A multi-channel data synchronous acquisition and analysis system of spectrophotometry

Abstract

The present application belongs to the technical field of spectral analysis, and particularly relates to a kind of multichannel data synchronous acquisition and analysis system of spectrophotometry, including memory, processor and computer program stored in memory and running on processor, when processor executes computer program, the following steps are realized, including: obtaining multi-wavelength absorbance data to generate multiple spectral curves;According to the characteristic degree of component absorption peak calculated by peak retention feature, the wavelength of component absorption peak is screened;According to the difference between absorbance value and waveform transverse span, the change degree of multiple frames is calculated, and the wavelength of unstable component absorption peak is identified;Absorbance change sequence is constructed and normalized, concentration change characteristics are calculated according to sequence similarity, and risk absorption peak wavelength is determined;Based on risk absorption peak wavelength, the absorbance data of multiple sub-samples is comprehensively analyzed.The present application effectively eliminates solution environment noise interference, improves the analysis accuracy and detection reliability of multichannel data acquisition.

Description

Technical Field

[0001] This invention relates to the field of spectral analysis technology. More specifically, this invention relates to a spectrophotometric multi-channel data synchronous acquisition and analysis system. Background Technology

[0002] During the production, processing, transportation, and storage of food, its nutritional components, additive content, and contaminant levels can change depending on process conditions, storage environment, and time. If the content of target components in food cannot be detected in a timely and accurate manner, it will not only affect food quality evaluation and production process control but may also increase food safety risks. Therefore, it is necessary to employ rapid, sensitive, and highly quantitative detection technologies to test for nutrients, food additives, harmful substances, and quality indicators in food.

[0003] Spectrophotometry utilizes the absorption characteristics of target substances at specific wavelengths of light. By measuring absorbance and combining it with a standard curve, quantitative analysis of target components can be achieved. It has significant application value in fields such as food nutrient analysis, food additive detection, contaminant detection, and food quality evaluation. Current spectrophotometric techniques for food detection typically employ multi-channel absorbance acquisition using multiple wavelengths within a range, and then construct absorbance spectral curves for food component analysis.

[0004] However, in actual sample testing, the absorption peaks obtained from the absorbance measurements may not accurately represent the true composition of the food. They could originate from scattering due to solvent turbidity, drift of some opaque micro-impurities, or inherent instrument noise. These transient absorption peaks formed by such interference factors in the solution typically cannot remain stably at a fixed wavelength during multi-frame continuous acquisition, thus reducing the reliability of the absorption peaks in the actual absorbance spectrum and decreasing the accuracy of the assessment of food composition. Summary of the Invention

[0005] To address the technical problem that existing spectrophotometric multi-channel data acquisition is susceptible to interference from noise in the solution environment, leading to decreased reliability of absorption peaks and reduced accuracy in food component detection, this invention provides a spectrophotometric multi-channel data synchronous acquisition and analysis system. The system includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it performs the following steps:

[0006] The absorbance data of multiple sub-samples after the total solution of the tested food is divided at multiple acquisition times are obtained, and multi-frame spectral curves of each sub-sample are generated.

[0007] For any subsample, extract the peak data points in each frame of the spectral curve, calculate the characteristic degree of the component absorption peak based on the peak retention characteristics of each wavelength in historical multi-frame data, and select the component absorption peak wavelengths based on the characteristic degree of the component absorption peak.

[0008] For each component absorption peak wavelength, the degree of change in multiple frames is calculated based on the difference in absorbance value and the difference in waveform lateral span of the peak data point at the absorption peak wavelength in historical multi-frame data. The unstable component absorption peak wavelength is identified based on the degree of change in multiple frames.

[0009] For each unstable component's absorption peak wavelength, an absorbance variation sequence is constructed and normalized to obtain a normalized variation sequence. Concentration variation characteristics are calculated based on the similarity between the normalized variation sequences, and risk absorption peak wavelengths are marked based on the concentration variation characteristics.

[0010] A comprehensive analysis was conducted based on the labeling results of different subsamples.

[0011] Preferably, the step of acquiring multi-wavelength absorbance data of multiple sub-samples after dividing the total solution of the tested food at multiple acquisition times, and generating multi-frame spectral curves for each sub-sample, includes:

[0012] The food to be tested is crushed and dissolved in a solvent to extract the target components. The insoluble solid residue is removed by filtration to obtain the test liquid. The test liquid is homogenized to form a total solution, and the total solution is divided into multiple subsamples.

[0013] A broadband light source emits continuous light to illuminate the transmitted subsample. A photodetector receives the transmitted monochromatic light of different wavelengths and converts it into an electrical signal to obtain the transmitted light intensity of each wavelength channel. The absorbance is calculated based on the transmitted light intensity and the initial incident light intensity. The spectral curve at the acquisition time is plotted with wavelength as the abscissa and absorbance as the ordinate.

[0014] Preferably, the characteristic degree of the component absorption peak satisfies the expression:

[0015] ;

[0016] In the formula, Indicates the first The degree of characteristic of the component absorption peaks at each wavelength; Indicates the first The total number of frames in which each wavelength appears as a peak data point across multiple historical frames; Indicates the first The total number of wave crest chains formed by each wavelength across multiple historical frames; Indicates the first The first wavelength The number of consecutive frames contained in a wave crest chain; This represents the normalization function.

[0017] Preferably, the method for obtaining the peak chain is as follows:

[0018] Regarding the first The wavelength, in the historical multi-frame temporal arrangement, will be the first wavelength. Adjacent frames with consecutive peak data points of a certain wavelength are connected to form a peak chain. A peak chain contains at least two consecutive frames of the first wavelength. If the peak data point of the i-th wavelength, If a wavelength has peak data points in multiple consecutive frames, then these consecutive frames together form a peak chain.

[0019] Preferably, the wavelengths of the component absorption peaks are selected based on the characteristic degree of the component absorption peaks, including:

[0020] If the characteristic degree of the component absorption peak at a given wavelength is greater than or equal to a preset threshold for the characteristic degree of the component absorption peak, then the wavelength is determined to be the component absorption peak wavelength of the food being tested.

[0021] Preferably, the degree of change in the multiple frames satisfies the expression:

[0022] ;

[0023] In the formula, Indicates the first The degree of multi-frame variation in the wavelength of the absorption peak of each component; Indicates the first The total number of frames in which the absorption peak wavelength of each component exists in historical multi-frame data points; and They represent the first The wavelength of the absorption peak of each component is the first in multiple historical frames. The and the first Absorbance values ​​of each peak data point; and They represent the first The wavelength of the absorption peak of each component is the first in multiple historical frames. The and the first The waveform lateral span at each peak data point, wherein the waveform lateral span is the lateral distance from the peak data point to the adjacent trough point on both sides; This represents the normalization function.

[0024] Preferably, identifying the absorption peak wavelength of unstable components based on the degree of change across multiple frames includes:

[0025] In response to a multi-frame variation in the wavelength of a component absorption peak that is greater than or equal to a preset multi-frame variation threshold, the wavelength of that component absorption peak is taken as the wavelength of the unstable component absorption peak.

[0026] Preferably, constructing an absorbance variation sequence for each unstable component's absorption peak wavelength includes:

[0027] The absorbance values ​​of each unstable component absorption peak wavelength in all historical frames are extracted and arranged in the order of acquisition time to form an absorbance sequence. The difference between adjacent absorbance values ​​in the absorbance sequence is calculated and the absolute value is taken. The absolute values ​​of the differences between all adjacent absorbance values ​​form an absorbance change sequence.

[0028] Preferably, the concentration change feature satisfies the expression:

[0029] ;

[0030] In the formula, Indicates the first The concentration variation characteristics of the absorption peak wavelength of the unstable component; Indicates the number of change values ​​within the normalized change sequence; Indicates the first In the normalized variation sequence of the absorption peak wavelength of the unstable component, the first... One change value; Indicates the relationship with the first The normalized variation sequence of the absorption peak wavelength of the unstable component with the highest similarity to the normalized variation sequence of the absorption peak wavelength of other unstable components is the first one. One change value; Represents an exponential function with the natural constant as its base;

[0031] The wavelength of the risk absorption peak based on the concentration change characteristics includes:

[0032] If the concentration change characteristic of the absorption peak wavelength of an unstable component is less than a preset concentration change characteristic threshold, it is determined to be a risky absorption peak wavelength.

[0033] Preferably, the comprehensive analysis based on the labeling results of different sub-samples includes:

[0034] For any given subsample, if its spectral curve contains an absorption peak wavelength marked as a risk absorption peak wavelength, the absorbance value of that subsample at that risk absorption peak wavelength is discarded. From other subsamples detected simultaneously, target subsamples that are not marked as risk absorption peak wavelengths are selected. Based on a comprehensive analysis of the absorbance values ​​of all target subsamples at the risk absorption peak wavelengths, the true absorption characteristics of the food components at those wavelengths are determined. If the same wavelength in all subsamples is marked as a risk absorption peak wavelength, the detection data at that wavelength is deemed invalid, and a re-acquisition command is output.

[0035] The beneficial effects of this invention are as follows: This invention calculates the characteristic degree of component absorption peaks and screens out component absorption peak wavelengths by analyzing the peak retention characteristics of each wavelength in historical multi-frame data. Utilizing the high stability of the true component absorption peaks at their wavelength positions, it eliminates transient noise absorption peaks induced by random events such as solvent scattering, impurity drift, and instrument noise, thus initially locking in the absorption peak wavelengths that truly reflect the food components. Furthermore, based on the differences in absorbance values ​​and waveform lateral span of the peak data points at the component absorption peak wavelengths in historical multi-frame data, it calculates the degree of change across multiple frames, identifying unstable component absorption peak wavelengths. By capturing the temporal fluctuations in absorbance values ​​and waveform morphology, it effectively identifies absorption peaks distorted by the superposition of transient interference factors. Subsequently, it constructs an absorbance change sequence... The sequences were normalized, and concentration variation characteristics were calculated based on the similarity between sequences to determine the wavelengths of risk absorption peaks. Utilizing the physical characteristic that absorption peaks exhibit fixed ratios and year-on-year changes when the chemical structure of food components remains unchanged, the dynamic changes in absorbance caused by static concentration or dilution of the solution were eliminated, accurately identifying the wavelengths of risk absorption peaks truly affected by noise. Finally, based on the labeled risk absorption peak wavelengths, anomaly removal and comprehensive analysis were performed on the absorbance data of multiple subsamples. By removing single, interfering anomaly data and comprehensively utilizing the true absorbance values ​​of undisturbed subsamples for multi-dimensional evaluation, the interference of solution environment noise on absorption peaks was ultimately eliminated, significantly improving the analytical accuracy of multi-channel spectrophotometric data and the reliability of food component detection. Attached Figure Description

[0036] Figure 1 This is a flowchart of a method for synchronous acquisition and analysis of multi-channel data using spectrophotometry provided in an embodiment of the present invention;

[0037] Figure 2 This is a wavelength absorbance spectrum curve of some food components. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0040] This invention discloses a spectrophotometric multi-channel data synchronous acquisition and analysis system, including a processor and a memory. The processor executes a computer program stored in the memory to implement a spectrophotometric multi-channel data synchronous acquisition and analysis method. (Refer to...) Figure 1 The method includes steps S1-S5:

[0041] S1: Obtain multi-wavelength absorbance data of multiple sub-samples after dividing the total solution of the tested food at multiple acquisition times, and generate multi-frame spectral curves for each sub-sample.

[0042] Specifically, the pretreatment for optical detection of the food sample involves: pulverizing the food sample, adding a specific solvent and stirring to fully dissolve the target component in the liquid phase, then filtering through filter paper to remove insoluble solid residues, obtaining a clear test liquid. Some detection methods require the addition of a colorimetric reagent to the test liquid to form a colored product with characteristic absorption peaks; for example, adding biuret reagent to detect proteins generates a purple complex. The test liquid is then thoroughly homogenized to form a total solution, which is then divided into multiple subsamples to ensure that the target analyte concentration is the same in all subsamples.

[0043] Multiple subsamples are injected into multiple cuvettes and simultaneously placed into multiple sample cells of a multichannel spectrophotometer. Broad-spectrum light sources such as deuterium lamps or tungsten halogen lamps are used to simultaneously emit continuous light onto the subsamples in each cuvette, allowing the light beam to penetrate each subsample. The transmitted polychromatic light is then dispersed by a grating into multiple continuous wavelengths of monochromatic light. Because different components in the food exhibit characteristic absorption properties for specific wavelengths of monochromatic light, this wavelength of monochromatic light undergoes selective attenuation during transmission. Therefore, the change in the spectral intensity distribution of the transmitted light directly reflects the absorption information of different components in the food.

[0044] Different wavelengths of monochromatic light, after transmission, simultaneously illuminate different pixels of the photodetector array. The photodetector converts the received light signal into an electrical signal to obtain the transmitted light intensity of the corresponding wavelength channel for each pixel. Based on the transmitted light intensity and the initial incident light intensity, the absorbance of the wavelength channel at the acquisition time is calculated. The photodetector is a charge-coupled device (CCD), complementary metal-oxide-semiconductor (CMOS), or photodiode array detector (PDA).

[0045] Furthermore, by plotting wavelength as the abscissa and absorbance as the ordinate, the spectral curve at the current moment can be obtained.

[0046] By periodically repeating the above transmission and detection process, multi-wavelength absorbance data within the entire measurable wavelength range over a period of time can be obtained. In this embodiment, the data is collected every 10 seconds for 10 minutes, resulting in 60 different spectral curves for each subsample, which are recorded as 60 frames of spectral data. In other embodiments, the implementer can set the acquisition frequency and acquisition duration according to actual detection needs.

[0047] For example, Figure 2 This is a spectrum of wavelength absorbance for some food components, with wavelength on the horizontal axis and absorbance on the vertical axis.

[0048] S2: For any subsample, extract the peak data points in each frame of the spectral curve, calculate the characteristic degree of the component absorption peak based on the peak retention characteristics of each wavelength in historical multi-frame data, and select the component absorption peak wavelengths based on the characteristic degree of the component absorption peak.

[0049] It should be noted that in the spectral curve constructed from single-frame acquired data, local maxima of absorbance at specific wavelengths form fluctuating absorption peaks, characterizing the strong absorption characteristics of specific substances in food at that wavelength. However, the absorption peaks obtained in actual detection may not all represent the true absorption peaks of food components. Some absorption peaks may originate from interference factors such as light scattering caused by solvent turbidity, random drift of opaque micro-impurities, and inherent instrument noise. These transient absorption peaks formed by such interference factors in solution typically cannot remain stably at a fixed wavelength position during multi-frame continuous acquisition. In contrast, the true absorption peaks corresponding to food components exhibit high stability at their wavelength positions. Therefore, this invention preliminarily screens the absorption peak wavelengths corresponding to food components by analyzing the temporal variation characteristics of absorption peaks at various wavelength positions in multi-frame continuous acquisition data.

[0050] Specifically, an adaptive multi-scale peak detection algorithm is used to extract all peak data points in each frame of the spectral curve. Traversing the entire wavelength range, if a peak data point has appeared in historical multi-frame data for a certain wavelength, that wavelength is marked as a suspected absorption peak wavelength. For each suspected absorption peak wavelength, its peak retention characteristics in historical multi-frame data are analyzed: since solvent scattering, impurity drift, and instrument noise are all random events, the absorption peaks induced by them appear less frequently and have a discrete spatial distribution in historical multi-frame data; while the actual absorption peaks corresponding to food components show a stable retention characteristic of continuous existence and a large number in historical multi-frame data. Therefore, if the number of peak data points for a certain wavelength in historical multi-frame data is smaller, and the continuity of the peak data points is lower, the absorption peak at that wavelength is more likely to be identified as a noise absorption peak; conversely, it is identified as a component absorption peak.

[0051] The degree of characteristic of the component absorption peaks at each wavelength is calculated based on the peak retention characteristics, satisfying the expression:

[0052] ;

[0053] In the formula, Indicates the first The degree of characteristic of the component absorption peaks at each wavelength; Indicates the first The larger the number of frames in which a wavelength has peak data points in historical frames, the more significant the wavelength's sustained retention characteristics are. Indicates the first The total number of peak chains formed by a wavelength across multiple historical frames, wherein the method for obtaining the peak chains is as follows: for the first wavelength... The wavelength, in the historical multi-frame temporal arrangement, will be the first wavelength. Adjacent frames with consecutive peak data points of a certain wavelength are connected to form a peak chain. A peak chain contains at least two consecutive frames of the first wavelength. If the peak data point of the i-th wavelength, If a wavelength has peak data points in multiple consecutive frames, then these consecutive frames together form a peak chain, from which statistical analysis can be performed. The wave crest chain, if the first If a wavelength does not have peak data points in two consecutive frames across multiple historical frames, then the characteristic degree of the component absorption peak is directly determined. This is to avoid calculation errors caused by a denominator of zero. Indicates the first The first wavelength The number of consecutive frames contained in a wave crest chain. Used to highlight the numerical contribution of long wave crest chains, characterizing the intensity of wave crest data points that continue to appear across multiple frames; This represents the average sustained intensity of all peak chains within the current wavelength. The larger this value, the more the peak distribution within the current wavelength conforms to the stable distribution characteristics of the component absorption peaks. This represents the normalization function. In this embodiment, maximum-minimum normalization is used, which is based on the total number of frames in which peak data points of all wavelengths appear in historical multi-frame data. Perform maximum and minimum value normalization, and calculate the average duration intensity based on the values ​​corresponding to all wavelengths. In other embodiments, the implementer can select the corresponding normalization function based on the actual implementation situation to perform maximum and minimum value normalization.

[0054] In practical food spectrophotometric detection, since an ideal noise-free environment is difficult to achieve, component absorption peaks and noise absorption peaks often coexist. After normalization processing in the component absorption peak characteristic degree calculation process, the component absorption peak characteristic degree can effectively divide the wavelengths with true component absorption peaks and wavelengths with only noise characteristics to the two ends of the normalized value range. Therefore, in this embodiment, the component absorption peak characteristic degree threshold is set to 0.5 to eliminate noise interference from the solution background environment and obtain the wavelengths of all components with absorption peaks. In other embodiments, the implementer can adaptively adjust the component absorption peak characteristic degree threshold according to the noise level and data distribution characteristics of the actual detection environment.

[0055] When the Characteristic degree of component absorption peaks at each wavelength When the wavelength is greater than or equal to the characteristic threshold of the component absorption peak, it is determined to be the wavelength corresponding to the component absorption peak of the current food being tested; when When the wavelength is less than the characteristic threshold of the component absorption peak, it is determined that the wavelength corresponds to the noise absorption peak.

[0056] S3: For the absorption peak wavelength of each component, calculate the degree of change in multiple frames based on the difference in absorbance value and the difference in the horizontal span of the waveform at the peak data point of the absorption peak wavelength in historical multi-frame data, and identify the absorption peak wavelength of unstable components based on the degree of change in multiple frames.

[0057] It should be noted that after the initial screening to obtain the absorption peak wavelengths of the components, transient interference factors such as solvent scattering, drift of minute impurities, and inherent instrument noise in the actual detection environment may still superimpose on the true absorption signal, causing fluctuations in the absorbance value or waveform morphology of the peak data points at the component absorption peak wavelengths. To identify the component absorption peak wavelengths affected by such transient interference, this invention further analyzes the temporal fluctuation characteristics of the peak data points at each component absorption peak wavelength across historical frames.

[0058] Specifically, for each component's absorption peak wavelength, the degree of multi-frame variation in the component's absorption peak wavelength is determined based on the historical multi-frame absorbance peak values ​​and peak shape changes of the component's absorption peak wavelength:

[0059] ;

[0060] In the formula, Indicates the first The degree of multi-frame variation in the wavelength of the absorption peak of each component; Indicates the first The total number of frames in which the absorption peak wavelengths of each component exist across multiple historical frames, and If it is a positive integer greater than or equal to 2, then If it equals 1, then directly set Equal to 0; and They represent the first The wavelength of the absorption peak of each component is the first in multiple historical frames. The and the first Absorbance values ​​of each peak data point; and They represent the first The wavelength of the absorption peak of each component is the first in multiple historical frames. The and the first The waveform lateral span at each peak data point, wherein the waveform lateral span is the lateral distance from the peak data point to the adjacent trough point on both sides; Indicates the first The average of the pairwise differences in absorbance values ​​of all peak data points in historical multi-frame data for each component absorption peak wavelength. The larger the value, the more drastic the fluctuation in absorbance value. Indicates the first The average difference between the pairwise lateral spans of the waveforms of each component absorption peak wavelength in all historical frames. The larger this value is, the more violent the waveform morphology fluctuations. This represents the normalization function. In this embodiment, a maximum-minimum normalization function is used, which is the maximum and minimum values ​​in the set formed by the mean of the pairwise differences of the absorbance values ​​corresponding to the absorption peak wavelengths of all components. Max-min normalization is performed, which involves finding the maximum and minimum values ​​in the set of mean differences between the pairwise waveform transverse spans corresponding to the absorption peak wavelengths of all components. In other embodiments, the implementer can select the corresponding normalization function based on the actual implementation situation to perform maximum and minimum value normalization.

[0061] After normalization processing during the multi-frame variation calculation, the multi-frame variation can effectively divide the absorption peak wavelengths of components with significant fluctuations and those of components with stable absorption peak wavelengths to the two ends of the normalized value range.

[0062] When the multi-frame variation of the component absorption peak wavelength is greater than or equal to a preset multi-frame variation threshold, the component absorption peak at that wavelength is determined to have an unstable variation, and the wavelength of that component absorption peak is recorded as an unstable component absorption peak wavelength. When the multi-frame variation of the component absorption peak wavelength is less than the preset multi-frame variation threshold, the component absorption peak at that wavelength is determined to be a stable component absorption peak. The value range of the multi-frame variation is [0,2]. Considering that after normalization processing in the multi-frame variation calculation process, the multi-frame variation can effectively divide the component absorption peak wavelengths with significant fluctuations and the component absorption peak wavelengths that remain stable to the two ends of the normalized value range, this embodiment sets the multi-frame variation threshold to 1. In other embodiments, the implementer can adaptively adjust the multi-frame variation threshold according to the noise level and data distribution characteristics of the actual detection environment.

[0063] S4: For the absorption peak wavelengths of each unstable component, construct an absorbance variation sequence and normalize it to obtain a normalized variation sequence. Calculate the concentration variation characteristics based on the similarity between the normalized variation sequences, and mark the risk absorption peak wavelengths based on the concentration variation characteristics.

[0064] It should be noted that, since food often contains sugars or salts, during spectrophotometric detection of food, if the test liquid contains sugars or salts, cell debris or macromolecular polymers will absorb water and swell during continuous standing, converting free water in the solution into bound water. This change leads to a reduction in solution volume, causing the food solution to concentrate. The dilution and concentration of the food solution cause synchronous changes in the concentration of all dissolved components, resulting in a systematic and proportional change in absorbance values ​​at various wavelengths. This means that multi-frame dynamic differences in absorbance are generated during detection. Therefore, directly judging the absorption peak wavelength of unstable components based on multi-frame changes may misinterpret absorbance fluctuations caused by normal solution concentration changes as noise components. Furthermore, a food component is usually composed of polymers of various chemical molecules. When its chemical structure remains unchanged, the absorbance changes of different components during concentration exhibit a high degree of synchronicity and correlation. For individual components without multiple peaks, the absorbance changes of the colored products formed by the colorimetric reaction also follow the same concentration-dependent law. In contrast, light scattering caused by turbidity does not precisely affect the absorbance change trend at a specific wavelength, the drift of minute impurities usually only causes random fluctuations in local wavelengths within the same time period, and the inherent noise of the instrument only manifests as irregular random disturbances. The absorbance changes caused by these interfering factors lack synchronous correlation between different wavelengths. Therefore, this invention further obtains the concentration change characteristics of wavelength-corresponding absorbance changes based on the synchronous correlation characteristics between the absorbance change sequence of the absorption peak wavelength of unstable components and the absorbance change sequences of the absorption peak wavelengths of other unstable components.

[0065] Specifically, an absorbance variation sequence is constructed at the absorption peak wavelength of each unstable component. This includes: extracting the absorbance values ​​of each unstable component's absorption peak wavelength across all historical frames and arranging them in chronological order to form an absorbance sequence; calculating the absolute values ​​of the differences between adjacent absorbance values ​​in the absorbance sequence; and using the absolute values ​​of the differences between all adjacent absorbance values ​​to form an absorbance variation sequence. Based on the maximum and minimum values ​​within the absorbance variation sequence, a maximum-minimum normalization is performed on the absorbance variation sequence. The normalized absorbance variation sequence is denoted as the normalized variation sequence. After maximum-minimum normalization, the original absolute values ​​of inter-frame differences are transformed into relative degrees of change.

[0066] The condensed variation features are calculated based on the similarity of the normalized variation sequences, satisfying the expression:

[0067] ;

[0068] In the formula, Indicates the first The concentration variation characteristics of the absorption peak wavelength of the unstable component; Indicates the number of change values ​​within the normalized change sequence; Indicates the first In the normalized variation sequence of the absorption peak wavelength of the unstable component, the first... One change value; Indicates the relationship with the first The normalized variation sequence of the absorption peak wavelength of the unstable component with the highest similarity to the normalized variation sequence of the absorption peak wavelength of other unstable components is the first one. One change value; This represents an exponential function with the natural constant as its base. Used to achieve inverse proportional normalization; Indicates the first The mean difference between the normalized variation sequence of the absorption peak wavelength of the unstable component and all variation values ​​of its most similar sequence is considered. The smaller this mean difference, the stronger the difference. The more similar the absorption peak wavelengths of the unstable components show to a certain extent, the greater the year-on-year variation. The stronger the concentration change characteristic of the absorption peak wavelength of the unstable component, the better. Among them, the concentration change is more pronounced with the first... The method for obtaining the normalized variation sequences of the absorption peak wavelengths of other unstable components with the highest similarity to the normalized variation sequence of the absorption peak wavelength of the first unstable component is as follows: Calculate the... The normalized variation sequence of the absorption peak wavelength of one unstable component is summed with the sum of the absolute values ​​of the differences between corresponding elements in the normalized variation sequences of the absorption peak wavelengths of all other unstable components. The normalized variation sequence of the absorption peak wavelengths of the other unstable component with the smallest sum of the absolute values ​​of the differences is taken as the normalized variation sequence with the highest similarity.

[0069] When the concentration change characteristic of the absorption peak wavelength of the unstable component is greater than or equal to the preset concentration change characteristic threshold, the absorption peak change at the wavelength of the unstable component is considered a solution concentration change. When the concentration change characteristic of the absorption peak wavelength of the unstable component is less than the preset concentration change characteristic threshold, the absorption peak at the wavelength of the unstable component is considered a real unstable change affected by noise, and is identified as a risk absorption peak wavelength. Considering that the maximum difference in the year-on-year change of the absorption peak of the same component can be considered a year-on-year change if it is within 0.2, the corresponding characteristic value boundary after substituting into the exponential function is 0.818. Therefore, this embodiment sets the concentration change characteristic threshold to 0.8. In other embodiments, the implementer can adaptively adjust the concentration change characteristic threshold according to the noise level and data distribution characteristics of the actual detection environment.

[0070] S5: Comprehensive analysis of food components based on the labeling results of risk absorption peak wavelengths of different subsamples.

[0071] Specifically, for any subsample, if its spectral curve contains an absorption peak wavelength marked as a risk absorption peak, it indicates that the absorbance value of the peak data point at that wavelength is affected by noise and lacks reliability as a basis for judgment. In this case, the absorbance value of that subsample at that wavelength is discarded and not used as reference data for subsequent analysis. Simultaneously, from other simultaneously detected subsamples, target subsamples that are not marked as risk absorption peak wavelengths at that wavelength are selected. A comprehensive analysis is performed based on the absorbance values ​​of all target subsamples at that wavelength to determine the true absorption characteristics of the food component at that wavelength. If the same wavelength in all subsamples is marked as a risk absorption peak wavelength, the detection data at that wavelength is deemed invalid, and a re-acquisition command is issued. This eliminates the interference of noise on absorption peaks, allowing analysts to obtain more accurate food component analysis conclusions based on the spectrophotometric detection results of wavelength channels without risk markers and the comprehensive analysis results of multiple subsamples.

[0072] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise expressly and specifically defined.

[0073] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A spectrophotometric multi-channel data synchronous acquisition and analysis system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it performs the following steps: The absorbance data of multiple sub-samples after the total solution of the tested food is divided at multiple acquisition times are obtained, and multi-frame spectral curves of each sub-sample are generated. For any subsample, extract the peak data points in each frame of the spectral curve, calculate the characteristic degree of the component absorption peak based on the peak retention characteristics of each wavelength in historical multi-frame data, and select the component absorption peak wavelengths based on the characteristic degree of the component absorption peak. The characteristic degree of the absorption peak of the component satisfies the expression: In the formula, Indicates the first The degree of characteristic of the component absorption peaks at each wavelength; Indicates the first The total number of frames in which each wavelength appears as a peak data point across multiple historical frames; Indicates the first The total number of wave crest chains formed by each wavelength across multiple historical frames; Indicates the first The first wavelength The number of consecutive frames contained in a wave crest chain; Represents the normalization function; For each component absorption peak wavelength, the degree of change in multiple frames is calculated based on the difference in absorbance value and the difference in waveform lateral span of the peak data point at the absorption peak wavelength in historical multi-frame data. The unstable component absorption peak wavelength is identified based on the degree of change in multiple frames. The degree of change across multiple frames satisfies the expression: In the formula, Indicates the first The degree of multi-frame variation in the wavelength of the absorption peak of each component; Indicates the first The total number of frames in which the absorption peak wavelength of each component exists in historical multi-frame data points; and They represent the first The wavelength of the absorption peak of each component is the first in multiple historical frames. The and the first Absorbance values ​​of each peak data point; and They represent the first The wavelength of the absorption peak of each component is the first in multiple historical frames. The and the first The waveform lateral span at each peak data point, wherein the waveform lateral span is the lateral distance from the peak data point to the adjacent trough point on both sides; For each unstable component's absorption peak wavelength, an absorbance variation sequence is constructed and normalized to obtain a normalized variation sequence. Concentration variation characteristics are calculated based on the similarity between these normalized variation sequences, and these characteristics satisfy the expression: In the formula, Indicates the first The concentration variation characteristics of the absorption peak wavelength of the unstable component; Indicates the number of change values ​​within the normalized change sequence; Indicates the first In the normalized variation sequence of the absorption peak wavelength of the unstable component, the first... One change value; Indicates the relationship with the first The normalized variation sequence of the absorption peak wavelength of the unstable component with the highest similarity to the normalized variation sequence of the absorption peak wavelength of other unstable components is the first one. One change value; Represents an exponential function with the natural constant as its base; Risk absorption peak wavelengths are marked based on concentration change characteristics, including: in response to the concentration change characteristics of unstable component absorption peak wavelengths being less than a preset concentration change characteristic threshold, they are determined to be risk absorption peak wavelengths. A comprehensive analysis was conducted based on the labeling results of different subsamples.

2. The spectrophotometric multi-channel data synchronous acquisition and analysis system according to claim 1, characterized in that, The process involves acquiring multi-wavelength absorbance data at multiple acquisition times for multiple sub-samples after dividing the total solution of the tested food, and generating multi-frame spectral curves for each sub-sample, including: The food to be tested is crushed and dissolved in a solvent to extract the target components. The insoluble solid residue is removed by filtration to obtain the test liquid. The test liquid is homogenized to form a total solution, and the total solution is divided into multiple subsamples. A broadband light source emits continuous light to illuminate the transmitted subsample. A photodetector receives the transmitted monochromatic light of different wavelengths and converts it into an electrical signal to obtain the transmitted light intensity of each wavelength channel. The absorbance is calculated based on the transmitted light intensity and the initial incident light intensity. The spectral curve at the acquisition time is plotted with wavelength as the abscissa and absorbance as the ordinate.

3. The spectrophotometric multi-channel data synchronous acquisition and analysis system according to claim 1, characterized in that, The method for obtaining the peak chain is as follows: Regarding the first The wavelength, in the historical multi-frame temporal arrangement, will be the first wavelength. Adjacent frames with consecutive peak data points of a certain wavelength are connected to form a peak chain. A peak chain contains at least two consecutive frames of the first wavelength. If the peak data point of the i-th wavelength, If a wavelength has peak data points in multiple consecutive frames, then these consecutive frames together form a peak chain.

4. The spectrophotometric multi-channel data synchronous acquisition and analysis system according to claim 1, characterized in that, The wavelengths of component absorption peaks are selected based on the characteristic degree of the component absorption peaks, including: If the characteristic degree of the component absorption peak at a given wavelength is greater than or equal to a preset threshold for the characteristic degree of the component absorption peak, then the wavelength is determined to be the component absorption peak wavelength of the food being tested.

5. The spectrophotometric multi-channel data synchronous acquisition and analysis system according to claim 1, characterized in that, The identification of unstable component absorption peak wavelengths based on the degree of change across multiple frames includes: In response to a multi-frame variation in the wavelength of a component absorption peak that is greater than or equal to a preset multi-frame variation threshold, the wavelength of that component absorption peak is taken as the wavelength of the unstable component absorption peak.

6. The spectrophotometric multi-channel data synchronous acquisition and analysis system according to claim 1, characterized in that, The process of constructing an absorbance variation sequence for each unstable component's absorption peak wavelength includes: The absorbance values ​​of each unstable component absorption peak wavelength in all historical frames are extracted and arranged in the order of acquisition time to form an absorbance sequence. The difference between adjacent absorbance values ​​in the absorbance sequence is calculated and the absolute value is taken. The absolute values ​​of the differences between all adjacent absorbance values ​​form an absorbance change sequence.

7. The spectrophotometric multi-channel data synchronous acquisition and analysis system according to claim 1, characterized in that, The comprehensive analysis based on the labeling results of different subsamples includes: For any given subsample, if its spectral curve contains an absorption peak wavelength marked as a risk absorption peak wavelength, the absorbance value of that subsample at that risk absorption peak wavelength is discarded. From other subsamples detected simultaneously, target subsamples that are not marked as risk absorption peak wavelengths are selected. Based on a comprehensive analysis of the absorbance values ​​of all target subsamples at the risk absorption peak wavelengths, the true absorption characteristics of the food components at those wavelengths are determined. If the same wavelength in all subsamples is marked as a risk absorption peak wavelength, the detection data at that wavelength is deemed invalid, and a re-acquisition command is output.

Images