A method and device for quantitative analysis of components of a multispectral mixture
By using a multispectral mixture component quantitative analysis method, a matrix is created using spectral information and then integrated and solved by multiple regression function inversion. This solves the problems of long time consumption and low accuracy in traditional methods, and achieves efficient and accurate component content detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- AOPU TIANCHENG (WUHAN) OPTOELECTRONICS TECHNOLOGY CO LTD
- Filing Date
- 2025-12-23
- Publication Date
- 2026-05-29
Smart Images
Figure CN121384840B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of information technology, and in particular to a method and apparatus for quantitative analysis of components in a multispectral mixture. Background Technology
[0002] Quantitative analysis of chemical mixtures has wide and important applications in chemical detection, food safety, public safety, and bioimaging. However, traditional methods for quantitative analysis of mixture components mostly employ chemical methods such as gas chromatography and chemical reactions, which are time-consuming and have relatively low accuracy. Summary of the Invention
[0003] The purpose of this application is to provide a method and apparatus for quantitative analysis of components in multispectral mixtures, thereby improving the efficiency and accuracy of spectral detection. The specific technical solution is as follows:
[0004] A first aspect of this application provides a method for quantitative analysis of components in a multispectral mixture, the method comprising:
[0005] Obtain the spectral information of the target mixture; create a spectral output matrix based on the spectral information;
[0006] Based on the spectral output matrix, the spectral characteristic peaks are integrated to obtain the spectral integral matrix;
[0007] Based on the spectral integral matrix, a multivariate regression function corresponding to the component content and light intensity in the mixture is created.
[0008] The content information of each component in the mixture is obtained by inverting the multivariate regression function.
[0009] In one possible implementation, the step of integrating the spectral characteristic peaks based on the spectral output matrix to obtain the spectral integral matrix includes:
[0010] Identify spectral characteristic peaks based on the spectral output matrix;
[0011] By using a preset numerical integration formula:
[0012]
[0013] Integrating the spectral characteristic peaks yields the spectral integral matrix, where I j Represents the spectral integral matrix. This represents the characteristic peak matrix.
[0014] In one possible implementation, the step of creating a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix includes:
[0015] Based on the spectral integral matrix, a multiple regression function corresponding to the component content and light intensity in the mixture is created:
[0016]
[0017] Among them, I i S represents the area of the i-th characteristic peak. i This represents the set of known contributing sources to the i-th characteristic peak. C represents the contribution coefficient corresponding to the j-th signal mode of the m-th component. m w represents the content of the m-th component. i b represents the proportionality coefficient of the peak intensity of the i-th feature peak in the fit. i This represents the bias of the unknown contribution source and its influence in fitting the i-th characteristic peak.
[0018] In one possible implementation, obtaining the spectral information of the target mixture includes: obtaining multiple sets of spectral information corresponding to multiple samples of the target mixture; and creating a spectral matrix based on the multiple sets of spectral information.
[0019] The step of creating a spectral output matrix based on the spectral information includes: performing baseline correction on the spectral matrix using a preset adaptive iterative weighted penalized least squares method to obtain the spectral output matrix.
[0020] In one possible implementation, the step of creating a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix; and inverting the multiple regression function to obtain the content information of each component in the mixture, includes:
[0021] Using the spectral integral matrix and a pre-created linear model:
[0022]
[0023] Inversion is performed to obtain the content information of each component in the mixture, where w is the preset proportional coefficient matrix, C is the constructed content feature matrix, b is the residual of the influence of unknown components on the characteristic peak intensity, and I is the spectral integral matrix.
[0024] A second aspect of this application provides a multispectral mixture component quantitative analysis device, the device comprising:
[0025] A spectral acquisition module is used to acquire the spectral information of the target mixture; and to create a spectral output matrix based on the spectral information.
[0026] The spectral integration module is used to integrate the spectral characteristic peaks based on the spectral output matrix to obtain the spectral integration matrix;
[0027] The function creation module is used to create a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix.
[0028] The function solving module is used to invert and solve the multivariate regression function to obtain the content information of each component in the mixture.
[0029] In one possible implementation, the spectral integration module is specifically used to identify spectral characteristic peaks based on the spectral output matrix; through a preset numerical integration formula:
[0030]
[0031] Integrating the spectral characteristic peaks yields the spectral integral matrix, where I j Represents the spectral integral matrix. This represents the characteristic peak matrix.
[0032] In one possible implementation, the function creation module is specifically used to create a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix:
[0033]
[0034] Among them, I i S represents the area of the i-th characteristic peak. i This represents the set of known contributing sources to the i-th characteristic peak. C represents the contribution coefficient corresponding to the j-th signal mode of the m-th component. m w represents the content of the m-th component. i b represents the proportionality coefficient of the peak intensity of the i-th feature peak in the fit. i This represents the bias of the unknown contribution source and its influence in fitting the i-th characteristic peak.
[0035] In one possible implementation, the spectral acquisition module is specifically used to acquire multiple sets of spectral information corresponding to multiple samples of the target mixture; create a spectral matrix based on the multiple sets of spectral information; and perform baseline correction on the spectral matrix using a preset adaptive iterative weighted penalized least squares method to obtain the spectral output matrix.
[0036] In one possible implementation, the step of creating a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix; and inverting the multiple regression function to obtain the content information of each component in the mixture, includes:
[0037] Using the spectral integral matrix and a pre-created linear model:
[0038]
[0039] Inversion is performed to obtain the content information of each component in the mixture, where w is the preset proportional coefficient matrix, C is the constructed content feature matrix, b is the residual of the influence of unknown components on the characteristic peak intensity, and I is the spectral integral matrix.
[0040] Another aspect of the embodiments of this application also provides an electronic device, including:
[0041] Memory, used to store computer programs;
[0042] The processor, when executing the program stored in the memory, implements any of the above-mentioned methods for quantitative analysis of components in a multispectral mixture.
[0043] In another aspect of the embodiments of this application, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements any of the above-described methods for quantitative analysis of components in a multispectral mixture.
[0044] In another aspect of the embodiments of this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the above-described methods for quantitative analysis of multispectral mixture components.
[0045] Beneficial effects of the embodiments in this application:
[0046] This application provides a method and apparatus for quantitative analysis of components in a multispectral mixture. The method includes: acquiring spectral information of a target mixture; creating a spectral output matrix based on the spectral information; integrating the spectral characteristic peaks based on the spectral output matrix to obtain a spectral integral matrix; creating a multivariate regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix; and inverting and solving the multivariate regression function to obtain the content information of each component in the mixture. Through the scheme of this application, after acquiring the spectral information of the target mixture, a matrix can be created based on the spectral information, and then the characteristic peaks can be integrated based on the matrix, thereby avoiding interference and improving detection efficiency. Finally, by constructing and inverting the multivariate regression function, the content information of each component in the mixture can be obtained, thereby improving detection accuracy.
[0047] Of course, implementing any product or method of this application does not necessarily require achieving all of the advantages described above at the same time. Attached Figure Description
[0048] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other embodiments can be obtained based on these drawings.
[0049] Figure 1 A schematic flowchart of a method for quantitative analysis of multispectral mixture components provided in this application embodiment;
[0050] Figure 2 A schematic diagram of a multispectral mixture component quantitative analysis device provided in an embodiment of this application;
[0051] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation
[0052] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art based on this application are within the scope of protection of this application.
[0053] A first aspect of the embodiments of this application provides a method for quantitative analysis of components in a multispectral mixture, see [link to relevant documentation]. Figure 1 , Figure 1 This is a schematic flowchart of a method for quantitative analysis of multispectral mixture components provided in an embodiment of this application. The method includes:
[0054] Step S11: Obtain the spectral information of the target mixture; create a spectral output matrix based on the spectral information;
[0055] Step S12: Based on the spectral output matrix, integrate the spectral characteristic peaks to obtain the spectral integral matrix;
[0056] Step S13: Based on the spectral integral matrix, create a multiple regression function corresponding to the component content and light intensity in the mixture;
[0057] Step S14: Invert the multivariate regression function to obtain the content information of each component in the mixture.
[0058] Corresponding to step S11 above, when acquiring the spectral information of the target mixture, multiple samples of the target mixture can be collected, and then multiple spectral acquisitions can be performed on each sample to obtain multiple sets of spectral information. A matrix can then be created based on these multiple sets of spectral information. In one possible implementation, acquiring the spectral information of the target mixture includes: acquiring multiple sets of spectral information corresponding to multiple samples of the target mixture; and creating a spectral matrix based on the multiple sets of spectral information. When creating the spectral output matrix based on the spectral information, the matrix corresponding to the spectral information can be preprocessed to obtain a preprocessed matrix, which is the spectral output matrix. Specifically, the preprocessing may include baseline correction. In one example, to acquire the spectral information of the target mixture, a certain number of chemical mixture samples can be collected, and then, based on a spectrometer and combining the chemical mechanisms of the mixture components and spectra, appropriate mixture spectra can be acquired to form a light intensity output (dependent variable) matrix. .in, Indicates the number of samples collected; This represents the wavenumber or number of wavelengths in the sample spectrum. When creating the spectral output matrix, different types of spectra (such as ultraviolet-visible, near-infrared, Raman, etc.) may be affected by background signals, with Raman spectra often exhibiting fluorescence background. To eliminate background interference, an adaptive iterative weighted penalized least squares method can be used to perform baseline correction on the acquired spectra, subtracting background components including fluorescence, thereby obtaining a corrected spectral output matrix. In one possible implementation, creating the spectral output matrix based on the spectral information includes: performing baseline correction on the spectral matrix using a preset adaptive iterative weighted penalized least squares method to obtain the spectral output matrix.
[0059] Corresponding to step S12 above, the spectral characteristic peaks are integrated based on the spectral output matrix to obtain the spectral integral matrix. This reduces the output dimension and also reduces noise interference. Specifically, the output matrix can be... Numerical integration is performed on the spectral characteristic peaks corresponding to specific molecular groups in the spectrum.
[0060] Corresponding to step S13 above, a multiple regression function is created based on the spectral integral matrix, relating the component content and light intensity in the mixture. Specifically, a corresponding multiple regression function can be created using a preset multiple regression model. This multi-source regression model can represent the relationship between the component content and light intensity in the mixture.
[0061] Corresponding to step S14 above, when the multivariate regression function is inverted and solved to obtain the content information of each component in the mixture, the nonlinear equation system can be solved by the constrained quasi-Newton method, thereby inverting and reconstructing the content of each component in the mixture.
[0062] In one possible implementation, the step of creating a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix, and then inverting and solving the multiple regression function to obtain the content information of each component in the mixture, includes: using the spectral integral matrix and a pre-created linear model.
[0063]
[0064] Inversion is performed to obtain the content information of each component in the mixture, where w is the preset proportional coefficient matrix, C is the constructed content feature matrix, b is the residual of the influence of unknown components on the characteristic peak intensity, and I is the spectral integral matrix.
[0065] As can be seen, the method of this application can be used to create a matrix after obtaining the spectral information of the target mixture, and then integrate the characteristic peaks based on the matrix to avoid affecting and improve detection efficiency. Finally, the content information of each component in the mixture can be obtained by constructing and inverting the multivariate regression function, thereby improving the detection accuracy.
[0066] In one possible implementation, the step of integrating the spectral characteristic peaks based on the spectral output matrix to obtain the spectral integral matrix includes:
[0067] Identify spectral characteristic peaks based on the spectral output matrix;
[0068] By using a preset numerical integration formula:
[0069]
[0070] Integrating the spectral characteristic peaks yields the spectral integral matrix, where I j Represents the spectral integral matrix. This represents the characteristic peak matrix. j corresponds to the common molecular group vibrational modes in the components to be predicted in the mixture. Assuming there are r types of molecular group vibrational modes j in the mixture components, the output dimension is reduced from p to r dimensions, and the output matrix becomes... .
[0071] In one possible implementation, the step of creating a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix includes:
[0072] Based on the spectral integral matrix, a multiple regression function corresponding to the component content and light intensity in the mixture is created:
[0073]
[0074] Among them, I i S represents the area of the i-th characteristic peak. iThis represents the set of known contributing sources to the i-th characteristic peak. C represents the contribution coefficient corresponding to the j-th signal mode of the m-th component. m w represents the content of the m-th component. i b represents the proportionality coefficient of the peak intensity of the i-th feature peak in the fit. i This represents the bias of the unknown contribution source and its influence in fitting the i-th characteristic peak.
[0075] To illustrate the multi-source regression function in the embodiments of this application, the following description is provided in conjunction with the creation process of the model corresponding to this function, including:
[0076] Traditional spectroscopic-based quantitative detection models for mixture components are primarily established based on spectra. To content The algorithm uses a partial least squares (PLS) model, and then uses collected samples for parameter training and fitting. This algorithm is purely data-driven, lacks sufficient model interpretability, and does not fully utilize the advantages of clearly defined chemical mechanisms and unique fingerprint properties. Based on these considerations, the starting point for the algorithm design in this application is the clearly defined chemical mechanisms and fingerprint characteristics of the spectrum. Specifically, when a sample is exposed to ultraviolet-visible light, near-infrared light, Raman laser, etc., due to transmission, reflection, or Raman scattering effects, certain vibrational modes of specific molecular groups in the sample will absorb the laser at specific frequencies. This specific absorption frequency corresponds to the wavenumber or wavelength. The greater the intensity of the characteristic peak corresponding to a specific wavenumber or wavelength in the sample, the more photons are absorbed at the frequency caused by the vibrational mode of that specific molecular group, which means a stronger signal for that specific molecular group vibrational mode. Each characteristic peak and its intensity correspond to a specific molecular group vibrational mode and its signal intensity; this is the unique fingerprint characteristic of the spectrum.
[0077] Specifically, the higher the concentration of a specific molecular group in a sample, the stronger the corresponding characteristic peak will be, and the following proportional relationship can be established:
[0078] (1)
[0079] Among them, I j Indicates the vibrational modes of molecular groups in the sample The corresponding characteristic peak intensity, which is the spectral integral matrix, S j Indicates the vibrational modes of molecular groups in the sample The signal strength.
[0080] Based on the chemical mechanism of molecular group vibrational modes, the signal intensity is affected by both the concentration of the molecular group and the molecular vibrational mode, thus the following mathematical relationship can be obtained:
[0081] (2)
[0082] Among them, C j Represents the vibrational modes of molecular groups The concentration of the corresponding molecular groups, A j Indicates the vibrational mode of the molecular group. The molecular structure influences it.
[0083] Furthermore, since the objective of this application is to determine the content of components, rather than the content of molecular groups, when several components share the same molecular group, theoretically the content of that molecular group should be the weighted sum of the contents of all components. Here, the weight represents the contribution of each component to the content of that molecular group, i.e., the ratio of the number of that molecular group in each molecule. Therefore, C in formula (2)... j Satisfy the following mathematical relationship:
[0084] (3)
[0085] in, Represents the vibrational modes of molecular groups in the m-th component. The number of corresponding molecular groups, C m This indicates the content of the m-th component.
[0086] Similarly, molecular structure affects A j It also satisfies the following relationship:
[0087] (4)
[0088] in, This represents the vibrational mode of the molecular groups of the m-th component. The resulting structural impact, C m This indicates the content of the m-th component.
[0089] Furthermore, by integrating formulas (2), (3), and (4), we can obtain the following results:
[0090] (5)
[0091] Combining equations (5) and (2), we obtain the following relationship:
[0092] (6)
[0093] Since the number of different molecular groups in different components is readily available, but the structural influence of different components on different molecular groups is uncertain, this application combines both into a single influence factor:
[0094] (7)
[0095] This impact factor This is called the peak intensity contribution coefficient of component m to the vibrational mode j of the molecular group. Therefore:
[0096] (8)
[0097] In this application, the ratio of the reference normalized spectra of each component is used as the peak intensity contribution factor. Further refinement of the model, unifying molecular vibrational mode j as signal mode j, allows for the establishment of a preliminary model for detecting the content of components in a mixture:
[0098] (9)
[0099] I i S represents the area of the i-th characteristic peak (absorption peak / Raman peak), which can be a sharp peak, an overlapping peak, or a broad peak. In a physicochemical sense, it represents the intensity of light that produces energy absorption or transfer at that peak position; i Represents all known components (component m, signal mode j, component content C) for the i-th characteristic peak. m A collection of contribution sources; Represents the m-th component's... The contribution coefficients corresponding to each signal mode are assigned based on the standard normalized spectra of the known components in the mixture; C m w represents the content of the m-th component; i The proportionality coefficient used to fit the peak intensity of the i-th characteristic peak, b i Used to fit the unknown (component m, signal mode j, component content C) in the i-th characteristic peak. m The bias of contribution and influence of the contributing source. Compared with the existing model from light intensity I to content C, the model in this application is based on chemical mechanism and establishes a hybrid model that combines mechanism and data-driven approach from the content C of the mixture components to light intensity I.
[0100] To illustrate the solutions of the embodiments of this application, the following description is provided in conjunction with a specific embodiment, including:
[0101] STEP 1: Data Acquisition and Processing; Three-dimensional cell mixture component determination is used as a preferred example of this invention. The three-dimensional cell is composed of a mixture of five components. Based on biochemical methods, the content of the five components at each voxel point of the collected cell samples is determined, forming 1K mixture samples, which constitute a component content input (independent variable) matrix. Based on hyperspectral bioimaging technology, three-dimensional volumetric hyperspectral data of cells were acquired, forming approximately 1K mixed spectra, each with a wavenumber length of 500, constituting a light intensity output (dependent variable) matrix. .
[0102] STEP 2: Data preprocessing; Based on the adaptive iterative weighted penalized least squares method (AirPLS algorithm), the acquired spectra are preprocessed. Perform baseline correction and subtract the fluorescence background (substrate).
[0103] STEP3: Spectral Feature Engineering; Process the output matrix The characteristic peak intensities corresponding to specific molecular groups are numerically integrated. Based on this, the output dimension is reduced from 500 dimensions to 5 dimensions, at which point the output matrix... .
[0104] STEP 4: Establish a hybrid model combining mechanism-driven and data-driven approaches; Based on the mechanism-driven and data-driven hybrid model proposed in this invention, the following model is established:
[0105]
[0106] As can be seen, compared to the PLS model which calculates the relationship between light intensity I and component content C, this model is based on chemical mechanisms and establishes a multiple regression model that calculates the relationship between component content C and light intensity I.
[0107] The peak intensity contribution coefficient is calculated based on the numerical integral of the characteristic peak intensity of the normalized spectrum. The possible values are as follows:
[0108]
[0109] STEP 5: Model Parameter Training and Optimization; During training, directly create a weighted sum of logarithmic features C to establish the following linear model:
[0110]
[0111] in, , representing the proportional coefficient matrix, requires training and optimization using samples. This represents the concentration feature matrix constructed based on the mechanism. , representing the residual (bias) representing the influence of the unknown component on the bee strength. The output matrix represents the intensity of spectral characteristic peaks. Based on 1K collected cell mixture samples, the parameters w and b are optimized using the least squares method.
[0112] STEP 6: Concentration Inversion and Model Evaluation; Spectral data of the collected cell mixture samples Leave-one-out cross-validation was used, and data preprocessing and feature engineering were performed in steps 2 and 3 to obtain the processed Raman peak features. Then, The input is fed into the model trained in step 5, and the constrained quasi-Newton method is used to solve the nonlinear equations, thereby reconstructing the content of the five components of the cell. The cross-validation results are shown in the table below:
[0113] Table 1 shows the errors in the component results corresponding to the mechanism modeling.
[0114]
[0115] Experimental results show that mechanism modeling achieved an average score of over 0.90 in predicting the content of components in cell mixtures, indicating good performance.
[0116] Specifically, compared to purely data-driven models and deep learning-based modeling methods, this application fully utilizes the physicochemical properties of spectra, combining mechanistic modeling with data-driven approaches to create a more interpretable model. This results in the following advantages: higher model detection accuracy. The model fully considers the chemical mechanisms of the spectrum and calculates the numerical integral of characteristic peaks based on the reference normalized spectra of each component in the mixture, using this integral as a peak intensity contribution coefficient to distinguish the different degrees of contribution of different components to the peak intensity, thereby improving the prediction accuracy of each component's concentration. It is more suitable for expanded applications in the detection of substance composition. This model can be used to model the prediction of the content of any specific mixture component, while the proportionality coefficient between component content and peak intensity, and the influence of unknown components on specific characteristic peaks, can be solved using a data-driven approach. If the modeling method is based on light intensity to component content, the influence of unknown mixture components is difficult to handle. Therefore, it can be extended to the detection of more substance composition contents. It performs better for detecting mixture components with large dimensional differences. Many mixture components exhibit this situation: one component's content is 50%–70%, while another component is only 1%–3%, with a large difference in dimensional range. Without using deep learning modeling methods that rely on large-scale data, purely machine learning-based modeling performs poorly in predicting outputs with significant dimensional differences. The mechanistic model established in this paper, however, performs better in predicting components with low content. It has lower requirements for sample size and scale, making it suitable for small-sample modeling and alleviating the data challenges associated with using deep learning models in spectral vertical domains.
[0117] A second aspect of this application provides a multispectral mixture component quantitative analysis device, see [link to relevant documentation]. Figure 2 , Figure 2 This is a schematic diagram of a multispectral mixture component quantitative analysis device provided in an embodiment of this application. The device includes:
[0118] The spectral acquisition module 201 is used to acquire the spectral information of the target mixture and create a spectral output matrix based on the spectral information.
[0119] The spectral integration module 202 is used to integrate the spectral characteristic peaks based on the spectral output matrix to obtain the spectral integration matrix;
[0120] The function creation module 203 is used to create a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix.
[0121] The function solving module 204 is used to perform inversion solving on the multivariate regression function to obtain the content information of each component in the mixture.
[0122] In one possible implementation, the spectral integration module is specifically used to identify spectral characteristic peaks based on the spectral output matrix; through a preset numerical integration formula:
[0123]
[0124] Integrating the spectral characteristic peaks yields the spectral integral matrix, where I j Represents the spectral integral matrix. This represents the characteristic peak matrix. j corresponds to the common molecular group vibrational modes in the components to be predicted in the mixture. Assuming there are r types of molecular group vibrational modes j in the mixture components, the output dimension is reduced from p to r dimensions, and the output matrix becomes... .
[0125] In one possible implementation, the function creation module is specifically used to create a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix:
[0126]
[0127] Among them, I i S represents the area of the i-th characteristic peak. i This represents the set of known contributing sources to the i-th characteristic peak. C represents the contribution coefficient corresponding to the j-th signal mode of the m-th component. m w represents the content of the m-th component. i b represents the proportionality coefficient of the peak intensity of the i-th feature peak in the fit. i This represents the bias of the unknown contribution source and its influence in fitting the i-th characteristic peak.
[0128] In one possible implementation, the spectral acquisition module is specifically used to acquire multiple sets of spectral information corresponding to multiple samples of the target mixture; create a spectral matrix based on the multiple sets of spectral information; and perform baseline correction on the spectral matrix using a preset adaptive iterative weighted penalized least squares method to obtain the spectral output matrix.
[0129] In one possible implementation, the step of creating a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix; and inverting the multiple regression function to obtain the content information of each component in the mixture, includes:
[0130] Using the spectral integral matrix and a pre-created linear model:
[0131]
[0132] Inversion is performed to obtain the content information of each component in the mixture, where w is the preset proportional coefficient matrix, C is the constructed content feature matrix, b is the residual of the influence of unknown components on the characteristic peak intensity, and I is the spectral integral matrix.
[0133] As can be seen, the scheme of this application can create a matrix based on the spectral information of the target mixture after obtaining the spectral information, and then perform integrals of the characteristic peaks based on the matrix, thereby avoiding interference and improving detection efficiency. Finally, the content information of each component in the mixture can be obtained by constructing and inverting the multivariate regression function, thereby improving the detection accuracy.
[0134] This application also provides an electronic device, such as... Figure 3 As shown, it includes:
[0135] Memory 301 is used to store computer programs;
[0136] When processor 302 executes a program stored in memory 301, it performs the following steps:
[0137] Obtain the spectral information of the target mixture; create a spectral output matrix based on the spectral information;
[0138] Based on the spectral output matrix, the spectral characteristic peaks are integrated to obtain the spectral integral matrix;
[0139] Based on the spectral integral matrix, a multivariate regression function corresponding to the component content and light intensity in the mixture is created.
[0140] The content information of each component in the mixture is obtained by inverting the multivariate regression function.
[0141] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.
[0142] The communication interface is used for communication between the aforementioned electronic devices and other devices.
[0143] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.
[0144] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.
[0145] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores a computer program that, when executed by a processor, implements the steps of any of the above-described methods for quantitative analysis of components in a multispectral mixture.
[0146] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the multispectral mixture component quantitative analysis methods described in the above embodiments.
[0147] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a solid-state drive (SSD), etc.
[0148] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0149] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on its differences from other embodiments. In particular, the embodiments for apparatus, electronic devices, and storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.
[0150] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application are included within the scope of protection of this application.
Claims
1. A method for quantitative analysis of components in a multispectral mixture, characterized in that, The method includes: Obtain the spectral information of the target mixture; create a spectral output matrix based on the spectral information; Based on the spectral output matrix, the spectral characteristic peaks are integrated to obtain the spectral integral matrix; Based on the spectral integral matrix, a multiple regression function corresponding to the component content and light intensity in the mixture is created. The content information of each component in the mixture is obtained by inverting the multivariate regression function. The step of integrating the spectral characteristic peaks based on the spectral output matrix to obtain the spectral integral matrix includes: Identify spectral characteristic peaks based on the spectral output matrix; By using a preset numerical integration formula: Integrating the spectral characteristic peaks yields the spectral integral matrix, where I j Represents the spectral integral matrix. Represents the characteristic peak matrix; The step of creating a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix includes: Based on the spectral integral matrix, a multiple regression function corresponding to the component content and light intensity in the mixture is created: Among them, I i S represents the area of the i-th characteristic peak. i Represents the set of known contribution sources for the i-th characteristic peak. This represents the contribution coefficient corresponding to the j-th signal mode of the m-th component. Represents the vibrational modes of molecular groups in the m-th component. The number of corresponding molecular groups, This represents the vibrational mode of the molecular groups of the m-th component. The resulting structural impact, C m w represents the content of the m-th component. i b represents the proportionality coefficient of the peak intensity of the i-th feature peak in the fit. i This represents the bias of the unknown contribution source and its influence in fitting the i-th characteristic peak.
2. The method according to claim 1, characterized in that, The acquisition of the spectral information of the target mixture includes: Obtain multiple sets of spectral information corresponding to multiple samples of the target mixture; create a spectral matrix based on the multiple sets of spectral information; The step of creating a spectral output matrix based on the spectral information includes: The spectral matrix is baseline-corrected by a preset adaptive iterative weighted penalized least squares method to obtain the spectral output matrix.
3. The method according to claim 2, characterized in that, The process involves creating a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix; and then inverting and solving the multiple regression function to obtain the content information of each component in the mixture, including: Using the spectral integral matrix and a pre-created linear model: Inversion is performed to obtain the content information of each component in the mixture, where w is the preset proportional coefficient matrix, C is the constructed content feature matrix, b is the residual of the influence of unknown components on the characteristic peak intensity, and I is the spectral integral matrix.
4. A multispectral mixture component quantitative analysis device, characterized in that, The device includes: A spectral acquisition module is used to acquire the spectral information of the target mixture; and to create a spectral output matrix based on the spectral information. The spectral integration module is used to integrate the spectral characteristic peaks based on the spectral output matrix to obtain the spectral integration matrix; The function creation module is used to create a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix. The function solving module is used to invert and solve the multivariate regression function to obtain the content information of each component in the mixture; The spectral integration module is specifically used to identify spectral characteristic peaks based on the spectral output matrix; through a preset numerical integration formula: Integrating the spectral characteristic peaks yields the spectral integral matrix, where I j Represents the spectral integral matrix. Represents the characteristic peak matrix; The function creation module is specifically used to create a multiple regression function corresponding to the component content and light intensity in the mixture based on the spectral integral matrix: Among them, I i S represents the area of the i-th characteristic peak. i This represents the set of known contributing sources to the i-th characteristic peak. This represents the contribution coefficient corresponding to the j-th signal mode of the m-th component. Represents the vibrational modes of molecular groups in the m-th component. The number of corresponding molecular groups, This represents the vibrational mode of the molecular groups of the m-th component. The resulting structural impact, C m w represents the content of the m-th component. i b represents the proportionality coefficient of the peak intensity of the i-th feature peak in the fit. i This represents the bias of the unknown contribution source and its influence in fitting the i-th characteristic peak.
5. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method described in any one of claims 1-3.
6. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains a computer program that, when executed by a processor, implements the method described in any one of claims 1-3.