutively identified as the target separation layer in multiple operations, an operational parameter optimization suggestion is triggered. The system analyzes twelve parameters, including the transmittance change rate, pressure curve, and flux attenuation value of the previous three layers for that layer, and generates a membrane module replacement or cleaning maintenance suggestion. The final setting result of each target separation layer is transmitted to the upper-level control system for coordinated adjustment of the pressure gradient mode parameters. If the layer determination fails, it automatically switches to a predefined safety mode and performs the separation operation using the default combination of L2-L3 layers.

[0066] The actuator has a hardware self-test function. Before each startup, it checks key indicators such as the cleanliness of the optical sensor lens, the energy value of the laser scattering unit, and the signal response time of the data acquisition card. Fault codes are managed in three levels: excessive transmittance signal drift triggers a yellow alarm, continuous loss of differential values ​​triggers an orange alarm, and interruption of particle size detection triggers a red shutdown command. Alarm events record detailed operating environment parameters, including suspension temperature, stirring rate, ambient light intensity, and other auxiliary diagnostic information.

[0067] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0068] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for extracting melanin decomposition components from Coix seed based on membrane separation, characterized in that, Includes the following steps: Obtain the pretreatment parameters of the raw material Coix seed and the optical characteristic data of the initial suspension; Based on the trend of optical characteristic data of the initial suspension with the processing time, the transmittance is divided into a stable stage and a transmittance jump stage. Based on the difference coefficient of particle size distribution of the suspension in the stable stage and the transmittance jump stage, the target separation level is screened. Under the target separation level, based on the transmembrane flux attenuation characteristics of the membrane module in the stable stage and the transmittance jump stage, combined with the membrane pore size distribution data and operating pressure parameters, the separation efficiency index of each membrane module is calculated. Based on the residual component concentration data and membrane fouling rate of the membrane module during the transmittance jump stage, and combined with the separation efficiency index, the cycle tolerance factor of each membrane module is determined; target membrane modules are screened according to the cycle tolerance factor, and a multi-stage membrane separation sequence is constructed by combining the permeate component data corresponding to the target separation level during the transmittance jump stage; a gradient extraction process for the melanin decomposition components of Coix seed is executed based on the multi-stage membrane separation sequence.

2. The method for extracting melanin decomposition components from Coix seed based on membrane separation according to claim 1, characterized in that, The preprocessing parameters include the average particle size and crushing temperature; the optical characteristic data include real-time transmittance and dynamic light scattering data; and the transmembrane flux decay characteristics include the initial flux decay rate and the steady-state flux maintenance time.

3. The method for extracting melanin decomposition components from Coix seed based on membrane separation according to claim 2, characterized in that, The method for obtaining the separation efficiency index includes: At any target separation level, membrane modules with cross-membrane flux records during the transmittance stabilization phase are designated as candidate membrane modules, and membrane modules with permeate composition data during the transmittance jump phase are designated as associated membrane modules. Any candidate membrane module is selected as the target candidate membrane module, and any associated membrane module is selected as the target associated membrane module. Based on the initial flux decay rate, steady-state flux maintenance duration, and membrane pore size distribution matching degree of the target candidate membrane module and the target associated membrane module, a flux synergy factor relative to the target associated membrane module is obtained. The pressure fluctuation similarity between the operating pressure parameters of the target candidate membrane module during the transmittance stabilization phase and the operating pressure parameters of the target associated membrane module during the transmittance jump phase is calculated to obtain the pressure characteristic factor corresponding to each pressure monitoring point. Using the flux synergy factor as a weight, the pressure characteristic factors corresponding to each pressure monitoring point are weighted and fused to obtain the membrane flux contribution of the target candidate membrane module relative to the target associated membrane module. The average of the membrane flux contribution of the target candidate membrane module relative to all associated membrane modules is used as the separation efficiency index of the target candidate membrane module.

4. The method for extracting melanin decomposition components from Coix seed based on membrane separation according to claim 3, characterized in that, The method for obtaining the flux synergy factor includes: An arbitrary operating time point is selected as the baseline node; based on the parameter differences between the target candidate membrane module and the target associated membrane module in the dimensions of average particle size and crushing temperature, the membrane material compatibility coefficient is determined; the ratio of the operating pressure parameter of the target associated membrane module at the baseline node to the operating pressure parameter of the target candidate membrane module at the baseline node is used as the first synergy coefficient; the ratio of the membrane material compatibility coefficient to the steady-state flux maintenance time of the target associated membrane module at the baseline node is used as the second synergy coefficient; the product of the first synergy coefficient and the second synergy coefficient is calculated to obtain the flux synergy factor of the target candidate membrane module relative to the target associated membrane module at the baseline node.

5. The method for extracting melanin decomposition components from Coix seed based on membrane separation according to claim 3, characterized in that, The method for obtaining the cyclic tolerance factor includes: Based on the average concentration of residual components and the membrane fouling rate at each operating time point, the fouling accumulation index of the target candidate membrane module is calculated. The product of the separation efficiency index and the fouling accumulation index of the target candidate membrane module is normalized and used as the cycle tolerance factor of the target candidate membrane module.

6. The method for extracting melanin decomposition components from Coix seed based on membrane separation according to claim 5, characterized in that, The method for obtaining the pollution accumulation index includes: The characteristic correlation value between the average concentration of residual components and the membrane fouling rate of the target candidate membrane module at each operating time point is calculated. The average of the characteristic correlation values ​​corresponding to all operating time points during the transmittance jump stage is used as the fouling accumulation index of the target candidate membrane module.

7. The method for extracting melanin decomposition components from Coix seed based on membrane separation according to claim 5, characterized in that, The construction of the multi-stage membrane separation sequence includes: Candidate membrane modules with a cycle tolerance factor greater than a preset tolerance threshold are selected as the main separation units in the multi-stage membrane separation sequence. The fouling accumulation index of each associated membrane module is obtained, and the product of the flux synergy factor between the main separation unit and the associated membrane module and the fouling accumulation index of the associated membrane module is used as the coupling matching degree of the associated membrane module relative to the main separation unit. For any main separation unit, a membrane assembly sequence is constructed in descending order of the coupling matching degree of the associated membrane module relative to the main separation unit, and the coupling matching degree of the associated membrane modules at the same level in the sequence is equal. The membrane assembly sequences of all main separation units constitute the multi-stage membrane separation sequence.

8. The method for extracting melanin decomposition components from Coix seed based on membrane separation according to claim 7, characterized in that, The gradient extraction process includes: In the multi-stage membrane separation sequence, the layer containing the first shared associated membrane module is designated as the primary separation layer; after the primary separation layer, the layer containing the associated membrane module with the highest coupling matching degree is designated as the secondary separation layer; the membrane module between the main separation unit and the primary separation layer is subjected to a first pressure gradient mode, the membrane module between the primary separation layer and the secondary separation layer is subjected to a second pressure gradient mode, and the membrane module between the secondary separation layer and the final membrane module is subjected to a third pressure gradient mode to perform transmembrane separation.

9. The method for extracting melanin decomposition components from Coix seed based on membrane separation according to claim 1, characterized in that, The division into the stable transmittance stage and the transmittance jump stage includes: The real-time transmittance values ​​at each time point are obtained to form a transmittance sequence. The first-order difference value of the transmittance sequence is calculated, and the time points with positive first-order difference values ​​are selected as transmittance increase nodes. The transmittance increase nodes are arranged in ascending order according to their corresponding first-order difference values ​​to obtain a difference sequence. The second-order difference value of the difference sequence is calculated, and the transmittance increase node corresponding to the maximum value of the second-order difference value is set as the jump characteristic node. The time period before the jump characteristic node is the transmittance stable stage, and the time period after the jump characteristic node is the transmittance jump stage.

10. The method for extracting melanin decomposition components from Coix seed based on membrane separation according to claim 9, characterized in that, The separation hierarchy of the screening targets includes: The difference between the average particle size of the suspension in the stable transmittance stage and the average particle size of the suspension in the jump transmittance stage is calculated and normalized to obtain the particle size distribution difference coefficient. If the particle size distribution difference coefficient is greater than the preset difference threshold, the separation level corresponding to the processing stage is taken as the target separation level.

Images