Dendrobium officinale powder release control method

Abstract

The invention provides a dendrobium officinale powder release control method, and relates to the technical field of medicine quality control, and the method comprises the following steps: obtaining and correcting hyperspectral images of dendrobium officinale powder under different temperature gradients; extracting an average spectrum and preprocessing by using an SNV algorithm; determining the polysaccharide content; dividing a calibration set and a test set; extracting a characteristic wavelength by using a CARS algorithm; establishing PLSR models under different temperature gradients; establishing an SO-PLS model by taking the characteristic wavelengths under all the temperature gradients as input; the probability risk that a test set sample exceeds a control limit is evaluated through bootstrap sampling, and whether the sample is released or not is determined. According to the method, temperature control and a hyperspectral technology are combined, so that the prediction precision is remarkably improved; and a probability risk assessment mechanism is introduced, so that the detection efficiency is improved while the quality control is ensured, and the problems of long time consumption and low efficiency of a traditional detection method are effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of drug quality control, and in particular to a release control method for Dendrobium officinale powder. Background Art

[0002] Dendrobium officinale Kimuraet & Migo, belonging to the genus Dendrobium of the Orchidaceae family, is a precious plant with both medicinal and edible value. The polysaccharides, bibenzyls, flavonoids and other active ingredients rich in Dendrobium officinale plants give it a variety of biological activities such as anti-aging, anti-fatigue, regulation of blood sugar, blood pressure, blood lipids, and relief of gastric ulcers. Among them, polysaccharides, as the core active ingredients, are not only the main carrier of pharmacological effects, but also an important basis for product quality control. In the production process of Dendrobium officinale products, the powder is a key intermediate product, and its polysaccharide content must be within the preset control range before it can be released to the subsequent process links for further processing. However, the phenol-sulfuric acid method, the polysaccharide detection method currently commonly used, has problems such as cumbersome operation and long time consumption, which leads to unnecessary stagnation in the production process. Therefore, the development of a rapid and non-destructive detection technology to achieve rapid release control of intermediates has become an urgent need in the industry.

[0003] Near-infrared spectroscopy and near-infrared spectral imaging, with their rapid, safe, and non-destructive nature, have already played a vital role in the quality characterization of agricultural products, foods, and pharmaceuticals. Temperature, a key factor influencing near-infrared spectroscopy, can cause variations in the intensity, width, and position of absorption peaks. To date, no theory precisely describes how temperature affects spectra. In early studies, temperature was often treated as a confounding factor. Researchers have expended considerable effort to correct for the temperature effects on spectral predictions. Existing techniques combine spectra measured at different temperatures to develop temperature compensation models, or reduce prediction errors by incorporating scattering effects and temperature variables into standard models. In fact, if temperature can be precisely controlled, it can be considered a correction tool and a source of information. Precise temperature control can significantly improve the reproducibility of measurement results. Furthermore, the vibrational states and energy level distributions of molecules vary under different temperature gradients. By integrating spectral fingerprints from multiple temperature gradients, a more comprehensive picture of the composition and content of a mixture can be obtained.

[0004] It is important to note that as an indirect measurement method, near-infrared spectroscopy often results in significant errors between the predicted and actual values of active ingredient content. These errors arise from a variety of factors, including sample characteristics such as particle size, moisture content, porosity, and the near-infrared response intensity of the analyte groups, as well as instrument and environmental influences such as dark current and ambient temperature and humidity. Given a limited sample set, building models to accurately predict the presence of certain active ingredients presents significant challenges. In large-scale industrial production, a more practical approach is to establish a relatively accurate model and then, based on the uncertainty of the model's predictions, use a probabilistic assessment, rather than a single prediction, to determine whether the sample is at risk of exceeding control limits. Samples with a probabilistic risk exceeding a preset value can be verified using precision chemical assays; samples with a probabilistic risk below the preset value can be released directly. This approach can significantly shorten testing time while ensuring that risks are manageable.

[0005] Although near-infrared spectroscopy has been applied in medicinal material testing, limited research has focused on the polysaccharide content of Dendrobium officinale powder. Existing methods mostly rely on spectral data at a single temperature, lacking the rich information provided by temperature variations and failing to consider the importance of model prediction uncertainty in actual production decisions. Consequently, existing technologies face challenges in quality control of Dendrobium officinale intermediates, including inefficiency and unscientific decision-making. Summary of the Invention

[0006] In order to solve the problems existing in the prior art, the purpose of the present invention is to provide a release control method for Dendrobium officinale powder that can effectively utilize the information brought by temperature changes and improve detection efficiency.

[0007] To achieve the above object, the present invention is implemented through the following technical solution: a method for controlling the release of Dendrobium officinale powder, comprising the following steps:

[0008] S1: Acquire hyperspectral images of Dendrobium officinale powder under different temperature gradients and perform black and white plate correction;

[0009] S2: Extract the average spectrum of Dendrobium officinale powder and preprocess the average spectrum using the standard normal variable transformation algorithm (SNV);

[0010] S3: Determination of polysaccharide content in Dendrobium officinale powder using the phenol-sulfuric acid method;

[0011] S4: Divide the samples into a calibration set and a test set, and the polysaccharide content range of the test set is included in the calibration set;

[0012] S5: Competitive adaptive reweighted sampling (CARS) algorithm is used to extract characteristic wavelengths related to polysaccharide content from the preprocessed average spectrum;

[0013] S6: Based on the spectral characteristic wavelength and polysaccharide content data under different temperature gradients, partial least squares regression (PLSR) models under different temperature gradients were established respectively;

[0014] S7: Taking the characteristic wavelength of the spectrum under all temperature gradients as input, a sequential orthogonal partial least squares model (SO-PLS) is established;

[0015] S8: Based on the sequential orthogonal partial least squares (SO-PLS) model, the bootstrap sampling method was used to predict the test set samples and calculate the probability risk of the sample polysaccharide content exceeding the control limit to control the release of Dendrobium officinale powder.

[0016] Further: in step S1, a semiconductor temperature control device is used to perform a gradient temperature change operation on the Dendrobium officinale powder, and the temperature gradient range is 0°C to 80°C.

[0017] Furthermore: the semiconductor temperature control device has both heating and cooling functions, and uses a PID controller to control the temperature, and its operating parameters include target temperature, heating / cooling time, sample insulation time and PID value.

[0018] Furthermore: in step S1, the hyperspectral scanning mode is line scanning, the optical mode is diffuse reflection, and the black and white plate correction formula is:

[0019]

[0020] Among them, I cal This is the hyperspectral image of Dendrobium officinale powder after correction, I raw is the original hyperspectral image of the powder, I white is the whiteboard hyperspectral image, I dark It is a blackboard hyperspectral image.

[0021] Furthermore: in step S2, the average spectrum extraction step includes: S21: converting the hyperspectral image into a pseudo-color image; S22: determining the Dendrobium officinale powder area on the pseudo-color image; S23: defining a circle inside the powder area with a threshold equal to the powder area or a threshold less than the powder area as a mask; S24: averaging the spectra of all pixel points in the mask to obtain the average spectrum of the Dendrobium officinale powder.

[0022] Furthermore: in step S4, the sample partitioning adopts a sample set partitioning algorithm based on joint xy distance (SPXY).

[0023] Further: in step S6, the steps of establishing the partial least squares regression (PLSR) model under different temperature gradients include:

[0024] S61: Select the search range for the number of principal components of partial least squares regression;

[0025] S62: The optimal number of principal components is determined based on the minimum root mean square error of validation (RMSECV) using a three-fold cross validation method.

[0026] S63: Establishing a partial least squares regression model using the determined optimal number of principal components and the spectral characteristic wavelengths under the corresponding temperature gradient;

[0027] S64: Through the coefficient of determination (R 2 ) and root mean square error of prediction (RMSEP) were used to evaluate the prediction performance of the partial least squares regression (PLSR) model under different temperature gradients.

[0028] Further: in step S7, the step of establishing the sequential orthogonal partial least squares model (SO-PLS) includes:

[0029] S71: Fitting the polysaccharide content to the spectral characteristic wavelength data at the first temperature by partial least squares regression;

[0030] S72: Calculate the matrix of the spectral characteristic wavelength data at the second temperature and the eigenvector of the first temperature orthogonalized;

[0031] S73: Fit the residuals to the orthogonalized matrix by partial least squares regression;

[0032] S74: Repeat the above steps until the spectral characteristic wavelength data under all temperature gradients are integrated into the model.

[0033] Furthermore, in step S8, the release control steps of Dendrobium officinale powder based on probability risk include:

[0034] S81: Use the bootstrap sampling method to perform multiple random samplings with replacement on the calibration set, and establish a new sequential orthogonal partial least squares model (SO-PLS) model each time;

[0035] S82: Use all bootstrap models to predict each test set sample to obtain a set of polysaccharide content prediction values;

[0036] S83: Based on the preset upper and lower control limits, calculate the proportion of predicted values exceeding the control limits as the probability risk;

[0037] S84: For samples with a probability risk lower than the preset threshold, direct release is allowed; for samples with a probability risk higher than the preset threshold, re-confirmation using the phenol-sulfuric acid method is required;

[0038] The bootstrap sampling times is 200 times, and each time 80%-90% of the samples are retained to re-establish the sequential orthogonal partial least squares model (SO-PLS) model; the preset threshold is 1%-2%.

[0039] Further: in step S3, the specific steps of the phenol-sulfuric acid method are:

[0040] S31: Accurately weigh Dendrobium officinale powder, place it in a conical flask, add water, and heat to reflux;

[0041] S32: After cooling, transfer to a volumetric flask, dilute to volume with water, shake well, and filter; measure the filtrate, add anhydrous ethanol, and refrigerate to precipitate;

[0042] S33: After centrifugation, discard the supernatant, wash the precipitate with ethanol, and centrifuge again; dissolve the precipitate in hot water and dilute to obtain the test solution;

[0043] S34: Measure the test solution, add phenol solution and sulfuric acid, and heat in a water bath; after cooling, measure the absorbance and calculate the polysaccharide content.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] First, this invention overcomes technical biases by transforming temperature from a traditional "interference factor" into an "information source." By acquiring hyperspectral images at varying temperature gradients, it captures the differences in the vibrational states and energy level distributions of molecules at different temperatures. These differences manifest themselves spectrally as variations in the intensity, position, and width of functional group absorption peaks. The SO-PLS model effectively integrates spectral signatures at different temperatures and extracts incremental information, achieving a more comprehensive description of material composition and significantly improving the accuracy of the polysaccharide content prediction model.

[0046] Second, while the traditional phenol-sulfuric acid method for determining polysaccharide content is cumbersome and time-consuming, the present invention's near-infrared hyperspectral method can complete acquisition and analysis within minutes, significantly reducing testing time. More importantly, the present invention incorporates a probabilistic risk assessment mechanism that directly determines sample release based on the uncertainty of the predicted results. This eliminates the need for chemical testing for lower-risk samples, further improving production efficiency and reducing unnecessary downtime during the production process.

[0047] Third, this invention breaks through the limitations of traditional single-predictive value judgments. By using bootstrap sampling to assess the uncertainty of model predictions, this method generates a probability distribution of polysaccharide content for each test sample, and then quantitatively assesses the risk of exceeding control limits. This maximizes production efficiency while ensuring product quality, balancing the relationship between testing efficiency and quality control.

[0048] Fourth, the technical solution proposed in this invention is not only applicable to the detection of polysaccharide content in Dendrobium officinale powder, but can also be expanded to the determination and quality control of active ingredients in other Chinese medicinal materials and natural products. This method can effectively enhance lean production, especially in pharmaceutical processes where strict control of intermediate quality consistency is required, and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a flow chart of a method for controlling the release of Dendrobium officinale powder based on temperature-controlled hyperspectral analysis of the present invention;

[0050] Figure 2 This is a schematic diagram of a temperature-controlled near-infrared hyperspectral imaging device according to an embodiment of the present invention;

[0051] Figure 3 This is a schematic diagram of a pseudo-color image generated at wavelengths of 1598.92 nm, 1398.65 nm, and 1198.49 nm according to an embodiment of the present invention;

[0052] Figure 4 Schematic diagram of a mask image on a pseudo-color image generated at wavelengths of 1598.92 nm, 1398.65 nm, and 1198.49 nm according to an embodiment of the present invention;

[0053] Figure 5 This is a schematic diagram of the average spectrum of a powder according to an embodiment of the present invention;

[0054] Figure 6 (A) is a schematic diagram of an average spectrum curve after SNV preprocessing according to an embodiment of the present invention; Figure 6 (B) is a schematic diagram of a partially enlarged view of the average spectrum after SNV preprocessing according to an embodiment of the present invention;

[0055] Figure 7 (A) is a schematic diagram of the probability distribution of the test set samples numbered 1-9 formed by the average value, the upper and lower control limits, and bootstrap sampling in one embodiment of the present invention; Figure 7 (B) is a schematic diagram of the probability distribution formed by the average value, the upper and lower control limits, and bootstrap sampling of the test set samples numbered 10-18 in one embodiment of the present invention; Figure 7 (C) is a schematic diagram of the probability distribution of the test set samples numbered 19-27 formed by the average value, the upper control limit, the lower control limit, and the bootstrap sampling in one embodiment of the present invention; Figure 7(D) is a schematic diagram of the probability distribution of the test set samples numbered 28-35 formed by the average value, the upper control limit, the lower control limit and the bootstrap sampling in one embodiment of the present invention. DETAILED DESCRIPTION

[0056] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0057] It should be noted that in the description of the present invention, SNV is the English abbreviation of the standard normal variable transformation algorithm, which has the same meaning as the standard normal variable transformation algorithm. Similarly, CARS is the English abbreviation of the competitive adaptive reweighted sampling algorithm, PLSR is the English abbreviation of the partial least squares regression model, SO-PLS is the English abbreviation of the sequential orthogonal partial least squares model, SPXY is the English abbreviation of the sample set partitioning algorithm, RMSECV is the English abbreviation of the root mean square error of verification, R 2 It is the abbreviation of the coefficient of determination, and RMSEP is the abbreviation of the root mean square error of prediction.

[0058] A method for controlling the release of Dendrobium officinale powder comprises the following steps:

[0059] S1: Acquire hyperspectral images of Dendrobium officinale powder under different temperature gradients and perform black and white plate correction;

[0060] S2: Extract the average spectrum of Dendrobium officinale powder and preprocess the average spectrum using the standard normal variable transformation algorithm (SNV);

[0061] S3: Determination of polysaccharide content in Dendrobium officinale powder using the phenol-sulfuric acid method;

[0062] S4: Divide the samples into a calibration set and a test set, and the polysaccharide content range of the test set is included in the calibration set;

[0063] S5: Competitive adaptive reweighted sampling (CARS) algorithm is used to extract characteristic wavelengths related to polysaccharide content from the preprocessed average spectrum;

[0064] S6: Based on the spectral characteristic wavelength and polysaccharide content data under different temperature gradients, partial least squares regression (PLSR) models under different temperature gradients were established respectively;

[0065] S7: Taking the characteristic wavelength of the spectrum under all temperature gradients as input, a sequential orthogonal partial least squares model (SO-PLS) is established;

[0066] S8: Based on the sequential orthogonal partial least squares (SO-PLS) model, the bootstrap sampling method was used to predict the test set samples and calculate the probability risk of the sample polysaccharide content exceeding the control limit to control the release of Dendrobium officinale powder.

[0067] Further: in step S1, a semiconductor temperature control device is used to perform a gradient temperature change operation on the Dendrobium officinale powder, and the temperature gradient range is 0°C to 80°C.

[0068] Furthermore: the semiconductor temperature control device has both heating and cooling functions, and uses a PID controller to control the temperature, and its operating parameters include target temperature, heating / cooling time, sample insulation time and PID value.

[0069] Furthermore: in step S1, the hyperspectral scanning mode is line scanning, the optical mode is diffuse reflection, and the black and white plate correction formula is:

[0070]

[0071] Among them, I cal This is the hyperspectral image of Dendrobium officinale powder after correction, I raw is the original hyperspectral image of the powder, I white is the whiteboard hyperspectral image, I dark It is a blackboard hyperspectral image.

[0072] Furthermore: in step S2, the average spectrum extraction step includes: S21: converting the hyperspectral image into a pseudo-color image; S22: determining the Dendrobium officinale powder area on the pseudo-color image; S23: defining a circle inside the powder area with a threshold equal to the powder area or a threshold less than the powder area as a mask; S24: averaging the spectra of all pixel points in the mask to obtain the average spectrum of the Dendrobium officinale powder.

[0073] Furthermore: in step S4, the sample partitioning adopts a sample set partitioning algorithm based on joint xy distance (SPXY).

[0074] Further: in step S6, the steps of establishing the partial least squares regression (PLSR) model under different temperature gradients include:

[0075] S61: Select the search range for the number of principal components of partial least squares regression;

[0076] S62: The optimal number of principal components is determined based on the minimum root mean square error of validation (RMSECV) using a three-fold cross validation method.

[0077] S63: Establishing a partial least squares regression model using the determined optimal number of principal components and the spectral characteristic wavelengths under the corresponding temperature gradient;

[0078] S64: Through the coefficient of determination (R 2 ) and root mean square error of prediction (RMSEP) were used to evaluate the prediction performance of the partial least squares regression (PLSR) model under different temperature gradients.

[0079] Further: in step S7, the step of establishing the sequential orthogonal partial least squares model (SO-PLS) includes:

[0080] S71: Fitting the polysaccharide content to the spectral characteristic wavelength data at the first temperature by partial least squares regression;

[0081] S72: Calculate the matrix of the spectral characteristic wavelength data at the second temperature and the eigenvector of the first temperature orthogonalized;

[0082] S73: Fit the residuals to the orthogonalized matrix by partial least squares regression;

[0083] S74: Repeat the above steps until the spectral characteristic wavelength data under all temperature gradients are integrated into the model.

[0084] Furthermore, in step S8, the release control steps of Dendrobium officinale powder based on probability risk include:

[0085] S81: Use the bootstrap sampling method to perform multiple random samplings with replacement on the calibration set, and establish a new sequential orthogonal partial least squares model (SO-PLS) model each time;

[0086] S82: Use all bootstrap models to predict each test set sample to obtain a set of polysaccharide content prediction values;

[0087] S83: Based on the preset upper and lower control limits, calculate the proportion of predicted values exceeding the control limits as the probability risk;

[0088] S84: For samples with a probability risk lower than the preset threshold, direct release is allowed; for samples with a probability risk higher than the preset threshold, re-confirmation using the phenol-sulfuric acid method is required;

[0089] The bootstrap sampling times is 200 times, and each time 80%-90% of the samples are retained to re-establish the sequential orthogonal partial least squares model (SO-PLS) model; the preset threshold is 1%-2%.

[0090] Further: in step S3, the specific steps of the phenol-sulfuric acid method are:

[0091] S31: Accurately weigh Dendrobium officinale powder, place it in a conical flask, add water, and heat to reflux;

[0092] S32: After cooling, transfer to a volumetric flask, dilute to volume with water, shake well, and filter; measure the filtrate, add anhydrous ethanol, and refrigerate to precipitate;

[0093] S33: After centrifugation, discard the supernatant, wash the precipitate with ethanol, and centrifuge again; dissolve the precipitate in hot water and dilute to obtain the test solution;

[0094] S34: Measure the test solution, add phenol solution and sulfuric acid, and heat in a water bath; after cooling, measure the absorbance and calculate the polysaccharide content.

[0095] The implementation process and effects of the method of the present invention are described in detail below through specific examples.

[0096] Example 1

[0097] like Figure 1 FIG. 1 is a method for controlling the release of Dendrobium officinale powder according to one embodiment of the present invention, comprising the following steps:

[0098] S1: Acquire hyperspectral images of Dendrobium officinale powder under different temperature gradients and perform black and white plate correction.

[0099] Specifically, a sufficient amount of Dendrobium officinale powder was collected and ground into a 300-mesh fine powder. A series of temperature sampling points were set within the 0°C to 80°C temperature range, and a semiconductor temperature control device was used to perform a gradient temperature change on the powder. After reaching a certain temperature sampling point, the sample was kept warm for a period of time, and its hyperspectral image was scanned. This hyperspectral image was then calibrated with a black and white plate.

[0100] Furthermore, the semiconductor temperature control device should have both heating and cooling functions, and use a PID controller to control the temperature. Its operating parameters include target temperature, heating / cooling time, sample holding time and PID value (P value, I value and D value). To ensure detection efficiency, the heating / cooling time shall not exceed 2 minutes, and the sample holding time shall not exceed 5 minutes. The hyperspectral scanning method is line scanning, and the optical mode is diffuse reflection. The sample is driven by a stepper motor through the lens. Its operating parameters include exposure time, stepper motor speed, start scanning position, end scanning position, distance between lens and sample, imaging resolution, maximum reflection intensity, spectral range, spectral resolution and number of wavelengths. The blackboard used for hyperspectral image correction is a black polytetrafluoroethylene lens cover, and the whiteboard is a white polytetrafluoroethylene board with diffuse reflection treatment on the surface. The correction formula for the black and white board is:

[0101]

[0102] Among them, I cal is the hyperspectral image after powder correction, Iraw is the original hyperspectral image of the powder, I white is the whiteboard hyperspectral image, I dark It is a blackboard hyperspectral image.

[0103] S2: Extract the average spectrum of Dendrobium officinale powder and preprocess the average spectrum using the standard normal variate transformation algorithm (SNV).

[0104] Specifically, to extract the average spectrum of the powder area, the hyperspectral image must first be converted into a pseudo-color image. The grayscale images under the pseudo-color image's R (red) channel, G (green) channel, and B (blue) channel are defined as grayscale images at 1598.92nm, 1398.65nm, and 1198.49nm of the hyperspectral image, respectively. Then, the powder area is visually determined on the pseudo-color image, and a circle within the powder area and slightly smaller than the powder area is defined as a mask. The average spectrum of the powder is obtained by averaging the spectra of all pixels within the mask.

[0105] The SNV algorithm is mainly used to eliminate the effects of solid particle size, surface scattering, and optical path changes on near-infrared diffuse reflectance spectra. The formula for SNV transformation is:

[0106]

[0107] Where x is the average spectrum, x SNV is the average spectrum after SNV processing, m is the number of wavelengths of the average spectrum, i = 1, 2, ..., m.

[0108] S3: The polysaccharide content of Dendrobium officinale powder was determined using the phenol-sulfuric acid method.

[0109] Specifically, the phenol-sulfuric acid method is a common method for determining polysaccharide content. It is based on the hydrolysis of polysaccharides under strong acid conditions to produce monosaccharides. Monosaccharides react with phenol to form colored compounds, and their absorbance is measured by colorimetry to estimate the polysaccharide content.

[0110] The specific steps are:

[0111] S31: Accurately weigh about 0.3 g of Dendrobium officinale powder, place it in a 250 mL conical flask, add 200 mL of water, and heat to reflux on a hot plate for 2 h.

[0112] S32: After cooling, transfer to a 250 mL volumetric flask. Wash the conical flask three times with a small amount of water. Combine the washes in the same volumetric flask, add water to the mark, shake well, and filter through a drying filter. Accurately measure 2 mL of the filtrate and place it in a 15 mL centrifuge tube. Accurately add 10 mL of anhydrous ethanol, shake well, and refrigerate at 4°C for 1 hour.

[0113] S33: Remove the pellet and centrifuge at 4000 rpm for 20 min. Discard the supernatant and wash the pellet twice with 80% ethanol (8 mL each time). Centrifuge at 4000 rpm for 20 min. Discard the supernatant and dissolve the pellet in hot water. Transfer the pellet to a 25 mL volumetric flask, cool, add water to the mark, and shake well to obtain the test solution.

[0114] S34: Accurately measure 1 mL of the test solution and place it in a 10 mL stoppered test tube. Quickly and accurately add 1 mL of 5% phenol solution. Shake well and accurately add 5 mL of sulfuric acid along the wall of the test tube. Shake well and heat in a boiling water bath for 20 minutes. After removing, cool in an ice water bath for 5 minutes. Measure the absorbance at 488 nm. Repeat the measurement three times for each sample. The formula for calculating the polysaccharide content of Dendrobium officinale powder is:

[0115]

[0116] Wherein, c is the polysaccharide concentration of the test solution, and M is the weight of the precisely weighed Dendrobium officinale powder.

[0117] S4: Use the sample set partitioning algorithm based on joint xy distance (SPXY) to divide the samples into a calibration set and a test set.

[0118] Specifically, when calculating the distance between samples, the SPXY algorithm not only considers the feature dimension direction (x vector), but also the true value dimension direction (y vector), which can increase the difference and representativeness between samples.

[0119] S5: The competitive adaptive reweighted sampling (CARS) algorithm is used to extract characteristic wavelengths related to the polysaccharide content from the preprocessed average spectrum.

[0120] Specifically, the CARS algorithm simulates biological evolution, adaptively competing and reweighting spectral wavelengths to efficiently select the most useful features for the model. It is suitable for feature selection tasks in high-dimensional data. Its core is to gradually eliminate relatively unimportant features through iterative optimization, thereby improving the model's predictive performance and interpretability. The CARS algorithm requires setting three hyperparameters: the number of principal components, the number of cross-validation folds, and the number of iterations. The specific process is as follows:

[0121] S51: Using the Monte Carlo sampling method, a certain proportion (usually 80%) of samples are selected from the calibration set each time for modeling, and the remaining samples are used as the validation set for verification to establish the PLS model. Set the number of Monte Carlo sampling times N and calculate the absolute value weight of the regression coefficient of the PLS model during each sampling process. The calculation formula is:

[0122]

[0123] Among them, |b i| is the absolute value of the regression coefficient of the i-th wavelength of the average spectrum after SNV processing, w i is the absolute value weight of the regression coefficient of the i-th wavelength of the average spectrum after SNV processing, and p is the number of wavelengths remaining after each sampling.

[0124] S52: Use the exponentially decaying function (EDF) to remove w i Relatively small wavelength. After the jth Monte Carlo sampling, the wavelength ratio R retained according to EDF j for:

[0125] R j =μe -kj

[0126] Among them, μ and k are two constants, P is the total number of wavelengths in the original spectrum.

[0127] S53: At each sampling, a specific number of wavelengths are selected from the previous sampling using Adaptive Reweighted Sampling (ARS), a PLS model is established, and the Root Mean Square Error of Validation (RMSEV) is calculated on the validation set.

[0128] S54: After N samplings are completed, N groups of candidate characteristic wavelength subsets and corresponding RMSEV values are obtained, and the characteristic wavelength subset corresponding to the minimum RMSEV value is selected as the final result of wavelength screening.

[0129] S6: Based on the spectral characteristic wavelength and polysaccharide content data under different temperature gradients, partial least squares regression (PLSR) models under different temperature gradients were established respectively.

[0130] Specifically, hyperspectral images under different temperature gradients were obtained. After black and white plate correction, average spectrum extraction and preprocessing, and characteristic wavelength extraction, the spectral characteristic wavelengths under different temperature gradients were obtained and used to establish PLS models under different temperature gradients. The PLS model used a three-fold cross-validation method to optimize the hyperparameters, and the hyperparameter was the number of principal components PC1. The root mean square error (RMSE) and the determination coefficient (R) were used to calculate the PLS model. 2 )Evaluate the accuracy of the PLS model.

[0131] The three-fold cross validation further divides the calibration set into a training set and a cross validation set. The RMSE of the training set, cross validation set, and test set are denoted as RMSEC, RMSECV, and RMSEP, respectively. 2 Respectively represented as R c 2 、R cv 2 and R p 2 The hyperparameters of the model are determined according to the principle of minimizing RMSECV, and then the model is fitted using the data. A better model should have a relatively high R p 2 and relatively low RMSEP.

[0132] S7: The spectral characteristic wavelengths under all temperature gradients are used as input to establish a sequential orthogonal partial least squares model (SO-PLS).

[0133] Specifically, the characteristic wavelength of the spectrum under all temperature gradients was used as input to establish the SO-PLS model. The SO-PLS model used three-fold cross validation to optimize the hyperparameters, and the hyperparameter was the number of principal components PC2. RMSE and R 2 The accuracy of the SO-PLS model was evaluated using the same method as the PLS model described above.

[0134] SO-PLS extracts incremental information from the spectral characteristic wavelength data under multiple temperature gradients, thereby gradually improving the prediction performance of the model. This can be summarized as follows:

[0135] S71: Assume that X1 and X2 represent two sets of preprocessed spectral data and Y is their common response. Y is fitted with X1 through PLS regression:

[0136]

[0137] in, is the eigenvector, is the loading and e1 is the residual.

[0138] S72: Calculate X2 and T X1 Orthogonal part:

[0139]

[0140] S73:e1 is regressed by PLS with Perform the fit:

[0141]

[0142] S74: Y is regressed by PLS with and Perform the fit:

[0143]

[0144] If there are more than two sets of spectral characteristic wavelength data involved in modeling, steps S72 to S74 are repeated.

[0145] It should be noted that under different temperature gradients, the vibrational states and energy level distributions of molecules vary, manifesting spectrally as shifts in the intensity, position, and width of functional group absorption peaks. The SO-PLS model can more effectively capture subtle differences between spectral signals measured at different temperatures, extracting incremental information to more comprehensively depict the composition and content of the mixture, thereby improving the accuracy of the polysaccharide content prediction model.

[0146] S8: Based on the SO-PLS model, the bootstrap sampling method was used to predict the test set samples and calculate the probability risk of the sample polysaccharide content exceeding the control limit to control the release of Dendrobium officinale powder.

[0147] Specifically, the calibration set was randomly sampled 200 times with replacement using the bootstrap sampling method, retaining 80%-90% of the samples each time to re-establish the SO-PLS model. After each SO-PLS model training, the polysaccharide content of the test set was re-predicted, ultimately yielding 200 predicted values for each sample in the test set. Then, based on the established upper and lower control limits, the probabilistic risk was calculated to determine whether the powder sample should be released for the next processing step.

[0148] Furthermore, the probability risk is calculated independently for each test set sample. The set of 200 content prediction values is denoted as Pred, and the upper and lower control limits are denoted as L upper and L lower If all elements in the set Pred are in L lower and L upper The probability risk RI of the sample is 0. If there is an element in the set Pred that is less than L lower or greater than L upper Count the number of elements exceeding the upper and lower limits (LN). The probability risk for this sample is LN / 200. Samples with a probability risk below 2% can be released to the next processing step. Otherwise, the polysaccharide content should be re-determined using the phenol-sulfuric acid method.

[0149] It should be noted that the present invention overcomes the technical prejudice that temperature is an interfering factor in near-infrared spectroscopy. Temperature, a key influencing factor in near-infrared spectroscopy, can cause changes in the intensity, width, and position of absorption peaks. To date, no theory exists that precisely explains how temperature affects the spectrum. In early studies, temperature was often treated as an interfering factor. Researchers have expended considerable effort to correct for the impact of temperature effects on spectral prediction. For example, S. Kawano et al. developed a temperature compensation model by combining spectra measured at different temperatures. M. Tarumi et al. reduced prediction errors by incorporating variables such as scattering effects and temperature into the standard model. The present invention utilizes a precise temperature control solution. Precise temperature control significantly improves the repeatability of measurement results. Furthermore, the vibrational states and energy level distributions of molecules vary under different temperature gradients. By integrating spectral fingerprint information from multiple temperature gradients, a more comprehensive picture of the material composition and content of a mixture can be obtained.

[0150] The present invention adopts a method for establishing a quantitative correction model between the spectral characteristic wavelength and the polysaccharide content under the full temperature gradient. The temperature of the Dendrobium officinale powder is precisely adjusted by a semiconductor temperature control device, and then a hyperspectral scanning method is performed to obtain a series of hyperspectral images of the powder under the temperature gradient. After the black and white plate correction, the mask area is determined in the powder area, and the average spectral signal is extracted. After SNV preprocessing and CARS characteristic wavelength extraction, the average spectrum is converted into a spectral characteristic wavelength. The SO-PLS model is established with the spectral characteristic wavelength under different temperature gradients as input and the polysaccharide content as output. Under different temperature gradients, the vibration state and energy level distribution of the material molecules will be different, which is manifested in the spectrum as changes in the intensity, position and width of the functional group absorption peak. The SO-PLS model can more effectively capture the slight differences between the spectral signals measured under different temperature conditions, extract the incremental information therein, and then more comprehensively depict the material composition and content information of the mixture, which is conducive to improving the accuracy of the polysaccharide content prediction model.

[0151] The present invention uses the bootstrap algorithm to measure the uncertainty of the SO-PLS model and generates a set containing multiple polysaccharide content prediction values through repeated sampling. By counting the number of elements in the set that exceed the upper and lower control limits, the probability risk of the current sample can be evaluated. When the probability risk is less than the preset value, the sample can be released to the next process link; otherwise, precise chemical measurements must be performed to confirm again whether its polysaccharide content exceeds the standard. While greatly improving detection efficiency and reducing detection costs, it also ensures the quality consistency of intermediate products in the pharmaceutical process, and can effectively improve the lean level of the Dendrobium officinale related industry.

[0152] The present invention precisely adjusts the temperature of the Dendrobium officinale powder through a semiconductor temperature control device, and then performs a hyperspectral scan to obtain a series of hyperspectral images of the powder under temperature gradients. After black and white plate correction, the mask area is determined in the powder area, and the average spectral signal is extracted. After SNV preprocessing and CARS characteristic wavelength extraction, the average spectrum is converted into a spectral characteristic wavelength. The SO-PLS model is established with the spectral characteristic wavelength under different temperature gradients as input and the polysaccharide content as output. Under different temperature gradients, the vibration state and energy level distribution of the material molecules will be different, which is manifested in the spectrum as changes in the intensity, position and width of the functional group absorption peak. The SO-PLS model can more effectively capture the slight differences between the spectral signals measured under different temperature conditions, extract the incremental information therein, and then more comprehensively depict the material composition and content information of the mixture, which is conducive to improving the accuracy of the polysaccharide content prediction model.

[0153] It should be noted that the spectral detection method in the present invention can not only use a near-infrared hyperspectral imaging system to obtain near-infrared spectral signals, but also use a Fourier near-infrared spectrometer or a Fabry-Perot near-infrared spectrometer to obtain near-infrared spectral signals.

[0154] It should be noted that the SO-PLS algorithm in this invention can also be replaced by the following algorithm: Multi-Block Partial Least Squares (MB-PLS): The MB-PLS algorithm is a statistical method for processing the relationships between multiple groups of data blocks. MB-PLS can extract principal components from data blocks from different sources for comprehensive analysis and establish a relationship model between them to explain the potential connections between the variables represented by different data blocks.

[0155] Example 2

[0156] The algorithm in this example was compiled using Python (v3.9.12; Python Software Foundation, 2022) and MATLAB R2018b. The following steps were included:

[0157] S1: 70 batches of Dendrobium officinale samples were collected, ground into powder and passed through a 300-mesh sieve to obtain the test samples. Figure 2 As shown, the temperature gradient points were set at 10°C and 50°C, and a semiconductor temperature control device (provided by Shenzhen Optoelectronics Technology Co., Ltd.) was used to perform the gradient temperature change operation.

[0158] The samples were packed and compacted in a pure copper mold with a 5 mm deep and 19.05 mm diameter hole. Each sample was initially temperature-controlled to 10°C and then to 50°C. A semiconductor temperature control unit with both heating and cooling functions was used to control the temperature using a PID controller. The operating settings were: 1.5-minute ramp-up / down times, 3-minute sample hold time, P value of 3000, I value of 150, and D value of 0. Hyperspectral scanning was performed using a line scan method with diffuse reflectance as the optical mode. A stepper motor drove the sample past the lens. Operating parameters were: 35 ms exposure time, 1.13 mm / s stepper motor speed, 150 mm starting scan position, 210 mm ending scan position, 16.5 cm lens-to-sample distance, 638 × 512 pixel resolution, a maximum reflection intensity of approximately 11,500, a spectral range of 898.47 nm to 1750.87 nm, a spectral resolution of 1.7 nm, and 512 wavelengths.

[0159] Because the grayscale image at a wavelength of 898.47 nm contained bad pixels, and the spectral curve between 1692.42 nm and 1750.87 nm contained significant noise, these wavelengths were removed. Ultimately, a total of 474 wavelengths in the hyperspectral image between 900.14 nm and 1690.75 nm were retained for subsequent analysis. After performing the black-and-white calibration, the spectra of all pixels in the hyperspectral image were converted from reflection intensity to reflectance, with reflectance ranging from 0 to 1.

[0160] S2: If Figure 3-Figure 4 As shown in the figure, the grayscale images of the hyperspectral image at 1598.92nm, 1398.65nm and 1198.49nm are selected as the R, G and B channels respectively to generate pseudo-color images. Due to the error in the stepper motor stroke, the image resolution of the pseudo-color image is between 786×638 and 829×638. The lower left corner of the pseudo-color image is used as the origin, the pixel coordinate point (340,430) is used as the center of the circle, and 70 pixels are used as the radius to draw a circle. The area inside the circle is used as the mask. Figure 5 As shown, the average spectrum of the powder is obtained by averaging the spectra of all pixels in the mask, as shown in Figure 6 As shown, the average spectrum is preprocessed using the SNV algorithm to obtain the preprocessed average spectrum. Figure 6 (A) is a schematic diagram of an average spectrum curve after SNV preprocessing according to an embodiment of the present invention; Figure 6 (B) is a schematic diagram of a partially enlarged view of the average spectrum after SNV preprocessing according to an embodiment of the present invention. Figure 5-Figure 6 The solid line is the spectrum curve at 10°C, and the dotted line is the spectrum curve at 50°C.

[0161] S3: The polysaccharide content was determined using the phenol-sulfuric acid method. The polysaccharide content of 70 samples ranged from 25.80% to 64.51%, with an average of 51.57% and a standard deviation of 6.62%.

[0162] S4: Using the SPXY algorithm, all samples were divided into a calibration set and a test set in a 1:1 ratio. After this division, the polysaccharide content of the calibration set ranged from 25.80% to 64.51%, with an average of 50.65% and a standard deviation of 7.67%. The polysaccharide content of the test set ranged from 39.50% to 64.24%, with an average of 52.48% and a standard deviation of 5.20%. The polysaccharide content range of the test set was included in the calibration set, ensuring good generalizability of the trained model.

[0163] S5: The CARS algorithm was used with 15 principal components, 3 cross-validation folds, and 10,000 iterations. The CARS algorithm was used to extract characteristic wavelengths from the preprocessed spectra at 10°C and 50°C. The characteristic wavelengths extracted at 10°C included 20 wavelengths, including 1086.79 nm, 1088.46 nm, and 1116.80 nm. The characteristic wavelengths extracted at 50°C included 33 wavelengths, including 900.14 nm, 938.46 nm, and 940.13 nm.

[0164] S6: Based on the characteristic wavelength of the spectrum at 10℃, a PLS model was established. The range of the principal component at 10℃ was set to 1 to 30. The number of principal components was optimized by the grid search method, and the optimal principal component of the three-fold cross-validation was determined to be 8. Based on the characteristic wavelength of the spectrum at 50℃, a PLS model was established. The range of the principal component at 50℃ was set to 1 to 30, and the optimal principal component of the three-fold cross-validation was determined to be 13. On the test set, the PLS model at 50℃ achieved better results than the PLS model at 10℃, R 2 The p value reached 0.850 and the RMSEP value reached 2.01%.

[0165] Table 1 Quantitative results of PLS model at 10℃ and 50℃

[0166]

[0167] S7: The characteristic wavelengths of the spectra obtained at 10°C and 50°C were used as input to establish the SO-PLS model. During the model construction process, the hyperparameters were optimized using the grid method, and the search range of the number of principal components was set to 1 to 30. After three-fold cross-validation, the optimal number of principal components was determined to be 18. The prediction performance of the SO-PLS model on the test set was improved compared to the 50°C PLS model, R 2 The p value increased by 0.045 and the RMSEP decreased by about 0.33%, indicating that the model has certain predictive ability.

[0168] Table 2 Quantitative results of SO-PLS model

[0169]

[0170] S8: Use the bootstrap sampling method to perform 200 random samplings with replacement on the calibration set, and retain 80%-90% of the samples each time to re-establish the SO-PLS model. After each SO-PLS model training is completed, the polysaccharide content of the test set is re-predicted. Finally, each sample in the test set corresponds to 200 content prediction values, and these predictions form a probability distribution. The upper and lower control limits are set according to the distribution range of the polysaccharide content of all samples. The upper control limit is set to the mean + 1.5 × standard deviation, and the lower control limit is set to the mean - 1.5 × standard deviation, which are 61.50% and 41.64% respectively. Figure 7 As shown in the figure, a schematic diagram of the probability distribution of 35 test set samples formed by the mean value, upper and lower control limits and bootstrap sampling is given. Figure 7 (A) is a schematic diagram of the probability distribution of the test set samples numbered 1-9 formed by the average value, the upper and lower control limits, and bootstrap sampling in one embodiment of the present invention; Figure 7 (B) is a schematic diagram of the probability distribution formed by the average value, the upper and lower control limits, and bootstrap sampling of the test set samples numbered 10-18 in one embodiment of the present invention; Figure 7 (C) is a schematic diagram of the probability distribution of the test set samples numbered 19-27 formed by the average value, the upper control limit, the lower control limit, and the bootstrap sampling in one embodiment of the present invention; Figure 7 (D) is a probability distribution diagram of the test set samples numbered 28-35 formed by the average value, the control upper and lower limits, and the bootstrap sampling, in one embodiment of the present invention. Figure 7 The middle control limit is represented by a dashed line, the upper control limit is represented by a dotted line, and the lower control limit is represented by a dotted line.

[0171] Taking the 1st, 2nd, 3rd and 4th test set samples as an example, the 1st sample is 100% within the control range, with a probability risk of 0, and is directly released to the next process link; the 2nd sample has a 27% probability of being within the control range, with a probability risk of 73%, and needs to be re-precision chemically determined, and its actual polysaccharide content is likely to exceed the control limit; the 3rd sample has a 93% probability of being within the control range, with a probability risk of 7%, which is higher than the preset standard of 2%, and also needs to be re-precision chemically determined; the 4th sample has a 99.5% probability of being within the control range, with a probability risk of 0.5%, which is lower than the preset standard of 2%, and is directly released to the next process link.

[0172] It can be seen from the above embodiments that the release control method of Dendrobium officinale powder based on temperature-controlled near-infrared hyperspectroscopy and probabilistic risk proposed in the present invention successfully solves the problems of low efficiency and unscientific decision-making in traditional methods by introducing temperature control, multi-temperature spectral information fusion and probabilistic risk assessment. It not only improves the accuracy of polysaccharide content prediction, but also significantly improves the detection efficiency, while realizing scientific quality risk control, and providing a new technical path for the production quality control of Dendrobium officinale and similar natural medicinal materials.

[0173] The above embodiments are intended only to illustrate the technical concepts and features of the present invention. Their purpose is to enable those skilled in the art to understand the contents of the present invention and implement them accordingly. They are not intended to limit the scope of protection of the present invention. Any equivalent changes or modifications made in accordance with the spirit of the present invention are intended to be covered by the scope of protection of the present invention.

Claims

1. A method for controlling the release of Dendrobium officinale powder, characterized in that: The following steps are involved: S1: Acquire hyperspectral images of Dendrobium officinale powder under different temperature gradients and perform black and white plate correction; S2: Extract the average spectrum of Dendrobium officinale powder and preprocess the average spectrum using the normal variable transformation algorithm; S3: Determination of polysaccharide content in Dendrobium officinale powder using the phenol-sulfuric acid method; S4: Divide the samples into a calibration set and a test set, and the polysaccharide content range of the test set is included in the calibration set; S5: A competitive adaptive reweighted sampling algorithm is used to extract characteristic wavelengths related to polysaccharide content from the preprocessed average spectrum; S6: Based on the spectral characteristic wavelength and polysaccharide content data under different temperature gradients, partial least squares regression models under different temperature gradients were established respectively; S7: Taking the characteristic wavelength of the spectrum under all temperature gradients as input, a sequential orthogonal partial least squares model is established; S8: Based on the sequential orthogonal partial least squares model, the bootstrap sampling method was used to predict the test set samples and calculate the probability risk of the sample polysaccharide content exceeding the control limit to control the release of Dendrobium officinale powder.

2. The method for controlling the release of Dendrobium officinale powder according to claim 1, wherein: In step S1, a semiconductor temperature control device is used to perform a gradient temperature change operation on the Dendrobium officinale powder, and the temperature gradient range is 0°C to 80°C.

3. The method for controlling the release of Dendrobium officinale powder according to claim 2, wherein: The semiconductor temperature control device has both heating and cooling functions and uses a PID controller to control the temperature. Its operating parameters include target temperature, heating time, cooling time, sample insulation time and PID value.

4. The method for controlling the release of Dendrobium officinale powder according to claim 1, wherein: In step S1, the hyperspectral scanning mode is line scanning, the optical mode is diffuse reflection, and the black and white plate correction formula is: Among them, I cal This is the hyperspectral image of Dendrobium officinale powder after correction, I raw is the original hyperspectral image of the powder, I white is the whiteboard hyperspectral image, I dark It is a blackboard hyperspectral image.

5. The method for controlling the release of Dendrobium officinale powder according to claim 1, wherein: In step S2, the average spectrum extraction step includes: S21: converting the hyperspectral image into a pseudo-color image; S22: determining the Dendrobium officinale powder area on the pseudo-color image; S23: defining a circle inside the powder area with a threshold equal to the powder area or a threshold less than the powder area as a mask; S24: averaging the spectra of all pixel points in the mask to obtain the average spectrum of the Dendrobium officinale powder.

6. The method for controlling the release of Dendrobium officinale powder according to claim 1, wherein: In step S4, the sample partitioning adopts a sample set partitioning algorithm based on the joint xy distance.

7. The method for controlling the release of Dendrobium officinale powder according to claim 1, wherein: In step S6, the establishment of the partial least squares regression model under different temperature gradients includes the following steps: S61: Select the search range of the number of principal components for partial least squares regression; S62: Determine the optimal number of principal components based on the principle of minimizing the root mean square error through the three-fold cross validation method; S63: Establishing a partial least squares regression model using the determined optimal number of principal components and the spectral characteristic wavelengths under the corresponding temperature gradient; S64: The prediction performance of the partial least squares regression model under different temperature gradients was evaluated by the coefficient of determination and the root mean square error of prediction.

8. The method for controlling the release of Dendrobium officinale powder according to claim 1, wherein: In step S7, the steps of establishing the sequential orthogonal partial least squares model include: S71: Fitting the polysaccharide content to the spectral characteristic wavelength data at the first temperature by partial least squares regression; S72: Calculate the matrix of the spectral characteristic wavelength data at the second temperature and the eigenvector of the first temperature orthogonalized; S73: Fit the residuals to the orthogonalized matrix by partial least squares regression; S74: Repeat the above steps until the spectral characteristic wavelength data under all temperature gradients are integrated into the model.

9. The method for controlling the release of Dendrobium officinale powder according to claim 1, wherein: In step S8, the release control steps of Dendrobium officinale powder based on probabilistic risk include: S81: Use the bootstrap sampling method to perform multiple random samplings with replacement on the calibration set, and establish a new sequential orthogonal partial least squares model each time; S82: Use all bootstrap models to predict each test set sample to obtain a set of polysaccharide content prediction values; S83: Based on the preset upper and lower control limits, calculate the proportion of predicted values exceeding the control limits as the probability risk; S84: For samples with a probability risk lower than the preset threshold, direct release is allowed; for samples with a probability risk higher than or equal to the preset threshold, re-confirmation using the phenol-sulfuric acid method is required; The bootstrap sampling times is 200 times, and 80%-90% of the samples are retained each time to re-establish the sequential orthogonal partial least squares model; the preset threshold is 1%-2%.

10. A method for controlling the release of Dendrobium officinale powder according to any one of claims 1 to 9, characterized in that: In step S3, the specific steps of the phenol-sulfuric acid method are: S31: Weigh the Dendrobium officinale powder, place it in a conical flask, add water, and heat to reflux; S32: After cooling, transfer to a volumetric flask, dilute to volume with water, shake well, and filter; measure the filtrate, add anhydrous ethanol, and refrigerate to precipitate; S33: After centrifugation, discard the supernatant, wash the precipitate with ethanol, and centrifuge again; dissolve the precipitate in hot water and dilute to obtain the test solution; S34: Measure the test solution, add phenol solution and sulfuric acid, and heat in a water bath; after cooling, measure the absorbance and calculate the polysaccharide content.

Images