A production control method and system for Polygonatum odoratum instant powder

Abstract

This invention discloses a production control method and system for Polygonatum odoratum instant powder, specifically relating to the field of production control technology. It involves acquiring real-time quality detection data of Polygonatum odoratum raw material batches, analyzing the component structure characteristics and quality fluctuation characteristics of each batch to generate quality assessment data; based on the quality assessment data, selecting candidate process routes from a process route library that match the raw material quality characteristics and calculating the product yield prediction data corresponding to the candidate process routes; based on the quality assessment data, comprehensively evaluating the active ingredient retention level and instant solubility performance indicators under the candidate process routes to generate process performance assessment data; fusing and analyzing the product yield prediction data and process performance assessment data to form optimal process route decision data; and dynamically switching the process parameter configuration of the production line according to the optimal process route decision data to achieve continuous production process control. This invention can effectively improve the batch stability of Polygonatum odoratum instant powder products.

Description

Technical Field

[0001] This invention relates to the field of production control technology, and more specifically, to a production control method and system for Polygonatum odoratum instant powder. Background Technology

[0002] Existing production control methods for Polygonatum odoratum instant powder are mostly based on fixed combinations of process parameters or a small number of pre-set process routes to complete production organization and process control. When there are differences in the source of Polygonatum odoratum raw materials, processing status and batch quality, the existing production control methods often cannot form a process route selection and process control arrangement that matches the differences in raw material batches.

[0003] Existing continuous production control methods for Polygonatum odoratum instant powder lack production decision-making based on batch-to-batch differences in raw materials. This makes it difficult to select appropriate process routes and configure process parameters according to the quality characteristics of each batch of raw materials, resulting in insufficient adaptability of the continuous production process to batch fluctuations in raw materials.

[0004] To address the aforementioned problems, a technical solution is provided. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a production control method and system for Polygonatum odoratum instant powder to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A production control method for Polygonatum odoratum instant powder includes the following steps:

[0008] S1: Obtain real-time quality inspection data of batches of Polygonatum sibiricum raw materials entering the continuous production line, standardize the real-time quality inspection data, and form a raw material quality characteristic dataset.

[0009] S2: Based on the raw material quality characteristic dataset, analyze the compositional structure characteristics and quality fluctuation characteristics of batches of Polygonatum odoratum raw materials to generate quality assessment data reflecting the differences in raw materials;

[0010] S3: Based on quality assessment data, select candidate process routes that match the quality characteristics of raw materials from the process route library, and calculate the product yield prediction data corresponding to the candidate process routes.

[0011] S4: Based on the quality assessment data, comprehensively evaluate the retention level of active ingredients and the instant solubility performance index under the candidate process routes to generate process performance assessment data.

[0012] S5: Integrate and analyze product yield prediction data with process performance evaluation data to form optimal process route decision data;

[0013] S6: Dynamically switch the process parameter configuration of the production line based on the optimal process route decision data.

[0014] In a preferred embodiment, S1 specifically refers to:

[0015] Obtain online testing data and offline sampling data corresponding to batches of Polygonatum sibiricum raw materials entering the continuous production line;

[0016] Online testing data and offline sampling data are correlated and aligned according to the collection time and production feeding sequence;

[0017] Based on the correlated and aligned online detection data and offline sampling data, detection index data characterizing the intrinsic quality status of batches of Polygonatum odoratum raw materials were extracted.

[0018] The test index data are preprocessed to form a raw material quality characteristic dataset.

[0019] In a preferred embodiment, S2 specifically refers to:

[0020] Based on the raw material quality characteristic dataset, the correlation between characteristic indicators of batches of Polygonatum sibiricum raw materials is identified;

[0021] Based on the correlation between characteristic indicators, determine the changing trend of the compositional structure characteristics of raw material batches;

[0022] By combining the changing trends of the composition and structure characteristics of raw material batches, the range of quality fluctuation characteristics is determined, and quality assessment data reflecting the differences in raw materials is generated.

[0023] In a preferred embodiment, S3 specifically refers to:

[0024] Based on the quality assessment data, read the applicable raw material quality constraint information for each process route in the process route library;

[0025] Matching score data is obtained by matching the applicable raw material quality constraint information with the quality assessment data;

[0026] Candidate process routes are selected based on the matching score data;

[0027] Based on the candidate process route, relevant historical production record data is retrieved and correlated with quality assessment data to generate product yield prediction data corresponding to the candidate process route.

[0028] In a preferred embodiment, S4 specifically refers to:

[0029] Based on quality assessment data, we obtained measured data of active ingredients and instant solubility performance from historical batches associated with candidate process routes.

[0030] A mapping relationship between the measured active ingredient data and the measured instant dissolution performance data of historical batches produced corresponding to the candidate process routes and the quality assessment data was established.

[0031] Based on the mapping relationship, the evaluation values ​​of the active ingredient retention level and rapid dissolution performance index corresponding to the candidate process route are calculated, and process performance evaluation data are generated.

[0032] In a preferred embodiment, S5 specifically refers to:

[0033] Determine the product yield level information for each candidate process route based on product yield prediction data;

[0034] Based on the process performance evaluation data, determine the active ingredient retention level and instant solubility level information corresponding to each candidate process route;

[0035] Based on product yield level information, active ingredient retention level information, and instant solubility performance information, the comprehensive score data of each candidate process route is calculated to generate the optimal process route decision data.

[0036] In a preferred embodiment, S6 specifically refers to:

[0037] Extract the real-time setpoints of the corresponding process parameters based on the optimal process route decision data;

[0038] Generate production line process configuration switching instructions based on real-time set values ​​of process parameters;

[0039] Adjust the current process parameter configuration of the production line according to the production line process configuration switching instruction, and dynamically execute the process route corresponding to the optimal process route decision data.

[0040] On the other hand, the present invention provides a production control system for Polygonatum odoratum instant powder, comprising:

[0041] Real-time acquisition module: Acquires real-time quality inspection data of batches of Polygonatum sibiricum raw materials entering the continuous production line, standardizes the real-time quality inspection data, and forms a raw material quality characteristic dataset;

[0042] Quality assessment module: Based on the raw material quality characteristic dataset, the module analyzes the compositional structure characteristics and quality fluctuation characteristics of batches of Polygonatum odoratum raw materials, and generates quality assessment data that reflects the differences in raw materials.

[0043] Route selection module: Based on quality assessment data, it selects candidate process routes from the process route library that match the quality characteristics of raw materials, and calculates the product yield prediction data corresponding to the candidate process routes.

[0044] Performance evaluation module: Based on quality evaluation data, it comprehensively evaluates the retention level of active ingredients and rapid dissolution performance under candidate process routes, and generates process performance evaluation data;

[0045] Fusion Decision Module: This module integrates and analyzes product yield prediction data with process performance evaluation data to generate optimal process route decision data.

[0046] Parameter switching module: Dynamically switches the process parameter configuration of the production line based on the optimal process route decision data.

[0047] The technical effects and advantages of the production control method and system for Polygonatum odoratum instant powder of the present invention are as follows:

[0048] By acquiring real-time quality inspection data of batches of Polygonatum sibiricum raw materials entering a continuous production line and standardizing it to form a raw material quality characteristic dataset, the comparability and usability of Polygonatum sibiricum raw material batch quality information under a unified scale are achieved. By analyzing the component structure characteristics and quality fluctuation characteristics of Polygonatum sibiricum raw material batches based on the raw material quality characteristic dataset and generating quality assessment data, the structured representation and decision-making expression of batch differences in Polygonatum sibiricum raw materials are achieved. By screening candidate process routes from a process route library based on quality assessment data and calculating product yield prediction data, a quantitative correlation between process route selection and output performance is achieved. By comprehensively evaluating the active ingredient retention level and instant solubility index under candidate process routes, process performance assessment data is generated, achieving a unified assessment standard for key quality indicators. By fusing and analyzing product yield prediction data and process performance assessment data, optimal process route decision data is formed, achieving a synergistic trade-off between output and quality. Finally, by dynamically switching the production line process parameter configuration based on the optimal process route decision data, adaptive control and stable operation of the continuous production process are achieved, thereby improving the pertinence of Polygonatum sibiricum instant powder production decisions. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of a production control method for Polygonatum odoratum instant powder according to the present invention;

[0050] Figure 2 This is a schematic diagram of the production control system for Polygonatum odoratum instant powder according to the present invention. Detailed Implementation

[0051] 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention. Example 1

[0052] Figure 1 The present invention provides a production control method for Polygonatum odoratum instant powder, which includes the following steps:

[0053] S1: Obtain real-time quality inspection data of batches of Polygonatum sibiricum raw materials entering the continuous production line, standardize the real-time quality inspection data, and form a raw material quality characteristic dataset.

[0054] S2: Based on the raw material quality characteristic dataset, analyze the compositional structure characteristics and quality fluctuation characteristics of batches of Polygonatum odoratum raw materials to generate quality assessment data reflecting the differences in raw materials;

[0055] S3: Based on quality assessment data, select candidate process routes that match the quality characteristics of raw materials from the process route library, and calculate the product yield prediction data corresponding to the candidate process routes.

[0056] S4: Based on the quality assessment data, comprehensively evaluate the retention level of active ingredients and the instant solubility performance index under the candidate process routes to generate process performance assessment data.

[0057] S5: Integrate and analyze product yield prediction data with process performance evaluation data to form optimal process route decision data;

[0058] S6: Dynamically switch the process parameter configuration of the production line based on the optimal process route decision data.

[0059] S1: Obtain real-time quality inspection data of batches of Polygonatum sibiricum raw materials entering the continuous production line, standardize the real-time quality inspection data, and form a raw material quality characteristic dataset, including:

[0060] Obtain online testing data and offline sampling data corresponding to batches of Polygonatum sibiricum raw materials entering the continuous production line;

[0061] Specifically, at the raw material inlet of the continuous production line, a spectrometer, a humidity detector, and a component content analyzer are installed to collect online detection data corresponding to each batch of Polygonatum odoratum raw material entering the continuous production line in real time. The spectrometer collects near-infrared spectral information of the raw material. The spectrometer uses a Fourier transform near-infrared spectrometer with a wavelength range of 780nm to 2500nm. For example, the scanning resolution is set to 8cm^-1, and the single scan time is 5 seconds. The acquired near-infrared spectral information is expressed in the form of a reflectance curve. The humidity detector collects the surface humidity information of the raw material in real time, expressed as a percentage of moisture content. The measurement accuracy of moisture content is set, for example, to ±0.5%. The component content analyzer uses a high-performance liquid chromatograph to analyze the content data of active polysaccharides and saponins in the Polygonatum odoratum raw material in real time. The detection wavelength is set to 203nm. The content of active polysaccharides and saponins is expressed as a mass percentage. For example, the content measurement accuracy is set to ±0.1%.

[0062] At fixed time intervals, such as every 60 minutes, samples of Polygonatum sibiricum raw materials entering the continuous production line are randomly selected and sent to the laboratory for offline quality analysis to obtain offline sampling data. Offline sampling data was obtained through manual laboratory testing: The sampled material was dried and ground to control the particle size range, for example, from 50 to 100 micrometers. The polysaccharide and saponin content in the Polygonatum sibiricum raw material samples was determined using ultraviolet spectrophotometry. The ground raw material sample was added to distilled water and extracted for 1 hour in an 80°C water bath. The extract was obtained by centrifugation and filtration. The absorbance was measured at 203 nm using an ultraviolet spectrophotometer. The polysaccharide and saponin content in the Polygonatum sibiricum raw material sample was calculated based on a pre-plotted standard curve, expressed as a mass percentage, with a measurement accuracy set, for example, ±0.1%. Simultaneously, the moisture content of the sample was determined using a drying weight loss method. The ground sample was weighed and placed in a constant temperature drying oven at 105°C for 4 hours. It was weighed again, and the moisture content was calculated based on the weight difference before and after drying, with a content accuracy set, for example, ±0.3%.

[0063] Online testing data and offline sampling data are correlated and aligned according to the collection time and production feeding sequence;

[0064] Specifically, based on the actual feeding sequence of the Polygonatum sibiricum raw material batches entering the continuous production line, the real-time online detection data is recorded and saved according to its corresponding collection timestamp; after obtaining the offline sampling data, it is marked and saved according to the actual feeding batch number and sampling time corresponding to the sample; based on the feeding batch number and timestamp, the online detection data and offline sampling data are established to establish a correspondence, that is, within the same feeding batch, the online detection data and offline sampling data are matched according to the timestamp to ensure that each group of raw material batch data contains both online detection data and offline sampling data.

[0065] Based on the correlated and aligned online detection data and offline sampling data, detection index data characterizing the intrinsic quality status of batches of Polygonatum odoratum raw materials were extracted.

[0066] Specifically, near-infrared spectral reflectance data, raw material moisture content data, and polysaccharide and saponin content data are extracted from the correlated online detection data and offline sampling data, respectively; the near-infrared spectral reflectance data, moisture content data, and polysaccharide and saponin content data are combined to generate detection index data.

[0067] Preprocess the test index data to form a raw material quality characteristic dataset;

[0068] Specifically, for spectral feature data, standardization is achieved using the standard normal transformation method, which uses the average and standard deviation of all batches of data for the corresponding wavelength as the basis for transformation. For moisture content data, polysaccharide content data, and saponin content data, normalization is used to achieve standardization, which involves subtracting the historical minimum value from each indicator data and then dividing by the difference between the historical maximum and the historical minimum value. Outlier detection is performed on all standardized data, for example, using a Z-score-based method to consider data with an absolute value greater than 3 as outliers and remove them. All data after outlier processing are summarized to form a raw material quality feature dataset.

[0069] S2: Based on the raw material quality characteristic dataset, the compositional structure characteristics and quality fluctuation characteristics of batches of Polygonatum sibiricum raw materials are analyzed to generate quality assessment data reflecting the differences in raw materials, including:

[0070] Based on the raw material quality characteristic dataset, the correlation between characteristic indicators of batches of Polygonatum sibiricum raw materials is identified;

[0071] Specifically, the raw material quality characteristic dataset includes near-infrared spectral reflectance data, raw material moisture content data, polysaccharide content data, and saponin content data. Principal component analysis (PCA) is performed on the near-infrared spectral reflectance data to reduce data dimensionality and extract representative principal components. A covariance matrix is ​​constructed based on the near-infrared spectral reflectance data, and eigenvalues ​​and eigenvectors are calculated. Then, the number of principal components is determined based on the cumulative variance contribution rate, for example, set between 85% and 95%. Finally, with the determined number of principal components, dimensionality reduction processing is performed on the near-infrared spectral reflectance data to obtain the spectral principal component data in the raw material quality characteristic dataset.

[0072] After obtaining the principal component data of the spectra, correlation coefficient analysis was used to identify the correlation between the principal component data and the data on raw material moisture content, polysaccharide content, and saponin content. The Pearson correlation coefficient method was used for correlation coefficient analysis. By calculating the correlation coefficients between the principal component data and the data on raw material moisture content, polysaccharide content, and saponin content, the strength and direction of the correlation between the characteristic indicators were determined. For example, a correlation coefficient with an absolute value greater than 0.7 was defined as a strong correlation, an absolute value between 0.3 and 0.7 as a moderate correlation, and a correlation less than 0.3 as a weak correlation, thus obtaining the correlation data between the characteristic indicators.

[0073] Based on the correlation between characteristic indicators, determine the changing trend of the compositional structure characteristics of raw material batches;

[0074] Specifically, based on the combination of characteristic indicators with strong correlations obtained from correlation coefficient analysis, regression analysis is used to determine the changing trends of component structure characteristics. Multiple linear regression analysis is employed to establish a regression model between the characteristic indicators and the changing trends of the component structure characteristics of raw material batches, expressed as:

[0075] ;in, This indicates the structural characteristics of the raw material batch to be analyzed; This represents the i-th feature index that is strongly correlated with the compositional and structural characteristics of the raw material batch; This represents the regression coefficient corresponding to the i-th feature index; Represents the constant term in the regression equation; The random error term is represented; the regression coefficients are determined using the least squares method. and By minimizing the sum of squared residuals between the predicted values ​​of the regression equation and the actual component structure characteristics data, the optimal solution of the model parameters is determined, thereby realizing the trend analysis of the changing characteristics of the component structure of raw material batches.

[0076] By combining the changing trends of the composition and structural characteristics of raw material batches, the range of quality fluctuation characteristics is determined, and quality assessment data reflecting the differences in raw materials is generated.

[0077] Specifically, based on the trend data of the changing characteristics of the raw material batch composition obtained from regression analysis, the fluctuation range of the composition structure characteristics is determined using the standard deviation analysis method. The standard deviation analysis formula is expressed as:

[0078] ;in, The standard deviation represents the trend of changes in the compositional structure characteristics, reflecting the degree of fluctuation in the compositional structure characteristics; This represents the numerical value indicating the trend of changes in the compositional characteristics of the raw material batch corresponding to the kth batch of feed. This represents the average value of the trend data of changes in the compositional structure characteristics corresponding to all batches of feed. This represents the total number of feed batches included in the dataset. The threshold for standard deviation analysis is determined by calculating the historical standard deviations of all component structure characteristic data. A significant fluctuation is defined as the current standard deviation being greater than 1.5 times the historical average standard deviation, while a fluctuation below 1.5 times is considered normal. This method is used to define the fluctuation range of the raw material batch component structure characteristics.

[0079] Based on the significant fluctuation range of component structure characteristics, three quality fluctuation levels—high fluctuation, medium fluctuation, and low fluctuation—are defined to quantify the magnitude of the fluctuation range. For example, batches with a standard deviation of component structure characteristics greater than twice the historical average standard deviation are classified as high fluctuation, those between 1.5 and 2 times are medium fluctuation, and those less than 1.5 times are low fluctuation. The fluctuation data of component structure characteristics for all raw material batches are then classified into levels according to the level definition rules to obtain corresponding fluctuation level labels. Finally, the fluctuation level label corresponding to each batch of raw materials is used as quality assessment data reflecting the differences in raw materials.

[0080] S3: Based on quality assessment data, select candidate process routes from the process route library that match the raw material quality characteristics, and calculate the product yield prediction data corresponding to the candidate process routes, including:

[0081] Based on the quality assessment data, read the applicable raw material quality constraint information for each process route in the process route library;

[0082] Specifically, the process route library is a pre-built database of process parameters that records multiple different production process routes for Polygonatum odoratum instant powder. Each production process route includes a combination of production control parameters and raw material quality constraints suitable for the corresponding process route, i.e., applicable raw material quality constraint information. The applicable raw material quality constraint information includes constraints on the polysaccharide content range, saponin content range, moisture content range, and the characteristic range of the main component data of the raw material spectrum. Among them, the polysaccharide content range constraints and saponin content range constraints are expressed in the form of upper and lower limits of mass percentage. For example, the applicable range of polysaccharide content is set to 8.0% to 12.0%, and the applicable range of saponin content is set to 2.0% to 5.0%. The moisture content range constraints are also expressed in the form of upper and lower limits of mass percentage. For example, the applicable range of moisture content is set to 10.0% to 15.0%. The raw material spectral principal component data characteristic range constraints are expressed in the form of upper and lower limits of spectral principal components. For example, the applicable ranges of the first two principal components are set to the value range of principal component 1 between -1.5 and 1.5, and the value range of principal component 2 between -2.0 and 2.0, respectively.

[0083] Matching score data is obtained by matching the applicable raw material quality constraint information with the quality assessment data;

[0084] Specifically, the quality assessment data of each batch of Polygonatum sibiricum raw materials is matched item by item with the applicable raw material quality constraint information corresponding to each process route in the process route library. A scoring-based matching algorithm is used: if a certain quality assessment data of a batch of Polygonatum sibiricum raw materials falls completely within the range set by the corresponding applicable raw material quality constraint information, it receives full marks, for example, a single item score is set at 10 points; if a certain quality assessment data of a batch of Polygonatum sibiricum raw materials is outside the range set by the corresponding applicable raw material quality constraint information, but the deviation does not exceed the set allowable threshold, the score is reduced accordingly based on the degree of deviation. For example, the allowable threshold is set to 10% above and below the corresponding range. If it exceeds the upper or lower limit but is within the 10% allowable threshold range, the score is linearly reduced according to the degree of deviation from the upper or lower limit. If it deviates from the 10% allowable threshold range, the score is 0 points. Then, the scores of all individual matching items for each batch of Polygonatum sibiricum raw materials for a set of process routes are summed to obtain the total matching score data. Finally, for each batch of Polygonatum sibiricum raw materials, the total matching score data is calculated for all process routes in the process route library, thus forming a matching score dataset.

[0085] Candidate process routes are selected based on the matching score data;

[0086] Specifically, for each batch of Polygonatum sibiricum raw materials, a screening threshold for candidate process routes is set based on the total matching score data of all process routes. This threshold is determined using historical data statistical methods; for example, based on the average historical process route matching score data, process routes with scores exceeding 80% of the historical average are considered candidate process routes to ensure a high degree of suitability. Process routes exceeding the screening threshold are marked, forming a set of candidate process routes for each batch of Polygonatum sibiricum raw materials. This ensures that production decisions can cover multiple alternative process routes within the optimal range.

[0087] Based on the candidate process route, retrieve the associated historical production record data and perform correlation mapping calculation with the quality assessment data to generate product yield prediction data corresponding to the candidate process route.

[0088] Specifically, historical production record data refers to the historical product yield data collected and stored during the actual operation of each process route. The historical product yield data is obtained as follows: for each process route, the product yield is the ratio of the weight of the Polygonatum odoratum instant powder produced during continuous production to the total weight of the raw materials entering the production line. The calculation precision of the product yield is set to, for example, two decimal places, and the yield is expressed as a percentage, such as a product yield of 85.00%. The product yield data is stored batch by batch and marked with its corresponding relationship to the process route.

[0089] For each candidate process route, historical product yield data corresponding to all process routes are retrieved from historical production records. Using the compositional characteristics and fluctuation levels of the corresponding batches of raw materials from the quality assessment data as input and the historical product yield data as output, a mapping relationship is established through a linear regression model. The formula for the linear regression model is:

[0090] ;in, For product yield prediction data; The j-th characteristic index in the component structure characteristics, such as including but not limited to polysaccharide content, saponin content and water content; This is the numerical value corresponding to the quality fluctuation level, for example, high fluctuation level is recorded as 3, medium fluctuation level as 2, and low fluctuation level as 1; These are the regression coefficients for the corresponding characteristic indicators and quality fluctuation levels, respectively; The constant term in the linear regression model is used. All model parameters are determined using the least squares method, specifically by minimizing the sum of squared residuals between the predicted data and historical product yield data. After determining the regression model parameters, based on the component structure characteristics and fluctuation levels of the quality assessment data, the predicted product yield data for each candidate process route for the current batch of Polygonatum odoratum raw materials is calculated.

[0091] S4: Based on quality assessment data, comprehensively evaluate the retention level of active ingredients and rapid dissolution performance under candidate process routes to generate process performance assessment data, including:

[0092] Based on quality assessment data, we obtained measured data of active ingredients and instant solubility performance from historical batches associated with candidate process routes.

[0093] Specifically, the measured data of active ingredients and instant solubility performance from historical batches were obtained through sampling and testing of each batch of Polygonatum odoratum instant powder during the historical production process of the candidate process route. The measured data of active ingredients includes the content of active polysaccharides and saponins in the Polygonatum odoratum instant powder, obtained using high-performance liquid chromatography (HPLC). All content results are expressed as a mass percentage, with a measurement accuracy set, for example, to ±0.05%.

[0094] Two grams of sample were randomly selected from each batch of Polygonatum odoratum instant powder and placed in a volumetric flask. A measured amount of distilled water was added and stirred at room temperature. The stirring speed was set to, for example, 120 revolutions per minute, and the stirring time was 3 minutes. After stirring was stopped, the sample was allowed to stand for 30 seconds. The time required for the sample to completely dissolve was observed and recorded, which was defined as the dissolution time. At the same time, the solution obtained after dissolution was filtered, and the residue was collected. The residue was dried and its mass was weighed. The proportion of the residue mass to the initial sample amount was calculated to obtain the insoluble content data of the corresponding batch of instant powder, expressed as a mass percentage, with a measurement accuracy of, for example, ±0.1%. Finally, the measured data of instant dissolution performance were expressed by combining the dissolution time and insoluble content of each batch.

[0095] A mapping relationship between the measured active ingredient data and the measured instant dissolution performance data of historical batches produced corresponding to the candidate process routes and the quality assessment data was established.

[0096] Specifically, the active polysaccharide content, saponin content, dissolution time, and insoluble matter content of each batch of Polygonatum odoratum instant powder from historical batch production data under the candidate process route are used as output variables, and the Polygonatum odoratum raw material quality assessment data input during the corresponding batch production is used as input variables. A mathematical mapping relationship from quality assessment data to measured process performance data is established using partial least squares regression. Both input and output variables are centered and normalized, for example, using minimum-maximum normalization to unify them to the [0,1] interval. Principal components of the input and output variables are iteratively extracted using the partial least squares regression algorithm. During the principal component extraction process, the number of iterations is determined, for example, by cross-validation. Iteration stops when the cross-validation error corresponding to the newly extracted principal component does not decrease. The regression coefficients are determined by minimizing the sum of squared residuals between the model prediction value and the actual historical measured data. The regression coefficients obtained by the partial least squares algorithm form a stable mapping relationship between the input and output variables.

[0097] Based on the mapping relationship, the evaluation values ​​of the active ingredient retention level and rapid solubility performance index corresponding to the candidate process routes are calculated, and process performance evaluation data are generated.

[0098] Specifically, for the quality assessment data of the current batch of Polygonatum sibiricum raw materials, the regression model formula of the mapping relationship is substituted, and the component structure characteristic index and quality fluctuation level value are used as input variables. Using the corresponding regression coefficients and constant terms, the predicted values ​​of active polysaccharide content, saponin content, dissolution time, and insoluble matter content of the current batch of Polygonatum sibiricum raw materials for each candidate process route are calculated respectively. The calculated predicted values ​​are used to determine the evaluation values ​​of the active ingredient retention level and the rapid dissolution performance index. The evaluation value of the active ingredient retention level index is obtained by weighted summation of the predicted values ​​of active polysaccharide content and saponin content. The weighting coefficient is set according to, for example, the contribution ratio of the two active ingredients in the actual efficacy, such as 60% for active polysaccharide and 40% for saponin. The evaluation value of the rapid dissolution performance index is obtained by using the predicted dissolution time... The predicted values ​​for dissolution time and insoluble content are converted into scores and then weighted and summed. The scoring is achieved by determining the best and worst performance limits for dissolution time and insoluble content based on historical data, then linearly converting these limits into a score from 0 to 100. For example, if the best dissolution time is 30 seconds and the worst dissolution time is 180 seconds, then a predicted dissolution time of 30 seconds receives 100 points, and 180 seconds receives 0 points. The best and worst dissolution times are calculated using linear interpolation. Similarly, the insoluble content is set at 0% for best and 5% for worst. Weights are then assigned to the converted scores, determined for example, based on their impact on instant solubility. For instance, the weight for dissolution time is set at 50%, and the weight for insoluble content is set at 50%. A weighted sum is then performed to obtain the instant solubility performance evaluation value. Finally, the evaluation values ​​for the active ingredient retention level and the instant solubility performance evaluation value for each candidate process route are combined to form the process performance evaluation data.

[0099] S5: Integrate and analyze product yield prediction data with process performance evaluation data to generate optimal process route decision data, including:

[0100] Determine the product yield level information for each candidate process route based on product yield prediction data;

[0101] Specifically, the product yield prediction data for each candidate process route is normalized. Based on the normalized product yield prediction data, product output level information is determined. Statistical analysis of historical production records determines the threshold range for level classification. For example, based on historical product yield prediction data, the product output of candidate process routes is divided into three levels according to the data distribution percentile method: high output level, medium output level, and low output level. The lower threshold for the high output level is set above the median of historical prediction data; the threshold range for the medium output level is set between the median of historical prediction data and the 20th percentile of historical data; and the threshold range for the low output level is set below the 20th percentile of historical data. Based on the threshold range, the normalized product yield prediction data is assigned to the corresponding level range, thus forming the product output level information for each candidate process route.

[0102] Based on the process performance evaluation data, determine the active ingredient retention level and instant solubility level information corresponding to each candidate process route;

[0103] Specifically, the grading of the active ingredient retention level index evaluation values ​​is as follows: Statistical analysis is performed on the active ingredient retention level index evaluation values ​​of each candidate process route. Grading thresholds are determined based on historical active ingredient retention level index evaluation value data. For example, an equal-frequency grading method is used, where historical evaluation value data is sorted and then divided into high, medium, and low grades with equal sample sizes. For instance, the first 1 / 3 of the historical data after sorting is classified as high grade, the next 1 / 3 as medium grade, and the remaining 1 / 3 as low grade. The current active ingredient retention level index evaluation value for each candidate process route is then compared to the historically defined thresholds to determine the grade, thereby obtaining the active ingredient retention level information.

[0104] The method for determining the instant solubility performance level information is as follows: For the instant solubility performance index evaluation value of each candidate process route, based on the historical instant solubility performance index evaluation data, the same equal frequency division method is used to determine the level division threshold. After sorting the historical instant solubility performance index evaluation data, three level ranges of high, medium and low are divided at equal frequency. For example, the first 1 / 3 of the data is divided into high level, the next 1 / 3 into medium level, and the remaining 1 / 3 into low level. Then, the instant solubility performance index evaluation value of each candidate process route is compared with the determined threshold range in turn to determine the level range and obtain the instant solubility performance level information.

[0105] Based on product yield level information, active ingredient retention level information, and instant solubility performance information, calculate the comprehensive score data of each candidate process route, and generate the optimal process route decision data.

[0106] Specifically, the product yield level, active ingredient retention level, and instant solubility level information corresponding to each candidate process route are converted into numerical scoring data. Different scores are assigned to the three level indicators, for example, 100 points for high level, 60 points for medium level, and 20 points for low level; and so on. The product yield level, active ingredient retention level, and instant solubility level information are quantitatively scored to obtain the individual level score data corresponding to each candidate process route.

[0107] A comprehensive score was calculated based on individual grade rating data. To reflect the importance of different indicators to the final process decision, the weights of product yield, active ingredient retention level, and instant solubility were determined using expert evaluation. The weights were determined using the analytic hierarchy process (AHP) to construct a judgment matrix comparing the importance of indicators. An expert panel compared each of the three indicators pairwise to provide a relative importance score. The eigenvector method was used to calculate the eigenvectors of the judgment matrix. The normalized results of the eigenvectors were used as the final weights for each indicator. For example, based on actual production process requirements and economic benefit goals, the weight of product yield was set to 0.4, the weight of active ingredient retention level to 0.35, and the weight of instant solubility to 0.25. Based on the determined weights and individual grade rating data, a weighted summation method was used to calculate the comprehensive score data for the candidate process routes.

[0108] The comprehensive score data of all candidate process routes are sorted, and the candidate process route with the highest score is selected as the optimal process route for producing Polygonatum odoratum instant powder from the current batch of raw materials. The specific production parameter combination of the candidate process route corresponding to the highest score is recorded. The production parameter combination includes drying temperature, extraction temperature, extraction time, and extractant concentration. For example, the drying temperature is set between 70℃ and 80℃, the extraction temperature is set between 50℃ and 60℃, the extraction time is set to 40 minutes to 60 minutes, and the extractant concentration is set to an ethanol solution with a volume fraction of 70% to 80%. The production parameter combination and the comprehensive score data of the corresponding candidate process routes together constitute the decision data for the optimal process route.

[0109] S6: Based on the optimal process route decision data, dynamically switch the process parameter configuration of the production line, including:

[0110] Extract the real-time setpoints of the corresponding process parameters based on the optimal process route decision data;

[0111] Specifically, the optimal process route decision data includes the specific combination of production parameters for the candidate process route corresponding to the highest score. These production parameter combinations include key process control parameters such as drying temperature, extraction temperature, extraction time, and extractant concentration. For example, each process control parameter might have a drying temperature range of 70℃ to 80℃, an extraction temperature range of 50℃ to 60℃, an extraction time range of 40 minutes to 60 minutes, and an extractant concentration range of 70% to 80% (v / v) ethanol solution. To achieve real-time production control, the parameter ranges need to be determined as real-time setpoints based on the quality assessment data, product yield prediction data, and process performance assessment data corresponding to the current batch of Polygonatum sibiricum raw materials.

[0112] Each process control parameter is processed individually. Taking the real-time setpoint of drying temperature as an example, based on the relationship between drying temperature and product yield, active ingredient retention level, and instant solubility in historical production data, a multi-objective optimization model of drying temperature and the above performance indicators is constructed to determine the real-time setpoint. The optimization model adopts the principle of maximizing comprehensive score, for example, using the response surface methodology. The response surface methodology model is constructed as follows: drying temperature is used as the independent variable, and historical batch product yield, active polysaccharide content, saponin content, dissolution time, and insoluble matter content are used as dependent variables. Based on historical production data, a second-order polynomial regression is used to fit the relationship between the above variables. The least squares method is used to determine the regression model parameters, and the parameter values ​​are derived from historical production data. The model's effectiveness is verified by the goodness of fit (e.g., the coefficient of determination R²), requiring a coefficient of determination greater than 0.8 for its use.

[0113] Based on the target values ​​of various performance indicators in the optimal process route decision data (e.g., maximizing product yield, maximizing active polysaccharide and saponin content, and minimizing dissolution time and insoluble matter content), the response surface model is comprehensively optimized. This comprehensive optimization employs a multi-objective weighted method, which transforms the optimization into a single-objective process by assigning specific weights to each performance indicator. For example, the weights for product yield, active polysaccharide content, saponin content, dissolution time, and insoluble matter content are set to 0.4, 0.2, 0.2, 0.1, and 0.1, respectively. The comprehensive performance indicator function is calculated using these weights. Then, using an algorithm such as particle swarm optimization, the drying temperature corresponding to the maximum value of the comprehensive performance indicator function is searched within the range of 70℃ to 80℃, with the real-time setpoint for drying temperature being 75℃.

[0114] Similarly, the extraction temperature, extraction time, and extractant concentration were determined to have real-time setpoints using response surface methodology (RSM) modeling and multi-objective optimization methods, respectively. Specifically, the optimization of the real-time extraction temperature setpoint involved using extraction temperature as the independent variable and product yield and active ingredient retention level as the main dependent variables. The setpoint was determined through RSM analysis, for example, a calculated real-time extraction temperature setpoint of 55℃. The real-time extraction time setpoint was optimized to, for example, 50 minutes. The real-time extractant concentration setpoint was optimized to, for example, a 75% (v / v) ethanol solution. Through RSM modeling and multi-objective optimization, the real-time setpoints for process parameters such as drying temperature, extraction temperature, extraction time, and extractant concentration were finally obtained.

[0115] Generate production line process configuration switching instructions based on real-time set values ​​of process parameters;

[0116] Specifically, the real-time setpoints for drying temperature, extraction temperature, extraction time, and extractant concentration are encoded and generated using digital instruction formats recognizable by the electronic control system. For example, the real-time setpoints are encoded as instruction information in JSON format; the encoded information of each real-time setpoint is combined into a production line process configuration switching instruction, and the generated production line process configuration switching instruction data packet is sent to the production line control system using a standard data interface protocol.

[0117] Adjust the current process parameter configuration of the production line according to the production line process configuration switching instruction, and dynamically execute the process route corresponding to the optimal process route decision data;

[0118] Specifically, upon receiving a real-time process configuration switching command, the continuous production line immediately uses electronic control actuators, including a PID temperature controller, solenoid valves, an automatic metering pump, and a frequency converter, to dynamically adjust the heating temperature of the drying unit to the real-time setpoint of 75°C; dynamically adjust the temperature control equipment of the extraction tank to the real-time setpoint of 55°C; dynamically adjust the stirring and extraction operation duration of the extraction tank to the real-time setpoint of 50 minutes via the PLC control system; and dynamically adjust the extractant concentration to a 75% volume fraction ethanol solution via the automatic proportioning pump and flow control valve. During the adjustment process, the real-time values ​​of each process parameter are fed back to the control interface in real time to ensure that the actual process state is highly consistent with the real-time setpoint. Example 2

[0119] The difference between Embodiment 2 and Embodiment 1 is that this embodiment introduces a production control system for Polygonatum odoratum instant powder.

[0120] Figure 2 A schematic diagram of the production control system for Polygonatum odoratum instant powder of the present invention is provided. The production control system for Polygonatum odoratum instant powder includes:

[0121] Real-time acquisition module: Acquires real-time quality inspection data of batches of Polygonatum sibiricum raw materials entering the continuous production line, standardizes the real-time quality inspection data, and forms a raw material quality characteristic dataset;

[0122] Quality assessment module: Based on the raw material quality characteristic dataset, the module analyzes the compositional structure characteristics and quality fluctuation characteristics of batches of Polygonatum odoratum raw materials, and generates quality assessment data that reflects the differences in raw materials.

[0123] Route selection module: Based on quality assessment data, it selects candidate process routes from the process route library that match the quality characteristics of raw materials, and calculates the product yield prediction data corresponding to the candidate process routes.

[0124] Performance evaluation module: Based on quality evaluation data, it comprehensively evaluates the retention level of active ingredients and rapid dissolution performance under candidate process routes, and generates process performance evaluation data;

[0125] Fusion Decision Module: This module integrates and analyzes product yield prediction data with process performance evaluation data to generate optimal process route decision data.

[0126] Parameter switching module: Dynamically switches the process parameter configuration of the production line based on the optimal process route decision data.

[0127] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0128] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0129] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0130] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0131] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0132] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0133] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0134] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0135] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A production control method for Polygonatum odoratum instant powder, characterized in that, Includes the following steps: S1: Obtain real-time quality inspection data of batches of Polygonatum sibiricum raw materials entering the continuous production line, standardize the real-time quality inspection data, and form a raw material quality characteristic dataset. S2: Based on the raw material quality characteristic dataset, analyze the compositional structure characteristics and quality fluctuation characteristics of batches of Polygonatum odoratum raw materials to generate quality assessment data reflecting the differences in raw materials; S3: Based on quality assessment data, candidate process routes matching the raw material quality characteristics are selected from the process route library, and the product yield prediction data corresponding to the candidate process routes is calculated. Specifically: based on the quality assessment data, the applicable raw material quality constraint information corresponding to each process route in the process route library is read; the applicable raw material quality constraint information is matched with the quality assessment data to obtain matching score data; candidate process routes are selected according to the matching score data; based on the candidate process routes, the associated historical production record data is retrieved and associated with the quality assessment data to generate the product yield prediction data corresponding to the candidate process routes. S4: Based on quality assessment data, comprehensively evaluate the active ingredient retention level and instant solubility performance under the candidate process routes to generate process performance assessment data; specifically: obtain the measured active ingredient and instant solubility performance data of historical batches associated with the candidate process routes based on the quality assessment data; establish a mapping relationship between the measured active ingredient and instant solubility performance data of historical batches corresponding to the candidate process routes and the quality assessment data; calculate the assessment values ​​of the active ingredient retention level and instant solubility performance corresponding to the candidate process routes according to the mapping relationship to generate process performance assessment data. S5: Integrate and analyze product yield prediction data with process performance evaluation data to form optimal process route decision data; Specifically, this involves: determining the product yield level information for each candidate process route based on product yield prediction data; and determining the active ingredient retention level and instant solubility level information for each candidate process route based on process performance evaluation data. Based on product yield level information, active ingredient retention level information, and instant solubility performance information, calculate the comprehensive score data of each candidate process route, and generate the optimal process route decision data. S6: Dynamically switch the process parameter configuration of the production line based on the optimal process route decision data.

2. The production control method for instant Polygonatum sibiricum powder according to claim 1, characterized in that, S1, specifically: Obtain online testing data and offline sampling data corresponding to batches of Polygonatum sibiricum raw materials entering the continuous production line; Online testing data and offline sampling data are correlated and aligned according to the collection time and production feeding sequence; Based on the correlated and aligned online detection data and offline sampling data, detection index data characterizing the intrinsic quality status of batches of Polygonatum odoratum raw materials were extracted. The test index data are preprocessed to form a raw material quality characteristic dataset.

3. The production control method for Polygonatum odoratum instant powder according to claim 2, characterized in that, S2, specifically: Based on the raw material quality characteristic dataset, the correlation between characteristic indicators of batches of Polygonatum sibiricum raw materials is identified; Based on the correlation between characteristic indicators, determine the changing trend of the compositional structure characteristics of raw material batches; By combining the changing trends of the composition and structure characteristics of raw material batches, the range of quality fluctuation characteristics is determined, and quality assessment data reflecting the differences in raw materials is generated.

4. The production control method for instant Polygonatum sibiricum powder according to claim 1, characterized in that, S6, specifically: Extract the real-time setpoints of the corresponding process parameters based on the optimal process route decision data; Generate production line process configuration switching instructions based on real-time set values ​​of process parameters; Adjust the current process parameter configuration of the production line according to the production line process configuration switching instruction, and dynamically execute the process route corresponding to the optimal process route decision data.

5. A production control system for Polygonatum odoratum instant powder, used to implement the production control method for Polygonatum odoratum instant powder according to any one of claims 1-4, characterized in that, include: Real-time acquisition module: Acquires real-time quality inspection data of batches of Polygonatum sibiricum raw materials entering the continuous production line, standardizes the real-time quality inspection data, and forms a raw material quality characteristic dataset; Quality assessment module: Based on the raw material quality characteristic dataset, the module analyzes the compositional structure characteristics and quality fluctuation characteristics of batches of Polygonatum odoratum raw materials, and generates quality assessment data that reflects the differences in raw materials. Route selection module: Based on quality assessment data, it selects candidate process routes from the process route library that match the quality characteristics of raw materials, and calculates the product yield prediction data corresponding to the candidate process routes. Performance evaluation module: Based on quality evaluation data, it comprehensively evaluates the retention level of active ingredients and rapid dissolution performance under candidate process routes, and generates process performance evaluation data; Fusion Decision Module: This module integrates and analyzes product yield prediction data with process performance evaluation data to generate optimal process route decision data. Parameter switching module: Dynamically switches the process parameter configuration of the production line based on the optimal process route decision data.

Images