Spectroscopy-based Chinese herbal medicine component identification method and system

Abstract

The invention relates to the technical field of data identification, in particular to a spectroscopy-based Chinese herbal medicine component identification method and system. The method comprises the following steps: carrying out difference wavelength characteristic analysis and wavelength calibration treatment on a quantitative Chinese herbal medicine sample to be detected to obtain spectrum calibration data of the Chinese herbal medicine sample to be detected; performing spectral feature recognition analysis and feature vector conversion processing on the spectral calibration data of the to-be-detected Chinese herbal medicine sample to obtain a spectral feature vector of the to-be-detected Chinese herbal medicine sample; performing vector classification processing and feature point set sampling processing on the spectral feature vectors of the Chinese herbal medicine sample to be detected to obtain spectral feature point sets of different components to be detected in the Chinese herbal medicine to be detected; and obtaining a spectral feature point set of known components of the Chinese herbal medicine, and carrying out incidence matrix mapping analysis and component extreme value identification analysis on the spectral feature point set of different to-be-detected components in the to-be-detected Chinese herbal medicine to obtain a component identification result of the to-be-detected Chinese herbal medicine sample. According to the invention, specific components in the Chinese herbal medicine can be identified more accurately and efficiently.

Description

Technical Field

[0001] The present invention relates to the technical field of data recognition, and in particular to a method and system for identifying components of Chinese herbal medicine based on spectroscopy. Background Art

[0002] Traditional Chinese herbal medicines are increasingly used in medicine and health care. However, the ingredients of Chinese herbal medicines are complex and diverse. Traditional methods for identifying Chinese herbal medicine ingredients usually rely on chemical analysis, which is time-consuming, labor-intensive and costly. Spectroscopy, as a fast and non-destructive analytical method, provides a new way to quickly identify Chinese herbal medicine ingredients. However, existing spectroscopic methods still have problems with recognition accuracy and low efficiency in identifying and representing Chinese herbal medicine ingredients. Summary of the invention

[0003] Based on this, it is necessary for the present invention to provide a method and system for identifying Chinese herbal medicine ingredients based on spectroscopy to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for identifying Chinese herbal medicine components based on spectroscopy comprises the following steps:

[0005] Step S1: selecting lasers of different frequencies through a spectral instrument to perform differential wavelength characteristic analysis and integration on a quantitative Chinese herbal medicine sample to be tested, so as to obtain spectral information data of the Chinese herbal medicine sample to be tested; performing spectral wavelength calibration processing on the spectral information data of the Chinese herbal medicine sample to be tested, so as to obtain spectral calibration data of the Chinese herbal medicine sample to be tested;

[0006] Step S2: performing spectral feature recognition analysis on the spectral calibration data of the Chinese herbal medicine sample to be tested to obtain spectral feature information data of the Chinese herbal medicine sample to be tested; performing feature vector conversion processing on the spectral feature information data of the Chinese herbal medicine sample to be tested to obtain a spectral feature vector of the Chinese herbal medicine sample to be tested;

[0007] Step S3: performing vector classification processing on the spectral feature vectors of the Chinese herbal medicine samples to be tested based on the spectral feature information data of the Chinese herbal medicine samples to be tested, so as to obtain spectral feature vectors of different components of the Chinese herbal medicine to be tested; performing feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, so as to obtain spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested;

[0008] Step S4: Obtain a set of spectral feature points of known components of Chinese herbal medicine, and perform association matrix mapping analysis on the spectral feature point sets of different components to be tested in the Chinese herbal medicine and the spectral feature point sets of known components of the Chinese herbal medicine, so as to obtain a spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicine; perform component maximum value recognition analysis based on the spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicine, and obtain the component identification result of the Chinese herbal medicine sample to be tested.

[0009] The present invention firstly selects lasers of different frequencies through a spectral instrument to perform spectral scanning analysis on the quantitative Chinese herbal medicine sample to be tested obtained after grinding treatment. The spectral scanning analysis process can fully understand the optical properties of different components in the sample to provide spectral information of various components. The spectral information of different components is also analyzed by differential wavelength characteristics. This step can effectively distinguish the spectral characteristics of different components in the sample, thereby providing strong support for subsequent data integration and processing. At the same time, by correspondingly integrating the spectral characteristic data of different components with differential wavelengths obtained by analysis, the absorption or reflection conditions of different components in the sample to be tested at different wavelengths are provided, providing the necessary original spectral information data for subsequent analysis. Subsequently, by performing spectral wavelength calibration processing on the spectral information data of the Chinese herbal medicine sample to be tested, possible errors of the spectral instrument can be eliminated to ensure the accuracy and reliability of the data. Spectral wavelength calibration processing is a key step to ensure the accuracy of subsequent analysis results, which helps to eliminate the influence of instrument errors on the results. Secondly, by performing spectral feature recognition and analysis on the spectral calibration data of the Chinese herbal medicine sample to be tested, the spectral feature information data of the sample can be obtained. This step can identify the key features in the sample spectrum by analyzing the characteristic peak value, valley position and other information in the spectral data, which is helpful for the subsequent identification and analysis of the components. In addition, by performing feature vector conversion processing on the spectral feature information data of the Chinese herbal medicine sample to be tested, the spectral feature information can be converted into a mathematical vector form, which is convenient for subsequent data processing and analysis. These processing steps provide basic data and feature vectors for subsequent component identification, thereby laying a foundation for further analysis. Then, by performing vector classification processing on the spectral feature vectors of the Chinese herbal medicine sample to be tested based on the spectral feature information data of the Chinese herbal medicine sample to be tested, the spectral feature vectors of different components can be obtained. This step can distinguish the feature vectors of different components by classifying the spectral feature information of the sample, thereby providing a basis for subsequent component identification. At the same time, by performing feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, representative feature points can be extracted, which is helpful for subsequent association analysis and component identification. Finally, by obtaining the spectral feature point set of known components of Chinese herbal medicine, and performing correlation matrix mapping analysis on the spectral feature point set of different components to be tested in the Chinese herbal medicine and the spectral feature point set of known components of Chinese herbal medicine, the spectral feature correlation matrix between the components to be tested and the known components of Chinese herbal medicine is obtained. This step establishes the relationship between the components to be tested and the known components by performing correlation analysis on the spectral feature points of the components to be tested and the known components, which provides a basis for the identification of the components to be tested. According to the correlation matrix, the component maximum value identification analysis can determine the feature point set with the strongest correlation with the known components in the sample, thereby identifying the components to be tested. This process provides reliable analysis results for the identification of components of Chinese herbal medicine samples, thereby improving the identification efficiency and accuracy of Chinese herbal medicine components.

[0010] Preferably, the present invention further provides a Chinese herbal medicine component identification system based on spectroscopy, which is used to perform the Chinese herbal medicine component identification method based on spectroscopy as described above, and the Chinese herbal medicine component identification system based on spectroscopy comprises:

[0011] The Chinese herbal medicine sample spectrum data acquisition module is used to select lasers of different frequencies through a spectral instrument to perform differential wavelength feature analysis and integration on the quantitative Chinese herbal medicine samples to be tested, so as to obtain the spectrum information data of the Chinese herbal medicine samples to be tested; perform spectrum wavelength calibration processing on the spectrum information data of the Chinese herbal medicine samples to be tested, so as to obtain the spectrum calibration data of the Chinese herbal medicine samples to be tested;

[0012] The Chinese herbal medicine spectrum feature vector conversion module is used to perform spectrum feature recognition and analysis on the spectrum calibration data of the Chinese herbal medicine sample to be tested, and obtain the spectrum feature information data of the Chinese herbal medicine sample to be tested; perform feature vector conversion processing on the spectrum feature information data of the Chinese herbal medicine sample to be tested, so as to obtain the spectrum feature vector of the Chinese herbal medicine sample to be tested;

[0013] The Chinese herbal medicine component spectral feature point set sampling module is used to perform vector classification processing on the spectral feature vectors of the Chinese herbal medicine samples to be tested based on the spectral feature information data of the Chinese herbal medicine samples to be tested, so as to obtain the spectral feature vectors of different components of the Chinese herbal medicine to be tested; perform feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, so as to obtain the spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested;

[0014] The spectral feature association recognition module for Chinese herbal medicine components is used to obtain the spectral feature point set of known components of Chinese herbal medicines, and to perform association matrix mapping processing on the spectral feature point sets of different components to be tested in the Chinese herbal medicines and the spectral feature point sets of known components of Chinese herbal medicines, so as to obtain the spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicines; and to perform component maximum value recognition analysis based on the spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicines, so as to obtain the component identification results of the Chinese herbal medicine samples to be tested.

[0015] In summary, the present invention provides a Chinese herbal medicine ingredient identification system based on spectroscopy, which is composed of a Chinese herbal medicine sample spectral data acquisition module, a Chinese herbal medicine spectral feature vector conversion module, a Chinese herbal medicine ingredient spectral feature point set sampling module and a Chinese herbal medicine ingredient spectral feature association identification module. It can realize any one of the Chinese herbal medicine ingredient identification methods based on spectroscopy described in the present invention, and is used to combine the operations between computer programs running on each module to realize any one of the Chinese herbal medicine ingredient identification methods based on spectroscopy. The internal structures of the system cooperate with each other, which can greatly reduce duplication of work and manpower investment, and can quickly and effectively provide a more accurate and efficient Chinese herbal medicine ingredient identification process, thereby simplifying the operation flow of the Chinese herbal medicine ingredient identification system based on spectroscopy. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0017] Figure 1 The figure is a schematic diagram of the steps of the method for identifying components of Chinese herbal medicine based on spectroscopy of the present invention;

[0018] Figure 2 for Figure 1 Detailed step flow diagram of step S1;

[0019] Figure 3 for Figure 2 Detailed step flow chart of step S15 in FIG. DETAILED DESCRIPTION

[0020] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0021] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0022] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0023] To achieve this, please refer to Figures 1 to 3 The present invention provides a method for identifying Chinese herbal medicine components based on spectroscopy, the method comprising the following steps:

[0024] Step S1: selecting lasers of different frequencies through a spectral instrument to perform differential wavelength characteristic analysis and integration on a quantitative Chinese herbal medicine sample to be tested, so as to obtain spectral information data of the Chinese herbal medicine sample to be tested; performing spectral wavelength calibration processing on the spectral information data of the Chinese herbal medicine sample to be tested, so as to obtain spectral calibration data of the Chinese herbal medicine sample to be tested;

[0025] Step S2: performing spectral feature recognition analysis on the spectral calibration data of the Chinese herbal medicine sample to be tested to obtain spectral feature information data of the Chinese herbal medicine sample to be tested; performing feature vector conversion processing on the spectral feature information data of the Chinese herbal medicine sample to be tested to obtain a spectral feature vector of the Chinese herbal medicine sample to be tested;

[0026] Step S3: performing vector classification processing on the spectral feature vectors of the Chinese herbal medicine samples to be tested based on the spectral feature information data of the Chinese herbal medicine samples to be tested, so as to obtain spectral feature vectors of different components of the Chinese herbal medicine to be tested; performing feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, so as to obtain spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested;

[0027] Step S4: Obtain a set of spectral feature points of known components of Chinese herbal medicine, and perform association matrix mapping analysis on the spectral feature point sets of different components to be tested in the Chinese herbal medicine and the spectral feature point sets of known components of the Chinese herbal medicine, so as to obtain a spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicine; perform component maximum value recognition analysis based on the spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicine, and obtain the component identification result of the Chinese herbal medicine sample to be tested.

[0028] In the embodiment of the present invention, please refer to Figure 1 FIG. 1 is a schematic diagram of the steps of the method for identifying Chinese herbal medicine components based on spectroscopy of the present invention. In this example, the steps of the method for identifying Chinese herbal medicine components based on spectroscopy include:

[0029] Step S1: selecting lasers of different frequencies through a spectral instrument to perform differential wavelength characteristic analysis and integration on a quantitative Chinese herbal medicine sample to be tested, so as to obtain spectral information data of the Chinese herbal medicine sample to be tested; performing spectral wavelength calibration processing on the spectral information data of the Chinese herbal medicine sample to be tested, so as to obtain spectral calibration data of the Chinese herbal medicine sample to be tested;

[0030] The embodiment of the present invention firstly crushes and grinds the Chinese herbal medicine sample to be tested by using a grinding bowl to ensure that the particle size of the Chinese herbal medicine sample is uniform and fine, and performs impurity screening to remove impurities and unnecessary parts in the ground Chinese herbal medicine sample from the sample. At the same time, the Chinese herbal medicine sample to be tested after impurity screening is homogenized by using a mixing and stirring method to ensure that the components in the sample can be evenly distributed, thereby obtaining a quantitative Chinese herbal medicine sample to be tested. Furthermore, by selecting a suitable dye or marker (including chemical dyes of corresponding chemical components, fluorescent markers, etc.) and adding it to the processed quantitative Chinese herbal medicine sample to be tested for dyeing and marking, and ensuring that the dye or marker is combined with the target component of the Chinese herbal medicine sample to be tested, and by using a spectral instrument (such as an ultraviolet-visible spectrophotometer (UV-Vis spectrophotometer) or an infrared spectrometer (FTIR spectrometer)) to select lasers under different frequency conditions to perform spectral scanning on the quantitative Chinese herbal medicine dyeing and marking sample to be tested, so as to obtain spectral information reflecting the absorption, reflection or transmission conditions at different wavelengths, and by using corresponding statistical methods (such as principal component analysis (PCA), partial least squares regression (PLS), spectral correlation analysis, etc.) to perform differential wavelength feature analysis on the spectral data of different components in the sample at different frequencies, so as to compare the spectral data between different components in the sample to find out the wavelength features with significant differences, so that it can distinguish similar components (such as cellulose, sugar, etc.) in the Chinese herbal medicine sample, and by integrating the different wavelength spectral feature data obtained by analysis, the spectral information data of the Chinese herbal medicine sample to be tested is obtained. . Secondly, the pre-processed spectral information data of the Chinese herbal medicine samples to be tested are processed by using corresponding spectral wavelength calibration methods (such as feature matching calibration algorithm, multi-scale calibration optimization algorithm, etc.) to adjust the spectral wavelength information of the Chinese herbal medicine samples to be tested and ensure the accuracy of the wavelength to calibrate the wavelength deviation that occurs during the detection process, and finally obtain the spectral calibration data of the Chinese herbal medicine samples to be tested.

[0031] Step S2: performing spectral feature recognition analysis on the spectral calibration data of the Chinese herbal medicine sample to be tested to obtain spectral feature information data of the Chinese herbal medicine sample to be tested; performing feature vector conversion processing on the spectral feature information data of the Chinese herbal medicine sample to be tested to obtain a spectral feature vector of the Chinese herbal medicine sample to be tested;

[0032] The embodiment of the present invention uses a spectral signal processing method to perform peak detection on the spectral calibration data of the Chinese herbal medicine sample to be tested, so as to identify the characteristics of each peak position in the spectrum of the Chinese herbal medicine sample to be tested, and calculate the intensity of the corresponding absorption peak range, including indicators such as peak height and peak area, and at the same time measure and record the distance characteristics between each adjacent spectral peak and valley in the spectrum of the aggregated Chinese herbal medicine sample to be tested. Secondly, its feature data is merged into an overall feature data set, and it is ensured that the corresponding relationship between different features matches correctly, so as to merge into a more comprehensive and integrated sample spectrum feature information data set. Then, by performing feature vector conversion on the merged spectral feature information data, the spectral feature information is converted into a feature vector form that is more suitable for machine learning and pattern recognition algorithm processing, such as a feature vector representation form of [peak position, absorption peak intensity, peak-to-valley distance, ...], and finally the spectral feature vector of the Chinese herbal medicine sample to be tested is obtained.

[0033] Step S3: performing vector classification processing on the spectral feature vectors of the Chinese herbal medicine samples to be tested based on the spectral feature information data of the Chinese herbal medicine samples to be tested, so as to obtain spectral feature vectors of different components of the Chinese herbal medicine to be tested; performing feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, so as to obtain spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested;

[0034] The embodiment of the present invention analyzes the spectral characteristic information data of the Chinese herbal medicine samples to be tested by using a time series analysis method, so as to organize the spectral characteristic information data of the Chinese herbal medicine samples to be tested according to a time series, so that it can reveal the dynamic characteristics of different components in the Chinese herbal medicine samples to be tested, including the fluctuation of component concentration and the change of spectral shape, and uses mathematical statistical methods or machine learning algorithms to perform pattern recognition and analysis to identify and understand the changing behavior patterns of the spectral characteristics of the Chinese herbal medicine samples to be tested, including the evolution trends of specific peaks, absorption peaks or peak valleys, and further reveal the time series characteristics of different components in the herbal samples. Secondly, by combining the characteristic change behavior patterns of different components in the Chinese herbal medicine samples to be tested, a machine learning classification algorithm (such as a support vector machine, a random forest, etc.) is used to classify the spectral characteristic vectors of the Chinese herbal medicine samples to be tested, so as to identify the spectral characteristic vectors of different components, thereby obtaining the spectral characteristic vectors of different components of the Chinese herbal medicine to be tested. Then, by extracting the key feature points representing different components from the classified spectral feature vectors of different components in the Chinese herbal medicine to be tested, the spectral characteristics of different components in the Chinese herbal medicine sample to be tested can be more prominently presented, and finally a set of spectral feature points of different components to be tested in the Chinese herbal medicine to be tested is obtained.

[0035] Step S4: Obtain a set of spectral feature points of known components of Chinese herbal medicine, and perform association matrix mapping analysis on the spectral feature point sets of different components to be tested in the Chinese herbal medicine and the spectral feature point sets of known components of the Chinese herbal medicine, so as to obtain a spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicine; perform component maximum value recognition analysis based on the spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicine, and obtain the component identification result of the Chinese herbal medicine sample to be tested.

[0036] The embodiment of the present invention first collects samples of known components of Chinese herbal medicines and obtains their spectral information data under experimental conditions, and obtains spectral feature point sets of known components of Chinese herbal medicines by using spectral instruments and other equipment, which include key feature points of known components of Chinese herbal medicines, thereby obtaining spectral feature point sets of known components of Chinese herbal medicines. At the same time, by constructing a suitable spectral feature correlation calculation formula, correlation calculation is performed on the spectral feature point sets of different components to be tested in the Chinese herbal medicines and the spectral feature point sets of known components of Chinese herbal medicines to quantify the similarities or differences between different components to be tested and known components in Chinese herbal medicines, and an initial correlation matrix between the components to be tested and known components of Chinese herbal medicines is constructed in a manner that the components to be tested in Chinese herbal medicines are in horizontal columns and the known components are in vertical columns. Secondly, the initial values ​​of the corresponding matrix areas in the initial correlation matrix are all set to zero, and the calculated spectral feature correlation is compared and judged by using a preset spectral feature correlation threshold. If it is greater than or equal to the preset The spectral feature correlation threshold is set, indicating that the similarity between the spectral feature points of the tested component and the spectral feature points of the known components corresponding to the spectral feature correlation is large, then the value of the corresponding matrix area in the initial correlation matrix is ​​increased by one; if it is less than the preset spectral feature correlation threshold, it means that the similarity between the spectral feature points of the tested component and the spectral feature points of the known components corresponding to the spectral feature correlation is small or irrelevant, then the value of the corresponding matrix area in the initial correlation matrix remains unchanged; after the corresponding spectral feature points are compared and judged, a sparse matrix form with inconsistent values ​​is obtained, thereby obtaining the spectral feature correlation matrix between the tested component and the known components of the Chinese herbal medicine. Then, the maximum value judgment is performed on the values ​​in the horizontal columns of the spectral feature correlation matrix between the tested component and the known components of the Chinese herbal medicine to identify the matrix area with the largest value, and the known components in the vertical columns corresponding to the matrix area are represented as the final component identification result of the tested component, and the same identification process is performed on each component to be tested, and finally the component identification result of the tested Chinese herbal medicine sample is obtained.

[0037] The present invention firstly selects lasers of different frequencies through a spectral instrument to perform spectral scanning analysis on the quantitative Chinese herbal medicine sample to be tested obtained after grinding treatment. The spectral scanning analysis process can fully understand the optical properties of different components in the sample to provide spectral information of various components. The spectral information of different components is also analyzed by differential wavelength characteristics. This step can effectively distinguish the spectral characteristics of different components in the sample, thereby providing strong support for subsequent data integration and processing. At the same time, by correspondingly integrating the spectral characteristic data of different components with differential wavelengths obtained by analysis, the absorption or reflection conditions of different components in the sample to be tested at different wavelengths are provided, providing the necessary original spectral information data for subsequent analysis. Subsequently, by performing spectral wavelength calibration processing on the spectral information data of the Chinese herbal medicine sample to be tested, possible errors of the spectral instrument can be eliminated to ensure the accuracy and reliability of the data. Spectral wavelength calibration processing is a key step to ensure the accuracy of subsequent analysis results, which helps to eliminate the influence of instrument errors on the results. Secondly, by performing spectral feature recognition and analysis on the spectral calibration data of the Chinese herbal medicine sample to be tested, the spectral feature information data of the sample can be obtained. This step can identify the key features in the sample spectrum by analyzing the characteristic peak value, valley position and other information in the spectral data, which is helpful for the subsequent identification and analysis of the components. In addition, by performing feature vector conversion processing on the spectral feature information data of the Chinese herbal medicine sample to be tested, the spectral feature information can be converted into a mathematical vector form, which is convenient for subsequent data processing and analysis. These processing steps provide basic data and feature vectors for subsequent component identification, thereby laying a foundation for further analysis. Then, by performing vector classification processing on the spectral feature vectors of the Chinese herbal medicine sample to be tested based on the spectral feature information data of the Chinese herbal medicine sample to be tested, the spectral feature vectors of different components can be obtained. This step can distinguish the feature vectors of different components by classifying the spectral feature information of the sample, thereby providing a basis for subsequent component identification. At the same time, by performing feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, representative feature points can be extracted, which is helpful for subsequent association analysis and component identification. Finally, by obtaining the spectral feature point set of known components of Chinese herbal medicine, and performing correlation matrix mapping analysis on the spectral feature point set of different components to be tested in the Chinese herbal medicine and the spectral feature point set of known components of Chinese herbal medicine, the spectral feature correlation matrix between the components to be tested and the known components of Chinese herbal medicine is obtained. This step establishes the relationship between the components to be tested and the known components by performing correlation analysis on the spectral feature points of the components to be tested and the known components, which provides a basis for the identification of the components to be tested. According to the correlation matrix, the component maximum value identification analysis can determine the feature point set with the strongest correlation with the known components in the sample, thereby identifying the components to be tested. This process provides reliable analysis results for the identification of components of Chinese herbal medicine samples, thereby improving the identification efficiency and accuracy of Chinese herbal medicine components.

[0038] Preferably, step S1 comprises the following steps:

[0039] Step S10: crushing and grinding the Chinese herbal medicine sample to be tested and sieving impurities to obtain a powdered Chinese herbal medicine sample to be tested; homogenizing the powdered Chinese herbal medicine sample to be tested to obtain a quantitative Chinese herbal medicine sample to be tested;

[0040] Step S11: performing component dyeing and labeling processing on the quantitative Chinese herbal medicine sample to be tested, so as to obtain a quantitative dyeing and labeling sample of the Chinese herbal medicine to be tested;

[0041] Step S12: Select lasers of different frequencies through a spectral instrument to perform spectral scanning analysis on the quantitative Chinese herbal medicine dyed and labeled samples to be tested, and obtain spectral data of different components in the sample at different frequencies; perform differential wavelength characteristic analysis on the spectral data of different components in the sample at different frequencies to obtain differential wavelength spectral information characteristic data of different components in the sample; perform data integration on the differential wavelength spectral information characteristic data of different components in the sample to obtain spectral information data of the Chinese herbal medicine sample to be tested;

[0042] Step S13: performing noise removal processing on the spectrum information data of the Chinese herbal medicine sample to be tested to obtain denoised spectrum information data of the Chinese herbal medicine sample to be tested;

[0043] Step S14: performing background correction processing on the spectral denoising information data of the Chinese herbal medicine sample to be tested, to obtain background correction data of the spectral data of the Chinese herbal medicine sample to be tested;

[0044] Step S15: performing spectral wavelength calibration processing on the spectral background correction data of the Chinese herbal medicine sample to be tested, to obtain spectral calibration data of the Chinese herbal medicine sample to be tested.

[0045] As an embodiment of the present invention, refer to Figure 2 As shown, Figure 1 Detailed step flow diagram of step S1 in FIG. 1 , in this embodiment, step S1 includes the following steps:

[0046] Step S10: crushing and grinding the Chinese herbal medicine sample to be tested and sieving impurities to obtain a powdered Chinese herbal medicine sample to be tested; homogenizing the powdered Chinese herbal medicine sample to be tested to obtain a quantitative Chinese herbal medicine sample to be tested;

[0047] The embodiment of the present invention grinds the Chinese herbal medicine sample to be tested by using a grinding bowl to ensure that the particle size of the Chinese herbal medicine sample is uniform and fine, and removes impurities and unnecessary parts in the ground Chinese herbal medicine sample from the sample through impurity screening, thereby obtaining a powdered Chinese herbal medicine sample to be tested. Then, the powdered Chinese herbal medicine sample to be tested is homogenized by using a mixing and stirring method to ensure that the components in the sample can be evenly distributed, and finally a quantitative Chinese herbal medicine sample to be tested is obtained.

[0048] Step S11: performing component dyeing and labeling processing on the quantitative Chinese herbal medicine sample to be tested, so as to obtain a quantitative dyeing and labeling sample of the Chinese herbal medicine to be tested;

[0049] The embodiment of the present invention selects a suitable dye or marker (including chemical dyes of corresponding chemical components, fluorescent markers, etc.) and adds it to the processed quantitative Chinese herbal medicine sample to be tested for dyeing and marking, and ensures that the dye or marker is combined with the target component of the Chinese herbal medicine sample to be tested, and finally obtains a quantitative dyed and labeled sample of the Chinese herbal medicine to be tested.

[0050] Step S12: Select lasers of different frequencies through a spectral instrument to perform spectral scanning analysis on the quantitative Chinese herbal medicine dyed and labeled samples to be tested, and obtain spectral data of different components in the sample at different frequencies; perform differential wavelength characteristic analysis on the spectral data of different components in the sample at different frequencies to obtain differential wavelength spectral information characteristic data of different components in the sample; perform data integration on the differential wavelength spectral information characteristic data of different components in the sample to obtain spectral information data of the Chinese herbal medicine sample to be tested;

[0051] The embodiment of the present invention uses a spectral instrument (such as a UV-Vis spectrophotometer or an infrared spectrometer (FTIR spectrometer)) to select lasers under different frequency conditions to perform spectral scanning on a quantitative sample of Chinese herbal medicine dyed and labeled to be tested, so as to place the quantitative sample of Chinese herbal medicine dyed and labeled to be tested in a sample chamber, and set the spectral scanning parameters (such as wavelength range, scanning speed, etc.) under different frequency states to collect spectral information, so as to obtain spectral information reflecting the absorption, reflection or transmission conditions under different wavelengths, thereby obtaining spectral data of different components in the sample at different frequencies. Secondly, by using corresponding statistical methods (such as principal component analysis (PCA), partial least squares regression (PLS), spectral correlation analysis, etc.), the spectral data of different components in the sample at different frequencies are analyzed for differential wavelength characteristics, so as to compare the spectral data between different components in the sample to find out the wavelength characteristics with significant differences, so that it can distinguish similar components (such as cellulose, sugar, etc.) in the Chinese herbal medicine sample, thereby obtaining differential wavelength spectral information characteristic data of different components in the sample. Then, by integrating the difference wavelength spectral characteristic data obtained by analysis, the spectral information data of the Chinese herbal medicine sample to be tested is finally obtained.

[0052] Step S13: performing noise removal processing on the spectrum information data of the Chinese herbal medicine sample to be tested to obtain denoised spectrum information data of the Chinese herbal medicine sample to be tested;

[0053] The embodiment of the present invention processes the spectral information data of the Chinese herbal medicine sample to be tested by using a suitable noise removal algorithm (such as smoothing filtering, wavelet denoising, etc.) to achieve the effect of removing noise, and effectively improves the signal-to-noise ratio of the data, reduces unnecessary interference, and at the same time retains the characteristics of the spectral signal as much as possible, and finally obtains the spectral denoised information data of the Chinese herbal medicine sample to be tested.

[0054] Step S14: performing background correction processing on the spectral denoising information data of the Chinese herbal medicine sample to be tested, to obtain background correction data of the spectral data of the Chinese herbal medicine sample to be tested;

[0055] The embodiment of the present invention processes the denoised spectrum denoising information data of the Chinese herbal medicine sample to be tested by using a suitable background correction method (including spectral baseline correction, zero point correction, etc.) to eliminate the background signals that may exist in the spectrum, and further eliminate the non-sample signals in the spectrum, and finally obtain the spectral background correction data of the Chinese herbal medicine sample to be tested.

[0056] Step S15: performing spectral wavelength calibration processing on the spectral background correction data of the Chinese herbal medicine sample to be tested, to obtain spectral calibration data of the Chinese herbal medicine sample to be tested.

[0057] The embodiment of the present invention processes the spectral background correction data of the Chinese herbal medicine sample to be tested after background correction by using corresponding spectral wavelength calibration methods (such as feature matching calibration algorithm, multi-scale calibration optimization algorithm, etc.) to adjust the spectral wavelength information of the Chinese herbal medicine sample to be tested, and ensure the accuracy of the wavelength to calibrate the wavelength deviation that occurs during the detection process, and finally obtain the spectral calibration data of the Chinese herbal medicine sample to be tested.

[0058] The present invention firstly grinds and sieves the impurities of the Chinese herbal medicine sample to be tested. This step can ensure the uniformity and purity of the sample. The Chinese herbal medicine sample can be effectively converted into a powder state by grinding and crushing, which is convenient for subsequent processing and analysis. At the same time, impurities can be removed by impurity sieving to improve the purity and accuracy of the sample, thereby ensuring the reliability of subsequent analysis, and corresponding homogenization treatment is performed to ensure that the components in the sample are evenly distributed. By using a specific dye to perform component dyeing and labeling processing on the quantitative Chinese herbal medicine sample to be tested, specific labeling of different components in the Chinese herbal medicine sample can be achieved, which helps to clarify the position and distribution of different components, and makes the different components in the sample clearly visible, thereby providing key information for subsequent spectral analysis. The component dyeing and labeling processing process can also enhance the visual identification of various components in the Chinese herbal medicine sample, and can also provide a reliable basis for subsequent spectral data acquisition, which helps to accurately capture the characteristics of various components. In addition, by selecting lasers of different frequencies through the spectral instrument to perform spectral scanning analysis on the quantitative Chinese herbal medicine dyeing and labeling samples to be tested, the spectral scanning and analysis process can fully understand the optical properties of the sample to provide spectral information of various components, and by performing differential wavelength feature analysis, this step can effectively distinguish the spectral characteristics of different components in the sample, thereby providing strong support for subsequent data integration and processing. Secondly, by using the spectral instrument to collect spectral data of the dyed and labeled Chinese herbal medicine samples, rich spectral information data can be obtained, which reflects the absorption, reflection or transmission at different wavelengths and provides detailed spectral characteristics for subsequent analysis. Through the high-precision acquisition of the spectral instrument, the optical properties of the Chinese herbal medicine samples can be accurately obtained, providing a reliable basis for subsequent data processing. Then, by performing noise removal processing on the spectral information data of the Chinese herbal medicine samples to be tested, the signal-to-noise ratio of the data can be effectively improved and unnecessary interference can be reduced, which helps to ensure the accuracy of subsequent analysis and makes the information extracted from the spectral data more reliable and accurate. Next, the background correction processing is performed on the spectral denoising information data of the Chinese herbal medicine samples to be tested in order to eliminate the background signals that may exist in the spectrum and further improve the accuracy of the data. This helps to eliminate non-sample signals, making the spectral background correction data of the Chinese herbal medicine samples to be tested purer, and can also enhance the significance of spectral features, which helps to more accurately identify and analyze the components in the Chinese herbal medicine samples. Finally, by performing spectral wavelength calibration processing on the spectral background correction data of the Chinese herbal medicine samples to be tested, the wavelength scales of different spectral data can be guaranteed to be consistent, thereby achieving accurate quantitative and qualitative analysis. The key to this step is that it can improve the comparability of spectral data, thereby providing reliable basic data guarantee for further component analysis and research.

[0059] Preferably, step S15 comprises the following steps:

[0060] Step S151: performing spectrum region segmentation processing on the spectrum background correction data of the Chinese herbal medicine sample to be tested according to a preset time division interval to obtain the spectrum data region of the Chinese herbal medicine sample to be tested;

[0061] Step S152: Calculate the degree of deviation of the spectral wavelength information data in the spectral data region of the Chinese herbal medicine sample to be tested using a spectral wavelength deviation degree calculation formula to obtain a wavelength deviation degree value of the spectral region of the Chinese herbal medicine sample to be tested;

[0062] Step S153: comparing and judging the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested according to a preset wavelength shift degree threshold value; when the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is greater than or equal to the preset wavelength shift degree threshold value, the spectral data region of the Chinese herbal medicine sample to be tested corresponding to the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is marked as a wavelength high-frequency shift region; when the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is less than the preset wavelength shift degree threshold value, the spectral data region of the Chinese herbal medicine sample to be tested corresponding to the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is marked as a wavelength low-frequency shift region;

[0063] Step S154: performing wavelength classification calibration processing on the spectrum wavelength information data in the wavelength high-frequency offset region to obtain the spectrum wavelength high-frequency offset calibration data of the Chinese herbal medicine to be tested;

[0064] Step S155: performing wavelength nonlinear calibration processing on the spectrum wavelength information data in the wavelength low-frequency offset region to obtain wavelength low-frequency offset calibration data of the Chinese herbal medicine spectrum to be tested;

[0065] Step S156: merging the calibration data of the high-frequency wavelength offset of the spectrum of the Chinese herbal medicine to be tested and the calibration data of the low-frequency wavelength offset of the spectrum of the Chinese herbal medicine to be tested to obtain the calibration data of the spectrum of the Chinese herbal medicine sample to be tested.

[0066] As an embodiment of the present invention, refer to Figure 3 As shown, Figure 2 Detailed step flow diagram of step S15 in FIG. 1 , in this embodiment, step S15 includes the following steps:

[0067] Step S151: performing spectrum region segmentation processing on the spectrum background correction data of the Chinese herbal medicine sample to be tested according to a preset time division interval to obtain the spectrum data region of the Chinese herbal medicine sample to be tested;

[0068] The embodiment of the present invention performs time region slicing and segmentation on the spectral background correction data of the Chinese herbal medicine sample to be tested according to preset time division intervals (for example, 5 minutes, 10 minutes, 1 hour, etc.) to divide the spectrum of the Chinese herbal medicine sample to be tested into multiple regions of equal time length, and finally obtain the spectral data region of the Chinese herbal medicine sample to be tested.

[0069] Step S152: Calculate the degree of deviation of the spectral wavelength information data in the spectral data region of the Chinese herbal medicine sample to be tested using a spectral wavelength deviation degree calculation formula to obtain a wavelength deviation degree value of the spectral region of the Chinese herbal medicine sample to be tested;

[0070] The embodiment of the present invention forms a suitable spectral wavelength shift degree calculation formula by combining the minimum spectral wavelength, the maximum spectral wavelength, the spectral wavelength variable parameter, the spectral specific reference wavelength, the spectral wavelength shift influence control index and related parameters of the spectral data region of the Chinese herbal medicine sample to be tested, and calculates the degree of shift of the spectral wavelength information data in the spectral data region of the Chinese herbal medicine sample to be tested, so as to quantify the degree of shift of the spectral wavelength in each spectral data region of the Chinese herbal medicine sample to be tested, and finally obtains the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested. In addition, the spectral wavelength shift degree calculation formula can also use any spectral wavelength shift detection algorithm in the field to replace the process of shift degree calculation, and is not limited to the spectral wavelength shift degree calculation formula.

[0071] Step S153: comparing and judging the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested according to a preset wavelength shift degree threshold value; when the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is greater than or equal to the preset wavelength shift degree threshold value, the spectral data region of the Chinese herbal medicine sample to be tested corresponding to the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is marked as a wavelength high-frequency shift region; when the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is less than the preset wavelength shift degree threshold value, the spectral data region of the Chinese herbal medicine sample to be tested corresponding to the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is marked as a wavelength low-frequency shift region;

[0072] The embodiment of the present invention compares and judges the calculated wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested by using a preset wavelength shift degree threshold. If the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is greater than or equal to the preset wavelength shift degree threshold, it means that there are more abnormal wavelength shifts in the spectral data region of the Chinese herbal medicine sample to be tested corresponding to the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested, and the spectral data region of the Chinese herbal medicine sample to be tested corresponding to the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is marked as a high-frequency wavelength shift region; if the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is less than the preset wavelength shift degree threshold, it means that there are fewer abnormal wavelength shifts in the spectral data region of the Chinese herbal medicine sample to be tested corresponding to the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested, and the spectral data region of the Chinese herbal medicine sample to be tested corresponding to the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is marked as a low-frequency wavelength shift region.

[0073] Step S154: performing wavelength classification calibration processing on the spectrum wavelength information data in the wavelength high-frequency offset region to obtain the spectrum wavelength high-frequency offset calibration data of the Chinese herbal medicine to be tested;

[0074] The embodiment of the present invention first collects the spectral wavelength information data in the divided wavelength high-frequency offset region and inputs it into a pre-set first-level FIFO calibration processing cache queue, and processes it by using a feature matching calibration algorithm (such as matching based on known pattern features, model fitting and other technologies) to reduce or eliminate the impact of high-frequency offset. At the same time, the calibration results are evaluated by using corresponding evaluation and analysis methods (including error analysis, fit evaluation, residual analysis and other methods) to comprehensively evaluate the success and failure of the offset calibration effect, and the spectral wavelength information data identified as failure is input into a pre-set second-level FIFO calibration processing cache queue, and processed by using a multi-scale calibration optimization algorithm (including calibration strategies based on multiple scales, etc.) to decompose the corresponding spectral data into spectral components of different scales, perform wavelength calibration separately, and then fuse the results to improve the globality and locality of the calibration, so that the calibration process can span spectral features of different scales, and then, by using weighted averaging, model fusion and other methods, the first-level calibration success data and the second-level calibration data are combined to integrate the effective results of the two-level calibration, so as to obtain a more comprehensive and accurate spectral calibration result, and finally obtain the spectral wavelength high-frequency offset calibration data of the Chinese herbal medicine to be tested.

[0075] Step S155: performing wavelength nonlinear calibration processing on the spectrum wavelength information data in the wavelength low-frequency offset region to obtain wavelength low-frequency offset calibration data of the Chinese herbal medicine spectrum to be tested;

[0076] The embodiment of the present invention processes the spectral wavelength information data in the low-frequency wavelength offset region by using nonlinear calibration methods such as polynomial fitting and spline interpolation, so as to more accurately calibrate the corresponding wavelength offset by comprehensively considering the spectral nonlinear transformation, and further improve the accuracy and stability of the spectral data, and finally obtain the spectral wavelength low-frequency offset calibration data of the Chinese herbal medicine to be tested.

[0077] Step S156: merging the calibration data of the high-frequency wavelength offset of the spectrum of the Chinese herbal medicine to be tested and the calibration data of the low-frequency wavelength offset of the spectrum of the Chinese herbal medicine to be tested to obtain the calibration data of the spectrum of the Chinese herbal medicine sample to be tested.

[0078] The embodiment of the present invention combines the high-frequency offset calibration data of the spectral wavelength of the Chinese herbal medicine to be tested and the low-frequency offset calibration data of the spectral wavelength of the Chinese herbal medicine to be tested, and performs further data processing on the combined calibration data, such as smoothing and interpolation, to ensure that the combined data is coherent in the entire wavelength range, thereby obtaining comprehensively processed and more accurate spectral calibration information data, and finally obtaining the spectral calibration data of the Chinese herbal medicine sample to be tested.

[0079] The present invention first performs spectral region segmentation processing on the spectral background correction data of the Chinese herbal medicine sample to be tested according to the preset time division interval. In this way, through the precise arrangement of the time division interval, the spectrum of the Chinese herbal medicine sample to be tested can be successfully divided into multiple regions, providing a more refined analysis unit, which is helpful for more accurate research on the wavelength characteristics in different time periods. At the same time, by using a suitable spectral wavelength shift degree calculation formula to calculate the degree of shift of the spectral wavelength information data in the spectral data area of ​​the Chinese herbal medicine sample to be tested, through this calculation, the degree of shift of the spectral wavelength in each area can be successfully obtained, thereby providing a basis for the subsequent determination of the degree of wavelength shift, which is helpful for more accurate identification of wavelength characteristics. Secondly, by comparing and judging the calculated wavelength shift degree values ​​of the spectral area of ​​the Chinese herbal medicine sample to be tested according to the preset wavelength shift degree threshold, this step can successfully divide the spectral area into a wavelength high-frequency shift area and a wavelength low-frequency shift area by setting the wavelength shift degree threshold, so that the effective classification of wavelength characteristics can be achieved, thereby laying a foundation for further processing of different areas in the future. Then, by performing wavelength classification calibration processing on the spectral wavelength information data in the high-frequency wavelength offset region, the wavelength high-frequency offset calibration data can be successfully obtained through this processing, making the wavelength information in this region more accurate and reliable, which helps to improve the resolution and accuracy of the spectral data. Next, by performing wavelength nonlinear calibration processing on the spectral wavelength information data in the low-frequency wavelength offset region, the wavelength low-frequency offset calibration data can be obtained through this processing, which further improves the accuracy and stability of the spectral data and helps to accurately capture the component information in the low-frequency wavelength offset region. Finally, by merging the high-frequency wavelength offset calibration data of the Chinese herbal medicine spectrum to be tested and the low-frequency wavelength offset calibration data of the Chinese herbal medicine spectrum to be tested, the comprehensive spectral calibration data of the Chinese herbal medicine sample to be tested can be successfully obtained, thereby achieving a comprehensive correction of different wavelength characteristics, and providing a high-quality and accurate data basis for the spectral analysis of the subsequent process.

[0080] Preferably, the calculation formula for the degree of spectrum wavelength shift in step S152 is specifically:

[0081]

[0082] Where Δθ is the wavelength shift value of the spectral region of the Chinese herbal medicine sample to be tested, θ min is the minimum spectral wavelength of the spectral data area of ​​the Chinese herbal medicine sample to be tested, θ max is the maximum spectral wavelength of the spectral data area of ​​the Chinese herbal medicine sample to be tested, θ is the spectral wavelength variable parameter of the spectral data area of ​​the Chinese herbal medicine sample to be tested, θ0 is the spectral specific reference wavelength, C is the spectral wavelength shift influence control index, and η is the correction coefficient of the wavelength shift degree value of the spectral area of ​​the Chinese herbal medicine sample to be tested.

[0083] The present invention obtains a spectrum wavelength deviation degree calculation formula by using a specific mathematical model and verifying it, which is used to calculate the deviation degree of spectrum wavelength information data in the spectrum data area of ​​the Chinese herbal medicine sample to be tested. The spectrum wavelength deviation degree calculation formula is a quantitative index obtained by integrating the spectrum wavelength and is used to describe the deviation of the spectrum wavelength in the area. By using the spectrum wavelength range parameter θ min and θ max To represent the minimum and maximum spectral wavelengths of the spectral data area of ​​the Chinese herbal medicine sample to be tested, these two parameters determine the range of the calculated wavelength shift degree. The different wavelength positions within the spectral range are represented by using spectral wavelength variable parameters. The reference benchmark in the calculation is adjusted by using the spectral specific reference wavelength parameter, thereby affecting the calculation result. Then, the influence of the offset is adjusted by using the spectral wavelength offset influence control index, which determines the contribution of the offset to the wavelength offset degree. In addition, the calculation result is adjusted by introducing a correction coefficient to take into account errors or other influencing factors that may occur during the calculation process. This formula integrates the various influencing factors of wavelength offset, and obtains a comprehensive index through integral calculation, which describes the overall wavelength offset in the spectral area of ​​the Chinese herbal medicine sample to be tested. It can also help identify the wavelength offset phenomenon in the spectrum, thereby providing a basis for the subsequent wavelength offset degree threshold judgment. In summary, this formula fully considers the wavelength offset degree value Δθ of the spectral area of ​​the Chinese herbal medicine sample to be tested, the minimum spectral wavelength θ of the spectral data area of ​​the Chinese herbal medicine sample to be tested, and the wavelength offset degree threshold value of the Chinese herbal medicine sample to be tested. min , the maximum spectral wavelength θ of the spectral data area of ​​the Chinese herbal medicine sample to be tested max , the spectral wavelength variable parameter θ of the spectral data area of ​​the Chinese herbal medicine sample to be tested, the spectral specific reference wavelength θ0, the spectral wavelength shift influence control index C, the correction coefficient η of the wavelength shift degree value of the spectral area of ​​the Chinese herbal medicine sample to be tested, and a functional relationship is formed according to the mutual correlation between the wavelength shift degree value Δθ of the spectral area of ​​the Chinese herbal medicine sample to be tested and the above parameters. This formula can realize the calculation process of the degree of deviation of spectral wavelength information data within the spectral data area of ​​the Chinese herbal medicine sample to be tested. At the same time, by introducing the correction coefficient η of the wavelength deviation degree value of the spectral area of ​​the Chinese herbal medicine sample to be tested, it can be adjusted according to the error occurring in the calculation process, thereby improving the accuracy and applicability of the spectral wavelength deviation degree calculation formula.

[0084] Preferably, step S154 includes the following steps:

[0085] Step S1541: inputting the spectrum wavelength information data in the wavelength high-frequency offset region into a preset first-level FIFO calibration processing cache queue to obtain high-frequency offset first-level FIFO calibration data; performing first-level calibration processing on the high-frequency offset first-level FIFO calibration data using a feature matching calibration algorithm to obtain wavelength high-frequency offset first-level calibration data;

[0086] The embodiment of the present invention first collects the spectrum wavelength information data in the divided wavelength high-frequency offset region and inputs it into a preset first-level FIFO calibration processing cache queue, thereby obtaining the high-frequency offset first-level FIFO calibration data. Then, the high-frequency offset first-level FIFO calibration data is processed by using a feature matching calibration algorithm (such as matching based on known pattern features, model fitting, etc.) to reduce or eliminate the impact of the high-frequency offset, and finally obtain the wavelength high-frequency offset first-level calibration data.

[0087] Step S1542: performing calibration effect evaluation and analysis on the wavelength high frequency offset primary calibration data to obtain wavelength high frequency offset primary calibration success data and fluctuation high frequency offset primary calibration failure data;

[0088] The embodiment of the present invention evaluates the wavelength high-frequency offset first-level calibration data by using corresponding evaluation and analysis methods (including error analysis, fit evaluation, residual analysis and other methods) to comprehensively evaluate the success and failure of the offset calibration effect, and finally obtains the wavelength high-frequency offset first-level calibration success data and the fluctuation high-frequency offset first-level calibration failure data, wherein the successful calibration data is regarded as an accurate capture of the spectral characteristics of the Chinese herbal medicine to be tested, while the failed calibration data may indicate the direction that needs further optimization.

[0089] Step S1543: inputting the fluctuation high frequency offset primary calibration failure data into the preset secondary FIFO calibration processing cache queue to obtain the high frequency offset secondary FIFO calibration data; performing secondary calibration processing on the high frequency offset secondary FIFO calibration data using a multi-scale calibration optimization algorithm to obtain the wavelength high frequency offset secondary calibration data;

[0090] The embodiment of the present invention first inputs the fluctuation high-frequency offset primary calibration failure data identified as failed into a preset secondary FIFO calibration processing cache queue, thereby obtaining the high-frequency offset secondary FIFO calibration data. Then, the high-frequency offset secondary FIFO calibration data is processed by using a multi-scale calibration optimization algorithm (including calibration strategies based on multiple scales, etc.) to decompose the corresponding spectral data into spectral components of different scales, perform wavelength calibration respectively, and then fuse the results to improve the globality and locality of the calibration, so that the calibration process can cross spectral features of different scales, and finally obtain the wavelength high-frequency offset secondary calibration data.

[0091] Step S1544: fusing the wavelength high frequency offset first-level calibration success data and the wavelength high frequency offset second-level calibration data to obtain the wavelength high frequency offset calibration data of the Chinese herbal medicine spectrum to be tested.

[0092] The embodiment of the present invention combines the successful data of the first-level calibration of wavelength high-frequency offset and the second-level calibration data of wavelength high-frequency offset by using methods such as weighted averaging and model fusion to integrate the effective results of the two-level calibration, thereby obtaining a more comprehensive and accurate spectral calibration result, and finally obtaining the wavelength high-frequency offset calibration data of the Chinese herbal medicine spectrum to be tested.

[0093] The present invention firstly inputs the spectral wavelength information data in the wavelength high-frequency offset region into the first-level FIFO calibration processing cache queue, and this cache queue is intended to effectively manage the data to be calibrated, thereby realizing a more efficient processing flow. The first-level FIFO data to be calibrated is also processed by adopting a feature matching calibration algorithm, and the algorithm can effectively identify and calibrate the high-frequency offset feature, thereby obtaining the wavelength high-frequency offset first-level calibration data after the first-level calibration process. The key to this step is that the feature of the wavelength high-frequency offset can be accurately captured through the precise application of the feature matching algorithm, providing a reliable basis for subsequent calibration. Secondly, when the calibration effect evaluation and analysis is performed on the wavelength high-frequency offset first-level calibration data, the success and failure of the calibration can be comprehensively evaluated. This evaluation process is not just a simple result record, but through in-depth analysis of the changes in the data before and after the calibration, the effectiveness of the calibration can be accurately judged. The successful calibration data is regarded as an accurate capture of the spectral features of the Chinese herbal medicine to be tested, and the failed calibration data may indicate the direction that needs to be further optimized. This evaluation analysis provides a deep understanding of the calibration process for subsequent analysis, which is helpful to optimize the calibration algorithm and subsequent processing. Then, once the data that failed the calibration is identified, it is input into the second-level FIFO calibration processing cache queue. In addition, the secondary FIFO data to be calibrated is processed by adopting a multi-scale calibration optimization algorithm to more efficiently solve the high-frequency offset characteristics. The use of this algorithm, combined with a multi-scale calibration method, can more comprehensively consider various situations of wavelength offset and provide more refined calibration results. Therefore, through this step, the secondary calibration data of wavelength high-frequency offset can be obtained, which provides a more reliable basis for further calibration. Finally, by fusing the successful data of the first-level calibration of wavelength high-frequency offset and the secondary calibration data of wavelength high-frequency offset, this fusion process aims to integrate the effective results of the two-level calibration, so as to obtain the final calibration data of the wavelength high-frequency offset of the Chinese herbal medicine spectrum to be tested. By comprehensively considering the results of the two-level calibration, a more comprehensive and accurate spectral calibration result can be obtained, thereby providing more reliable data support for subsequent spectral analysis.

[0094] Preferably, step S2 comprises the following steps:

[0095] Step S21: performing spectral peak recognition analysis on the spectral calibration data of the Chinese herbal medicine sample to be tested, and obtaining a spectral peak position feature data set of the Chinese herbal medicine sample to be tested;

[0096] The embodiment of the present invention uses a spectral signal processing method to perform peak detection on the spectral calibration data of the Chinese herbal medicine sample to be tested, so as to identify the various peaks in the spectrum of the Chinese herbal medicine sample to be tested and determine their specific positions, and at the same time effectively aggregate the identified spectral peak positions of the Chinese herbal medicine sample to be tested into a feature set, and finally obtain a spectral peak position feature data set of the Chinese herbal medicine sample to be tested.

[0097] Step S22: performing absorption peak intensity detection on the spectral calibration data of the Chinese herbal medicine sample to be tested, and obtaining a characteristic data set of absorption peak intensity of the spectral calibration data of the Chinese herbal medicine sample to be tested;

[0098] The embodiment of the present invention first identifies the specific position of the absorption peak in the spectrum of the Chinese herbal medicine sample to be tested through a spectral signal processing method, and then defines the range of the absorption peak around the identified absorption peak position and calculates the intensity of the absorption peak range, including peak height, peak area and other indicators. At the same time, the intensity characteristics of each absorption peak in the spectrum of the aggregated Chinese herbal medicine sample to be tested are measured and recorded, and finally a data set of absorption peak intensity characteristics of the spectrum of the Chinese herbal medicine sample to be tested is obtained.

[0099] Step S23: performing spectral peak-to-valley distance detection on the spectral calibration data of the Chinese herbal medicine sample to be tested, and obtaining a spectral peak-to-valley distance characteristic data set of the Chinese herbal medicine sample to be tested;

[0100] The embodiment of the present invention first identifies the specific positions of peaks and valleys in the spectrum of the Chinese herbal medicine sample to be tested by using a spectral signal processing method, and then detects the peak-to-valley distance of the spectrum between the identified peak-to-valley positions, that is, the distance between two adjacent spectral peaks and valleys, while simultaneously measuring and recording the distance characteristics between each peak and valley in the spectrum of the Chinese herbal medicine sample to be tested, and finally obtains a peak-to-valley spacing characteristic data set of the Chinese herbal medicine sample spectrum to be tested.

[0101] Step S24: performing feature merging processing on the characteristic data set of peak position of the spectrum of the Chinese herbal medicine sample to be tested, the characteristic data set of absorption peak intensity of the spectrum of the Chinese herbal medicine sample to be tested, and the characteristic data set of peak-to-valley spacing of the spectrum of the Chinese herbal medicine sample to be tested, to obtain characteristic information data of the spectrum of the Chinese herbal medicine sample to be tested;

[0102] The embodiment of the present invention combines the spectral peak position characteristic data set of the Chinese herbal medicine sample to be tested, the spectral absorption peak intensity characteristic data set of the Chinese herbal medicine sample to be tested, and the spectral peak-to-valley spacing characteristic data set of the Chinese herbal medicine sample to be tested into an overall characteristic data set, and ensures that the correspondence between different features matches correctly, so as to merge them into a more comprehensive and integrated sample spectral characteristic information data set, and finally obtain the spectral characteristic information data of the Chinese herbal medicine sample to be tested.

[0103] Step S25: performing feature vector conversion processing on the spectral feature information data of the Chinese herbal medicine sample to be tested, and obtaining the spectral feature vector of the Chinese herbal medicine sample to be tested.

[0104] The embodiment of the present invention performs feature vector conversion on the merged spectral feature information data of the Chinese herbal medicine sample to be tested, so as to convert the spectral feature information into a feature vector form that is more suitable for machine learning and pattern recognition algorithm processing, such as a feature vector representation of [peak position, absorption peak intensity, peak-to-valley distance, ...], and finally obtains the spectral feature vector of the Chinese herbal medicine sample to be tested.

[0105] The present invention firstly performs spectral peak recognition analysis on the spectral calibration data of the Chinese herbal medicine sample to be tested, and can effectively identify the peak position characteristics in the spectrum of the Chinese herbal medicine sample to be tested. This analysis process helps to determine the peak with significant properties in the spectrum, and provides important information about the molecular structure and chemical composition of the sample. By accurately identifying the peak position, a spectral feature data set of the sample can be established, thereby providing basic data guarantee for subsequent analysis and processing. Secondly, by performing absorption peak intensity detection on the spectral calibration data of the Chinese herbal medicine sample to be tested, the intensity characteristics of each absorption peak in the spectrum can be accurately measured and recorded, and these characteristics reflect the concentration and relative abundance of various compounds in the sample. By detecting the absorption peak intensity, the absorption characteristics of the spectrum of the Chinese herbal medicine sample to be tested can be obtained, thereby providing important data support for further analysis. Then, by performing spectral peak-to-valley distance detection on the spectral calibration data of the Chinese herbal medicine sample to be tested, the distance characteristics between each peak and valley in the spectrum can be measured and recorded, and these distance characteristics reflect the interaction and distribution between different compounds in the sample, thereby providing important information on the internal structure and composition of the sample for the subsequent processing process, which helps to further understand the characteristics and composition of the sample. Next, by merging the peak position, absorption peak intensity, and peak-to-valley spacing characteristic data sets of the Chinese herbal medicine sample spectra to be tested, a more comprehensive and integrated sample spectral characteristic information data set can be obtained. This comprehensive processing process helps to better understand the spectral characteristics of the samples and provides a more reliable data basis for subsequent analysis and processing. Finally, by performing feature vector conversion processing on the spectral characteristic information data of the Chinese herbal medicine sample to be tested, this processing step aims to convert the spectral characteristic information into a feature vector form that is more suitable for machine learning and pattern recognition algorithm processing. Through the conversion of feature vectors, sample spectral data can be used more effectively, and more accurate classification, identification, and predictive analysis can be performed. The implementation of this step provides necessary preparation for further data processing and application.

[0106] Preferably, step S3 comprises the following steps:

[0107] Step S31: performing time series analysis on the spectral characteristic information data of the Chinese herbal medicine sample to be tested, and obtaining the spectral characteristic time series information data of the Chinese herbal medicine sample to be tested;

[0108] The embodiment of the present invention analyzes the spectral characteristic information data of the Chinese herbal medicine samples to be tested by using a time series analysis method, so as to organize the spectral characteristic information data of the Chinese herbal medicine samples to be tested according to a time series, and ensure that each time point corresponds to the correct spectral feature, so as to reveal the dynamic characteristics of different components in the Chinese herbal medicine samples to be tested, including the fluctuation of component concentration and the change of spectral shape, and finally obtain the spectral characteristic time series information data of the Chinese herbal medicine samples to be tested.

[0109] Step S32: performing characteristic change pattern analysis on the time series information data of the spectrum characteristics of the Chinese herbal medicine sample to be tested, and obtaining the behavior pattern data of the spectrum characteristics change of the Chinese herbal medicine sample to be tested;

[0110] The embodiment of the present invention uses mathematical statistical methods or machine learning algorithms to perform pattern recognition and analysis on the spectral characteristic time series information data of the Chinese herbal medicine samples to be tested, so as to identify and understand the changing behavior patterns of the spectral characteristics of the Chinese herbal medicine samples to be tested, including the evolution trends of specific peaks, absorption peaks or valleys, and further reveal the time series characteristics of different components in the herbal samples, and finally obtain the spectral characteristic changing behavior pattern data of the Chinese herbal medicine samples to be tested.

[0111] Step S33: performing vector classification processing on the spectral feature vector of the Chinese herbal medicine sample to be tested based on the spectral feature change behavior pattern data of the Chinese herbal medicine sample to be tested, so as to obtain spectral feature vectors of different components of the Chinese herbal medicine to be tested;

[0112] The embodiment of the present invention classifies the spectral feature vectors of the Chinese herbal medicine samples to be tested by combining the feature change behavior patterns of different components in the spectral feature change behavior pattern data of the Chinese herbal medicine samples to be tested, so as to identify the spectral feature vectors of different components and finally obtain the spectral feature vectors of different components of the Chinese herbal medicine to be tested.

[0113] Step S34: performing feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, and obtaining spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested.

[0114] The embodiment of the present invention extracts key feature points representing different components from the spectral feature vectors of different components in the classified Chinese herbal medicine to be tested, so that the spectral characteristics of different components inside the Chinese herbal medicine sample to be tested can be more prominently presented, and finally a set of spectral feature points of different components to be tested in the Chinese herbal medicine to be tested is obtained.

[0115] The present invention firstly performs time series analysis on the spectral characteristic information data of the tested Chinese herbal medicine samples, and can obtain the changes of the spectral characteristics over time, which helps to reveal the dynamic characteristics in the sample, including the fluctuation of the component concentration and the change of the spectral shape. Through the in-depth analysis of the time series information, it is possible to obtain useful insights into the evolution law of the spectrum of the herbal sample, thereby providing a basis for the subsequent characteristic change pattern analysis. Secondly, by performing characteristic change pattern analysis on the spectral characteristic time series information data of the tested Chinese herbal medicine samples, it is possible to identify and understand the change behavior pattern of the spectral characteristics of the tested Chinese herbal medicine samples. This step helps to discover the evolution trend of specific peaks, absorption peaks or peak valleys, and further reveal the time series characteristics of different components in the herbal sample. Through in-depth research on the characteristic change pattern, more accurate pattern information can be provided for the subsequent vector classification processing, thereby providing more detailed feature characterization for the analysis of Chinese herbal medicine samples. Then, by performing vector classification processing on the spectral feature vectors of the Chinese herbal medicine samples to be tested based on the spectral feature change behavior pattern data of the Chinese herbal medicine samples to be tested, the spectral feature vectors of the Chinese herbal medicine samples to be tested can be classified more accurately, thereby identifying the spectral features of different components. The implementation of this step provides a powerful tool for in-depth understanding of the spectral characteristics of each component in the sample, which helps to analyze and identify different components more accurately. Finally, by performing feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, the key feature point sets of different components in the spectrum can be extracted, and the spectral characteristics of different components inside the sample can be more prominently presented, which helps to refine the analysis of the internal structure of the sample, thereby providing important support for a more comprehensive description of the spectral characteristics of different components of the Chinese herbal medicine to be tested. The feature point set sampling processing step can also further improve the spectral resolution capability of different components, thereby providing a useful data basis for further research and application.

[0116] Preferably, step S4 comprises the following steps:

[0117] Step S41: obtaining a set of spectral feature points of known components of Chinese herbal medicine;

[0118] The embodiment of the present invention first collects samples of known components of Chinese herbal medicine and obtains their spectral information data under experimental conditions. Then, a spectral feature point set of the known components of Chinese herbal medicine is obtained by using a spectral instrument and other equipment, which includes key feature points of the known components of Chinese herbal medicine as a reference for correlation calculation with the components to be tested, and finally a spectral feature point set of the known components of Chinese herbal medicine is obtained.

[0119] Step S42: using a spectral feature correlation calculation formula to perform correlation calculation on the spectral feature point sets of different components to be tested in the Chinese herbal medicine and the spectral feature point sets of known components of the Chinese herbal medicine, so as to obtain the spectral feature correlation between the spectral feature points of the components to be tested and the spectral feature points of the known components;

[0120] The embodiment of the present invention forms a suitable spectral feature correlation calculation formula by combining the number parameters of the spectral feature points of the components to be tested, the number parameters of the spectral feature points of the known components, the minimum wavelength within the spectral range, the maximum wavelength within the spectral range, the integral wavelength variable parameter within the spectral range, the spectral intensity parameter, the spectral intensity distribution mean, the spectral intensity distribution standard deviation, the exponential function, the wavelength distribution mean within the spectral range, the wavelength distribution standard deviation within the spectral range and related parameters to perform correlation calculation on the spectral feature point set of different components to be tested in the Chinese herbal medicine to be tested and the spectral feature point set of known components of the Chinese herbal medicine to quantify the similarity or difference between different components to be tested and the known components in the Chinese herbal medicine, and finally obtain the spectral feature correlation between the spectral feature points of the components to be tested and the spectral feature points of the known components. In addition, the spectral feature correlation calculation formula can also use any correlation degree detection algorithm in the field to replace the correlation calculation process, and is not limited to the spectral feature correlation calculation formula.

[0121] Step S43: constructing an initial correlation matrix between the components to be tested and the known components of the Chinese herbal medicine, wherein the initial values ​​of the corresponding matrix areas in the initial correlation matrix are all zero, and comparing and judging the spectral feature correlation between the spectral feature points of the components to be tested and the spectral feature points of the known components according to a preset spectral feature correlation threshold, if the spectral feature correlation between the spectral feature points of the components to be tested and the spectral feature points of the known components is greater than or equal to the preset spectral feature correlation threshold, the value of the corresponding matrix area in the initial correlation matrix is ​​increased by one; if the spectral feature correlation between the spectral feature points of the components to be tested and the spectral feature points of the known components is less than the preset spectral feature correlation threshold, the value of the corresponding matrix area in the initial correlation matrix remains unchanged, so as to obtain the spectral feature correlation matrix between the components to be tested and the known components of the Chinese herbal medicine;

[0122] In the embodiment of the present invention, an initial correlation matrix between the components to be tested and the known components of the Chinese herbal medicine is constructed in a manner that the components to be tested are arranged in horizontal columns and the known components are arranged in vertical columns, and the initial values ​​of the corresponding matrix areas in the initial correlation matrix are all set to zero. At the same time, a spectral feature correlation degree between the spectral feature points of the components to be tested and the spectral feature points of the known components is compared and judged by using a preset spectral feature correlation threshold. If the spectral feature correlation degree between the spectral feature points of the components to be tested and the spectral feature points of the known components is greater than or equal to the preset spectral feature correlation threshold, it means that the spectral feature points of the components to be tested corresponding to the spectral feature correlation degree are more similar to the spectral feature points of the known components. Then the initial correlation matrix is ​​set to zero. The value of the corresponding matrix area in the correlation matrix is ​​increased by one; if the spectral feature correlation between the spectral feature point of the component to be tested and the spectral feature point of the known component is less than the preset spectral feature correlation threshold, it means that the spectral feature point of the component to be tested corresponding to the spectral feature correlation is small in similarity or irrelevant to the spectral feature point of the known component, and the value of the corresponding matrix area in the initial correlation matrix remains unchanged; after the spectral feature point set of different components to be tested in the Chinese herbal medicine to be tested and the corresponding spectral feature points in the spectral feature point set of the known components of the Chinese herbal medicine are compared and judged, a sparse matrix form with inconsistent values ​​is obtained, and finally the spectral feature correlation matrix between the components to be tested and the known components of the Chinese herbal medicine is obtained.

[0123] Step S44: performing component maximum value identification analysis based on the spectral feature correlation matrix between the components to be tested and the known components of the Chinese herbal medicine to obtain the component identification result of the Chinese herbal medicine sample to be tested.

[0124] The embodiment of the present invention performs maximum value judgment on the values ​​in the horizontal columns of the spectral feature correlation matrix between the components to be tested and the known components of the Chinese herbal medicine to identify the matrix area with the largest value, and represents the known components in the vertical columns corresponding to the matrix area as the final component identification result of the component to be tested. At the same time, the same identification processing process is performed on each component to be tested, and finally the component identification result of the Chinese herbal medicine sample to be tested is obtained.

[0125] The present invention first obtains the spectral feature point set of known components of Chinese herbal medicine in order to establish benchmark data, which contains the key feature points of known components and can be used as a reference for the components to be tested. The key to this step is to ensure that the spectral feature point set of known components is representative to improve the accuracy of subsequent association calculations. Secondly, by using a suitable spectral feature correlation calculation formula to perform association calculations on the spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested and the spectral feature point sets of known components of the Chinese herbal medicine, the spectral feature correlation is calculated, which helps to quantify the similarity or difference between different components and provides a basis for the subsequent association matrix construction. Accurate correlation calculation is an important step to ensure the accuracy of component identification. Then, by constructing the initial association matrix between the components to be tested in the Chinese herbal medicine and the known components, and comparing and judging the spectral feature correlation according to the preset association threshold, the key to this step is to determine the appropriate association threshold to filter out irrelevant or low-similarity spectral feature points, thereby constructing a more accurate association matrix. Finally, component maximum value identification analysis is performed based on the association matrix. By analyzing the values ​​in the association matrix, the set of feature points with the strongest correlation with the known components in the Chinese herbal medicine sample to be tested can be determined, thereby identifying the components to be tested. The key to this step is to ensure the accurate identification of the components to be tested through an in-depth interpretation of the association matrix, thereby providing reliable results for the identification and analysis process of the components in the Chinese herbal medicine to be tested.

[0126] Preferably, the spectral feature correlation calculation formula in step S42 is specifically:

[0127]

[0128] In the formula, R ij is the spectral feature correlation between the spectral feature point of the i-th component to be tested and the spectral feature point of the j-th known component, n is the total number of spectral feature points of the components to be tested in the spectral feature point set of different components to be tested in the Chinese herbal medicine to be tested, i is the index item of the spectral feature point of the component to be tested, m is the total number of spectral feature points of the known components in the spectral feature point set of the known components of the Chinese herbal medicine, j is the index item of the spectral feature point of the known components, and λ min is the minimum wavelength in the spectral range, λ max is the maximum wavelength in the spectral range, λ is the integral wavelength variable parameter in the spectral range, I i (λ) is the spectral intensity of the spectral characteristic point of the i-th component to be measured at the spectral wavelength λ, is the mean value of the spectral intensity distribution of the spectral characteristic point of the i-th component to be measured, σ i is the standard deviation of the spectral intensity distribution of the uth spectral characteristic point of the component to be measured, I j (λ) is the spectral intensity of the jth known component spectral feature point at the spectral wavelength λ, is the mean value of the spectral intensity distribution of the jth known component spectral feature point, σ j is the standard deviation of the spectral intensity distribution of the jth known component spectral feature point, exp is the exponential function, λ0 is the mean value of the wavelength distribution within the spectral range, Δλ is the standard deviation of the wavelength distribution within the spectral range, and ε is the correction value of the spectral feature correlation.

[0129] The present invention obtains a spectral feature correlation calculation formula by using a specific mathematical model and verifying it, which is used for correlation calculation of spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested and spectral feature point sets of known components of the Chinese herbal medicine. The spectral feature correlation calculation formula is obtained by comparing the spectral intensity distribution of the spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested and the spectral feature point sets of known components of the Chinese herbal medicine in the entire spectral range, and is calculated by weights of an exponential function and a Gaussian distribution. The value range of the correlation is between 0 and 1, and the closer to 1, the higher the similarity between the two spectral feature points. The range and complexity of the correlation calculation are determined by using a quantitative parameter. The scope of the integral is limited by using a spectral range limitation parameter, and it is ensured that only spectral information within a specific wavelength range is considered. The spectral performance of different components at different wavelengths is compared by using the spectral intensity of the spectral feature points of the components to be tested and known. The spectral intensity distribution mean is used to normalize the spectral intensity so as to better compare their relative changes. Secondly, the degree of change of the spectral intensity distribution is measured by using the standard deviation of the spectral intensity distribution, and the shape of the Gaussian distribution is adjusted by using the two parameters of the wavelength distribution mean and standard deviation, thereby affecting the calculation of the spectral feature correlation. In addition, the calculation results are adjusted by introducing correction values ​​to consider possible errors or other influencing factors. This spectral feature correlation calculation formula integrates the spectral information at different wavelength positions, and obtains the correlation between components through weight calculation and integration, providing a basis for further analysis. In this way, the similarity between the component to be tested and the known components can be quantified, thereby providing key data for the subsequent component maximum value identification analysis. In summary, this formula fully considers the spectral feature correlation R between the spectral feature point of the i-th component to be tested and the spectral feature point of the j-th known component. ij , the total number of spectral feature points of the tested components in the spectral feature point set of different tested components in the tested Chinese herbal medicine n, the index item number i of the spectral feature points of the tested components, the total number of spectral feature points of the known components in the spectral feature point set of known components of Chinese herbal medicine m, the index item number j of the spectral feature points of the known components, the minimum wavelength λ in the spectral range min , the maximum wavelength λ in the spectral range max , integral wavelength variable parameter λ within the spectral range, spectral intensity I of the spectral characteristic point of the i-th component to be measured at the spectral wavelength λ i (λ), the mean value of the spectral intensity distribution of the spectral characteristic point of the i-th component to be measured The standard deviation σ of the spectral intensity distribution of the spectral characteristic point of the i-th component to be measured i , the spectral intensity I of the jth known component spectral characteristic point at the spectral wavelength λ j (λ), the mean value of the spectral intensity distribution of the jth known component spectral feature point The standard deviation σ of the spectral intensity distribution of the jth known component spectral characteristic point j , exponential function exp, wavelength distribution mean λ0 within the spectral range, wavelength distribution standard deviation Δλ within the spectral range, correction value ε of spectral feature correlation, according to the spectral feature correlation R between the spectral feature point of the i-th component to be measured and the spectral feature point of the j-th known component ij The correlation between the above parameters constitutes a functional relationship This formula can realize the correlation calculation process of the spectral feature point sets of different components to be tested in the Chinese herbal medicine and the spectral feature point sets of known components in the Chinese herbal medicine. At the same time, by introducing the correction value ε of the spectral feature correlation, it can be adjusted according to the errors occurring in the calculation process, thereby improving the accuracy and applicability of the spectral feature correlation calculation formula.

[0130] Preferably, the present invention further provides a Chinese herbal medicine component identification system based on spectroscopy, which is used to perform the Chinese herbal medicine component identification method based on spectroscopy as described above, and the Chinese herbal medicine component identification system based on spectroscopy comprises:

[0131] The Chinese herbal medicine sample spectrum data acquisition module is used to select lasers of different frequencies through a spectral instrument to perform differential wavelength feature analysis and integration on the quantitative Chinese herbal medicine samples to be tested, so as to obtain the spectrum information data of the Chinese herbal medicine samples to be tested; perform spectrum wavelength calibration processing on the spectrum information data of the Chinese herbal medicine samples to be tested, so as to obtain the spectrum calibration data of the Chinese herbal medicine samples to be tested;

[0132] The Chinese herbal medicine spectrum feature vector conversion module is used to perform spectrum feature recognition and analysis on the spectrum calibration data of the Chinese herbal medicine sample to be tested, and obtain the spectrum feature information data of the Chinese herbal medicine sample to be tested; perform feature vector conversion processing on the spectrum feature information data of the Chinese herbal medicine sample to be tested, so as to obtain the spectrum feature vector of the Chinese herbal medicine sample to be tested;

[0133] The Chinese herbal medicine component spectral feature point set sampling module is used to perform vector classification processing on the spectral feature vectors of the Chinese herbal medicine samples to be tested based on the spectral feature information data of the Chinese herbal medicine samples to be tested, so as to obtain the spectral feature vectors of different components of the Chinese herbal medicine to be tested; perform feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, so as to obtain the spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested;

[0134] The spectral feature association recognition module for Chinese herbal medicine components is used to obtain the spectral feature point set of known components of Chinese herbal medicines, and to perform association matrix mapping processing on the spectral feature point sets of different components to be tested in the Chinese herbal medicines and the spectral feature point sets of known components of Chinese herbal medicines, so as to obtain the spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicines; and to perform component maximum value recognition analysis based on the spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicines, so as to obtain the component identification results of the Chinese herbal medicine samples to be tested.

[0135] In summary, the present invention provides a Chinese herbal medicine ingredient identification system based on spectroscopy, which is composed of a Chinese herbal medicine sample spectral data acquisition module, a Chinese herbal medicine spectral feature vector conversion module, a Chinese herbal medicine ingredient spectral feature point set sampling module and a Chinese herbal medicine ingredient spectral feature association identification module. It can realize any one of the Chinese herbal medicine ingredient identification methods based on spectroscopy described in the present invention, and is used to combine the operations between computer programs running on each module to realize any one of the Chinese herbal medicine ingredient identification methods based on spectroscopy. The internal structures of the system cooperate with each other, which can greatly reduce duplication of work and manpower investment, and can quickly and effectively provide a more accurate and efficient Chinese herbal medicine ingredient identification process, thereby simplifying the operation flow of the Chinese herbal medicine ingredient identification system based on spectroscopy.

[0136] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0137] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for identifying Chinese herbal medicine ingredients based on spectroscopy, characterized in that: The following steps are involved: Step S1: using a spectral instrument to select lasers of different frequencies to perform differential wavelength characteristic analysis and integration on a quantitative Chinese herbal medicine sample to be tested, so as to obtain spectral information data of the Chinese herbal medicine sample to be tested; Performing spectral wavelength calibration processing on the spectral information data of the Chinese herbal medicine sample to be tested to obtain spectral calibration data of the Chinese herbal medicine sample to be tested; Step S2: performing spectral feature recognition analysis on the spectral calibration data of the Chinese herbal medicine sample to be tested, and obtaining spectral feature information data of the Chinese herbal medicine sample to be tested; Performing feature vector conversion processing on the spectral feature information data of the Chinese herbal medicine sample to be tested to obtain the spectral feature vector of the Chinese herbal medicine sample to be tested; Step S3: performing vector classification processing on the spectral feature vectors of the Chinese herbal medicine samples to be tested based on the spectral feature information data of the Chinese herbal medicine samples to be tested, so as to obtain spectral feature vectors of different components of the Chinese herbal medicine to be tested; performing feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, so as to obtain spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested; Step S4: obtaining a spectral feature point set of known components of Chinese herbal medicine, and performing correlation matrix mapping analysis on the spectral feature point sets of different components to be tested in the Chinese herbal medicine and the spectral feature point sets of known components of the Chinese herbal medicine, so as to obtain a spectral feature correlation matrix between the components to be tested and the known components of the Chinese herbal medicine; According to the spectral feature correlation matrix between the components to be tested and the known components of the Chinese herbal medicine, component maximum value identification analysis is performed to obtain the component identification results of the Chinese herbal medicine sample to be tested.

2. The method for identifying Chinese herbal medicine components based on spectroscopy according to claim 1, characterized in that: Step S1 includes the following steps: Step S10: crushing and grinding the Chinese herbal medicine sample to be tested and sieving impurities to obtain a powdered Chinese herbal medicine sample to be tested; homogenizing the powdered Chinese herbal medicine sample to be tested to obtain a quantitative Chinese herbal medicine sample to be tested; Step S11: performing component dyeing and labeling processing on the quantitative Chinese herbal medicine sample to be tested, so as to obtain a quantitative dyeing and labeling sample of the Chinese herbal medicine to be tested; Step S12: Select lasers of different frequencies through a spectral instrument to perform spectral scanning analysis on the quantitative Chinese herbal medicine dyed and labeled samples to be tested, and obtain spectral data of different components in the sample at different frequencies; perform differential wavelength characteristic analysis on the spectral data of different components in the sample at different frequencies to obtain differential wavelength spectral information characteristic data of different components in the sample; perform data integration on the differential wavelength spectral information characteristic data of different components in the sample to obtain spectral information data of the Chinese herbal medicine sample to be tested; Step S13: performing noise removal processing on the spectrum information data of the Chinese herbal medicine sample to be tested to obtain denoised spectrum information data of the Chinese herbal medicine sample to be tested; Step S14: performing background correction processing on the spectral denoising information data of the Chinese herbal medicine sample to be tested, to obtain background correction data of the spectral data of the Chinese herbal medicine sample to be tested; Step S15: performing spectral wavelength calibration processing on the spectral background correction data of the Chinese herbal medicine sample to be tested, to obtain spectral calibration data of the Chinese herbal medicine sample to be tested.

3. The method for identifying Chinese herbal medicine components based on spectroscopy according to claim 2, characterized in that: Step S15 includes the following steps: Step S151: performing spectrum region segmentation processing on the spectrum background correction data of the Chinese herbal medicine sample to be tested according to a preset time division interval to obtain the spectrum data region of the Chinese herbal medicine sample to be tested; Step S152: Calculate the degree of deviation of the spectral wavelength information data in the spectral data region of the Chinese herbal medicine sample to be tested using a spectral wavelength deviation degree calculation formula to obtain a wavelength deviation degree value of the spectral region of the Chinese herbal medicine sample to be tested; Step S153: comparing and judging the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested according to a preset wavelength shift degree threshold value; when the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is greater than or equal to the preset wavelength shift degree threshold value, the spectral data region of the Chinese herbal medicine sample to be tested corresponding to the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is marked as a wavelength high-frequency shift region; when the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is less than the preset wavelength shift degree threshold value, the spectral data region of the Chinese herbal medicine sample to be tested corresponding to the wavelength shift degree value of the spectral region of the Chinese herbal medicine sample to be tested is marked as a wavelength low-frequency shift region; Step S154: performing wavelength classification calibration processing on the spectrum wavelength information data in the wavelength high-frequency offset region to obtain the spectrum wavelength high-frequency offset calibration data of the Chinese herbal medicine to be tested; Step S155: performing wavelength nonlinear calibration processing on the spectrum wavelength information data in the wavelength low-frequency offset region to obtain wavelength low-frequency offset calibration data of the Chinese herbal medicine spectrum to be tested; Step S156: merging the calibration data of the high-frequency wavelength offset of the spectrum of the Chinese herbal medicine to be tested and the calibration data of the low-frequency wavelength offset of the spectrum of the Chinese herbal medicine to be tested to obtain the calibration data of the spectrum of the Chinese herbal medicine sample to be tested.

4. The method for identifying Chinese herbal medicine components based on spectroscopy according to claim 3, characterized in that: The calculation formula for the degree of spectrum wavelength shift in step S152 is specifically: Where Δθ is the wavelength shift value of the spectral region of the Chinese herbal medicine sample to be tested, θ min is the minimum spectral wavelength of the spectral data area of ​​the Chinese herbal medicine sample to be tested, θ max is the maximum spectral wavelength of the spectral data area of ​​the Chinese herbal medicine sample to be tested, θ is the spectral wavelength variable parameter of the spectral data area of ​​the Chinese herbal medicine sample to be tested, θ0 is the spectral specific reference wavelength, C is the spectral wavelength shift influence control index, and η is the correction coefficient of the wavelength shift degree value of the spectral area of ​​the Chinese herbal medicine sample to be tested.

5. The method for identifying Chinese herbal medicine components based on spectroscopy according to claim 3, characterized in that: Step S154 includes the following steps: Step S1541: inputting the spectrum wavelength information data in the wavelength high-frequency offset region into a preset first-level FIFO calibration processing cache queue to obtain high-frequency offset first-level FIFO calibration data; performing first-level calibration processing on the high-frequency offset first-level FIFO calibration data using a feature matching calibration algorithm to obtain wavelength high-frequency offset first-level calibration data; Step S1542: performing calibration effect evaluation and analysis on the wavelength high frequency offset primary calibration data to obtain wavelength high frequency offset primary calibration success data and fluctuation high frequency offset primary calibration failure data; Step S1543: inputting the fluctuation high frequency offset primary calibration failure data into the preset secondary FIFO calibration processing cache queue to obtain the high frequency offset secondary FIFO calibration data; performing secondary calibration processing on the high frequency offset secondary FIFO calibration data using a multi-scale calibration optimization algorithm to obtain the wavelength high frequency offset secondary calibration data; Step S1544: fusing the wavelength high frequency offset first-level calibration success data and the wavelength high frequency offset second-level calibration data to obtain the wavelength high frequency offset calibration data of the Chinese herbal medicine spectrum to be tested.

6. The method for identifying Chinese herbal medicine components based on spectroscopy according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: performing spectral peak recognition analysis on the spectral calibration data of the Chinese herbal medicine sample to be tested, and obtaining a spectral peak position feature data set of the Chinese herbal medicine sample to be tested; Step S22: performing absorption peak intensity detection on the spectral calibration data of the Chinese herbal medicine sample to be tested, and obtaining a characteristic data set of absorption peak intensity of the spectral calibration data of the Chinese herbal medicine sample to be tested; Step S23: performing spectral peak-to-valley distance detection on the spectral calibration data of the Chinese herbal medicine sample to be tested, and obtaining a spectral peak-to-valley distance characteristic data set of the Chinese herbal medicine sample to be tested; Step S24: performing feature merging processing on the characteristic data set of peak position of the spectrum of the Chinese herbal medicine sample to be tested, the characteristic data set of absorption peak intensity of the spectrum of the Chinese herbal medicine sample to be tested, and the characteristic data set of peak-to-valley spacing of the spectrum of the Chinese herbal medicine sample to be tested, to obtain characteristic information data of the spectrum of the Chinese herbal medicine sample to be tested; Step S25: performing feature vector conversion processing on the spectral feature information data of the Chinese herbal medicine sample to be tested, and obtaining the spectral feature vector of the Chinese herbal medicine sample to be tested.

7. The method for identifying Chinese herbal medicine components based on spectroscopy according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: performing time series analysis on the spectral characteristic information data of the Chinese herbal medicine sample to be tested, and obtaining the spectral characteristic time series information data of the Chinese herbal medicine sample to be tested; Step S32: performing characteristic change pattern analysis on the time series information data of the spectrum characteristics of the Chinese herbal medicine sample to be tested, and obtaining the behavior pattern data of the spectrum characteristics change of the Chinese herbal medicine sample to be tested; Step S33: performing vector classification processing on the spectral feature vector of the Chinese herbal medicine sample to be tested based on the spectral feature change behavior pattern data of the Chinese herbal medicine sample to be tested, so as to obtain spectral feature vectors of different components of the Chinese herbal medicine to be tested; Step S34: performing feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, and obtaining spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested.

8. The method for identifying Chinese herbal medicine components based on spectroscopy according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: obtaining a set of spectral feature points of known components of Chinese herbal medicine; Step S42: using a spectral feature correlation calculation formula to perform correlation calculation on the spectral feature point sets of different components to be tested in the Chinese herbal medicine and the spectral feature point sets of known components of the Chinese herbal medicine, so as to obtain the spectral feature correlation between the spectral feature points of the components to be tested and the spectral feature points of the known components; Step S43: constructing an initial correlation matrix between the components to be tested and the known components of the Chinese herbal medicine, wherein the initial values ​​of the corresponding matrix areas in the initial correlation matrix are all zero, and comparing and judging the spectral feature correlation between the spectral feature points of the components to be tested and the spectral feature points of the known components according to a preset spectral feature correlation threshold value; if the spectral feature correlation between the spectral feature points of the components to be tested and the spectral feature points of the known components is greater than or equal to the preset spectral feature correlation threshold value, the value of the corresponding matrix area in the initial correlation matrix is ​​increased by one; if the spectral feature correlation between the spectral feature points of the components to be tested and the spectral feature points of the known components is less than the preset spectral feature correlation threshold value, the value of the corresponding matrix area in the initial correlation matrix remains unchanged, so as to obtain the spectral feature correlation matrix between the components to be tested and the known components of the Chinese herbal medicine; Step S44: performing component maximum value identification analysis based on the spectral feature correlation matrix between the components to be tested and the known components of the Chinese herbal medicine to obtain the component identification result of the Chinese herbal medicine sample to be tested.

9. The method for identifying Chinese herbal medicine components based on spectroscopy according to claim 8, characterized in that: The calculation formula of the spectral feature correlation degree in step S42 is specifically: In the formula, R ij is the spectral feature correlation between the spectral feature point of the i-th component to be tested and the spectral feature point of the j-th known component, n is the total number of spectral feature points of the components to be tested in the spectral feature point set of different components to be tested in the Chinese herbal medicine to be tested, i is the index item of the spectral feature point of the component to be tested, m is the total number of spectral feature points of the known components in the spectral feature point set of the known components of the Chinese herbal medicine, j is the index item of the spectral feature point of the known components, and λ min is the minimum wavelength in the spectral range, λ max is the maximum wavelength in the spectral range, λ is the integral wavelength variable parameter in the spectral range, I i (λ) is the spectral intensity of the spectral characteristic point of the i-th component to be measured at the spectral wavelength λ, is the mean value of the spectral intensity distribution of the spectral characteristic point of the i-th component to be measured, σ i is the standard deviation of the spectral intensity distribution of the uth spectral characteristic point of the component to be measured, I j (λ) is the spectral intensity of the jth known component spectral feature point at the spectral wavelength λ, is the mean value of the spectral intensity distribution of the jth known component spectral feature point, σ j is the standard deviation of the spectral intensity distribution of the jth known component spectral feature point, exp is the exponential function, λ0 is the mean value of the wavelength distribution within the spectral range, Δλ is the standard deviation of the wavelength distribution within the spectral range, and ε is the correction value of the spectral feature correlation.

10. A Chinese herbal medicine component identification system based on spectroscopy, characterized in that: Used to perform the method for identifying Chinese herbal medicine components based on spectroscopy as claimed in claim 1, the Chinese herbal medicine component identification system based on spectroscopy comprises: The Chinese herbal medicine sample spectrum data acquisition module is used to select lasers of different frequencies through a spectral instrument to perform differential wavelength feature analysis and integration on the quantitative Chinese herbal medicine samples to be tested, so as to obtain the spectrum information data of the Chinese herbal medicine samples to be tested; perform spectrum wavelength calibration processing on the spectrum information data of the Chinese herbal medicine samples to be tested, so as to obtain the spectrum calibration data of the Chinese herbal medicine samples to be tested; The Chinese herbal medicine spectrum feature vector conversion module is used to perform spectrum feature recognition and analysis on the spectrum calibration data of the Chinese herbal medicine sample to be tested, and obtain the spectrum feature information data of the Chinese herbal medicine sample to be tested; perform feature vector conversion processing on the spectrum feature information data of the Chinese herbal medicine sample to be tested, so as to obtain the spectrum feature vector of the Chinese herbal medicine sample to be tested; The Chinese herbal medicine component spectral feature point set sampling module is used to perform vector classification processing on the spectral feature vectors of the Chinese herbal medicine samples to be tested based on the spectral feature information data of the Chinese herbal medicine samples to be tested, so as to obtain the spectral feature vectors of different components of the Chinese herbal medicine to be tested; perform feature point set sampling processing on the spectral feature vectors of different components of the Chinese herbal medicine to be tested, so as to obtain the spectral feature point sets of different components to be tested in the Chinese herbal medicine to be tested; The spectral feature association recognition module for Chinese herbal medicine components is used to obtain the spectral feature point set of known components of Chinese herbal medicines, and to perform association matrix mapping processing on the spectral feature point sets of different components to be tested in the Chinese herbal medicines and the spectral feature point sets of known components of Chinese herbal medicines, so as to obtain the spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicines; and to perform component maximum value recognition analysis based on the spectral feature association matrix between the components to be tested and the known components of the Chinese herbal medicines, so as to obtain the component identification results of the Chinese herbal medicine samples to be tested.

Images