Spectral analysis line screening method and system

Abstract

The invention provides a spectral analysis line screening method and system, and the method comprises the steps: collecting a plasma spectrum based on a laser probe, taking the plasma spectrum as an original spectrum, calculating a corresponding average spectrum, and carrying out the spectrum peak recognition of the average spectrum through wavelet transformation; extracting predetermined features of spectrum peaks from the original spectrum, and performing classified scanning on each spectrum peak based on the predetermined features of the spectrum peaks; the classified scanning results form a column vector matrix, and an accumulated correlation coefficient between column phasors in the column vector matrix is calculated; arranging the accumulative correlation coefficients in an ascending order, and selecting spectrum peaks corresponding to the accumulative correlation coefficients which are ranked in the front as analysis line clusters; extracting predetermined features of the analysis line cluster from the original data, and constructing a multi-spectral peak classification model through a machine learning algorithm based on the features of the analysis line cluster; and screening the spectral analysis lines to be identified through the multi-spectral peak classification model. Through the scheme, the dimension of the spectral peak feature vector can be reduced, the model data volume is reduced, the analysis line screening efficiency is improved, and the classification accuracy is guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of spectral analysis, and in particular relates to a spectral analysis line screening method and system. Background Art

[0002] Laser probe technology, also known as laser-induced breakdown spectroscopy (LIBS), is widely used in rock and mineral detection, deep sea and deep space exploration, environmental pollution detection, industrial production and other fields. In these fields, there are both qualitative analysis and quantitative detection. When using LIBS technology for qualitative analysis, it is usually necessary to screen the analysis line.

[0003] To screen the spectral analysis lines, we must first select a specific number of analysis lines in the spectrum, then extract the spectral intensity of the selected analysis lines from the original spectral data, and then combine it with specific analysis algorithms such as principal component analysis and support vector machines to establish a clustering or classification model to achieve screening. However, in portable and handheld LIBS systems with relatively weak computing power, in order to ensure the accuracy of the analysis results and avoid the problems of excessive computation and low efficiency in the later application of the established analysis model, the amount of modeling and prediction data needs to be as small as possible. Therefore, the selected analysis lines should have strong characterization capabilities and a small number.

[0004] At present, relevant technicians in this field have conducted some research in this regard. For example, the published document "A method for automatic peak finding of LIBS spectrum based on continuous wavelet transform" (Spectroscopy and Spectral Analysis. Vol. 34, No.7, pp1969-1972) mentions a method for peak finding of LIBS spectrum using continuous wavelet transform. In this method, 21 peaks were identified in a LIBS spectrum of Fe3O4 sample (382.274-385.347 nm, 300 pixels) through continuous wavelet transform based on scale parameter and translation parameter optimization, as well as ridge correction and signal-to-noise ratio screening. However, the identification process of this method is relatively complicated and the parameter optimization is difficult. Since the identified peaks are not further evaluated and utilized, their characterization ability and qualitative and quantitative analysis utility cannot be known. Moreover, when screening using the signal-to-noise ratio, background estimation is required for each wavelength. Manually specifying the background has efficiency problems, and automatically identifying the background will introduce evaluation errors.

[0005] At the same time, full spectrum classification is also a feasible line selection and classification scheme. Although the classification method based on the full spectrum has simple steps, the amount of full spectrum data is large, which will inevitably lead to low efficiency in the early modeling and later prediction application of LIBS technology. In addition, the introduction of too many approximate features in the full spectrum data can easily cause the problem of non-positive definiteness of the feature matrix, making modeling more difficult.

[0006] Therefore, how to accurately and efficiently select analytical lines with strong characterization capabilities from the original spectrum is a key issue that needs to be solved urgently. Summary of the invention

[0007] In view of this, an embodiment of the present invention provides a spectral analysis line screening method and system for solving the problems of complex process, low accuracy and low efficiency in current analysis line screening.

[0008] In a first aspect of an embodiment of the present invention, a spectral analysis line screening method is provided, comprising: Collect plasma spectra of a preset number of samples based on a laser probe; Taking the plasma spectrum as the original spectrum, calculating the average spectrum corresponding to the original spectrum, and performing spectrum peak identification on the average spectrum by wavelet transformation; Extracting predetermined features of spectral peaks from the original spectrum, and classifying and scanning each spectral peak based on the predetermined features of the spectral peaks, wherein the predetermined features are used to characterize information features of the spectral peaks; The classification scanning results are formed into a column vector matrix, and the cumulative correlation coefficient between any column vector in the column vector matrix and all other column phasors is calculated; Arrange the cumulative correlation coefficients in ascending order, and select the spectrum peaks corresponding to the cumulative correlation coefficients with a preset number of first order as the analysis line clusters; Extracting predetermined features of the analysis line clusters from the original spectrum, and constructing a multi-spectral peak classification model through a machine learning algorithm based on the predetermined features of the analysis line clusters; The spectral analysis lines to be identified are screened through the multi-spectral peak classification model.

[0009] In a second aspect of an embodiment of the present invention, a spectral analysis line screening system is provided, comprising: A spectrum acquisition module, used for acquiring plasma spectra of a preset number of samples based on a laser probe; A spectrum peak identification module, used to take the plasma spectrum as the original spectrum, calculate the average spectrum corresponding to the original spectrum, and perform spectrum peak identification on the average spectrum by wavelet transformation; A spectrum peak classification module is used to extract predetermined features of spectrum peaks from the original spectrum, and classify and scan each spectrum peak based on the predetermined features of the spectrum peaks, wherein the predetermined features are used to characterize the information features of the spectrum peaks; A spectrum peak selection module is used to form a column vector matrix from the classification scanning results, calculate the cumulative correlation coefficients between any column vector in the column vector matrix and all other column vectors, arrange the cumulative correlation coefficients in ascending order, and select the spectrum peaks corresponding to the cumulative correlation coefficients with a preset number of rankings as analysis line clusters; A model building module is used to extract predetermined features of the analysis line clusters from the original spectrum, and to build a multi-spectral peak classification model based on the predetermined features of the analysis line clusters through a machine learning algorithm; The analysis line screening module is used to screen the spectral analysis lines to be identified through a multi-spectral peak classification model.

[0010] In a third aspect of an embodiment of the present invention, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in the first aspect of the embodiment of the present invention when executing the computer program.

[0011] In a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method provided in the first aspect of the embodiment of the present invention are implemented.

[0012] In the embodiment of the present invention, the peak features in the original spectrum are extracted to classify and scan each peak, the cumulative correlation coefficient between the peak feature vectors is calculated, and the peak is selected as the analysis line cluster according to the cumulative correlation coefficient, and then a multi-peak classification model is constructed to screen the analysis line, which can not only simplify the peak feature vector and reduce the amount of data in the modeling process, but also ensure the accuracy of the spectral analysis line screening and improve the calculation efficiency of the model. At the same time, it avoids the problems of human interference, error introduction, modeling difficulties and low efficiency in practical application in the traditional spectral analysis line screening. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0014] Figure 1 A schematic flow chart of a spectrum analysis line screening method provided by one embodiment of the present invention; Figure 2 A schematic diagram of a spectrum peak identification result provided by an embodiment of the present invention; Figure 3 A schematic diagram of calculating cumulative correlation coefficients provided by one embodiment of the present invention; Figure 4 A schematic diagram of a curve showing the relationship between the number of analysis lines and the classification accuracy provided by an embodiment of the present invention; Figure 5 A schematic structural diagram of a spectrum analysis line screening system provided by one embodiment of the present invention; Figure 6 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0015] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0016] It should be understood that the term "including" and other similar expressions in the specification or claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions, such as a process, method, system, or device including a series of steps or units is not limited to the listed steps or units. In addition, "first" and "second" are used to distinguish different objects, not to describe a specific order.

[0017] See also Figure 1 , a schematic flow chart of a spectral analysis line screening method provided by an embodiment of the present invention includes: S101, collecting plasma spectra of a preset number of samples based on a laser probe; The laser probe, also known as the laser induced breakdown spectrometer, is a material composition analyzer based on the laser induced breakdown spectroscopy technology. The laser induced breakdown spectroscopy uses high-energy lasers to excite the sample surface to produce a high-temperature, high-electron density plasma, and analyzes the ion or atomic spectrum of the plasma-emitted light. The plasma spectrum is a laser probe of the electromagnetic radiation spectrum from infrared to vacuum ultraviolet bands emitted from the inside of the plasma.

[0018] The plasma spectra of a predetermined number of samples are obtained based on the laser probe. For the predetermined number of samples, multiple collection points at different positions can be set on the samples, and each collection point collects multiple spectra to obtain multiple plasma spectra corresponding to each sample.

[0019] For example, 12 igneous rock samples are spectra collected, each sample is 1, 10 points are collected at different positions of each sample, 10 spectra are collected at each point, and the number of pixels of each spectrum is 4096. Therefore, the total number of plasma spectra of the samples is 12×10×10, and the original spectrum data matrix can be expressed as N 4096ⅹ1200 .

[0020] S102, taking the plasma spectrum as the original spectrum, calculating the average spectrum corresponding to the original spectrum, and performing spectral peak identification on the average spectrum by wavelet transform; The average spectrum is the result obtained by averaging the spectrum data of multiple samples. By averaging the spectrum values ​​of multiple samples at each wavelength point, an average spectrum can be obtained, thereby reducing noise, reducing the difference between samples, and extracting common information. Wavelet transform is a mathematical analysis method that decomposes the spectral signal into a series of wavelet functions, thereby analyzing the spectral signal at different scales.

[0021] The original spectra are averaged to obtain the average spectrum, and the average spectrum is decomposed by wavelet transform to obtain the peak information in the spectrum.

[0022] For example, Figure 2 As shown, using the Matlab function mean(N 4096，1200, 2) The original spectral data are averaged to obtain the average spectrum, and then the wavelet transform function cwt(y, scales, 'mexh') is used to identify the spectral peaks based on wavelet changes. In this way, 282 spectral peaks in the average spectrum can be identified, which are recorded as P 282ⅹ2 .

[0023] S103, extracting predetermined features of spectrum peaks from the original spectrum, and classifying and scanning each spectrum peak based on the predetermined features of the spectrum peaks, wherein the predetermined features are used to characterize information features of the spectrum peaks; The classification scan refers to classifying the spectrum peaks respectively through the classification model and outputting the classification labels.

[0024] Specifically, the predetermined features of a single wavelength spectrum peak are extracted from the original spectrum to obtain the one-dimensional spectrum predetermined features corresponding to each spectrum peak, and after the one-dimensional spectrum predetermined features are divided into a training set and a test set in proportion, a single spectrum peak classification model is trained and tested, and the single spectrum peak classification model is a linear discriminant analysis model; Based on the trained single peak classification models, K-fold cross validation is performed on the spectral data in the test set to obtain the peak classification results of the original spectrum by each single peak classification model, where each peak category corresponds to a predicted label value.

[0025] The predetermined feature is spectral intensity, spectral peak area, spectral peak intensity ratio or spectral peak internal standard intensity value. For example, spectral intensity and spectral peak intensity ratio of spectral peaks of different wavelengths are extracted from the original spectrum.

[0026] For spectral peaks of different wavelengths, corresponding single spectral peak classification models are established and trained. For example, if there are 282 spectral peaks, 282 single spectral peak classification models need to be established. Based on the trained single spectral peak classification models, K-fold cross-validation is performed on the spectral data in the test set to obtain the spectral peak classification results of the original spectrum by each single spectral peak classification model. Each spectral peak category corresponds to a predicted label value, and the number of such values ​​is the number of spectra of each category in the test set.

[0027] For example, assume that there are N s samples, the number of sample categories is N type , then the number of spectra of each type of sample is n / N type At this time, assuming that the correct predicted label value of the first class sample is 1, the wrong predicted label value may be 2 to the value N type , but the total number of predicted label values ​​for each class of samples is fixed: .

[0028] The spectral intensity of the 282 wavelengths was extracted in turn, a single spectrum peak classification model was constructed, and the classification was cross-validated. s and N type Both are 12, the number of spectra in each category is n / N type is 100, and the ratio of the model's training set to the test set data is N train :N test is 4:1, then the total number of predicted label values ​​for each type of sample .

[0029] S104, forming a column vector matrix with the classification scanning results, and calculating the cumulative correlation coefficient between any column vector in the column vector matrix and all other column phasors; The classification scanning results can be classified by a single spectrum peak classification model and obtained through K-fold cross validation to form a predicted label value. The predicted label value of each round of K-fold cross validation forms a one-dimensional column vector. The one-dimensional column vectors of multiple rounds of scanning can form a column vector matrix.

[0030] Specifically, the spectrum peaks of all wavelengths are classified and scanned in turn, and the predicted label values ​​of each round of K-fold cross-validation are vertically spliced ​​to obtain a one-dimensional column vector composed of the predicted label values, and the column vector matrix is ​​formed based on the classification scanning results of all wavelengths; Substitute all one-dimensional column vectors into the Pearson correlation coefficient calculation formula in turn, calculate the Pearson correlation coefficient between any column vector and all other column vectors, and add up the obtained Pearson correlation coefficients to obtain the cumulative correlation coefficient corresponding to each spectrum peak.

[0031] Substitute all column vectors into the Pearson correlation coefficient calculation formula in turn to obtain the Pearson correlation coefficient R between any column vector and all other vectors ij (L i , L j ), and add up the obtained correlation coefficients to get the cumulative correlation coefficient Finally, each wavelength corresponding to each spectrum peak has a cumulative correlation coefficient, and the cumulative correlation coefficient column vector C can be obtained. wⅹ1 For details, please refer to Figure 3 .

[0032] For example, for each round of cross-validation after modeling and classification for each wavelength, a column vector L of predicted label values ​​can be obtained for each type of sample Nsampleⅹ1 = L 20ⅹ1 ; After vertically splicing the predicted label value column vectors of all samples, we can get L Nsample*Ntypeⅹ1 = L 240ⅹ1 ; After 5-fold cross validation, we can get the column vector L of all predicted label values ​​corresponding to the spectrum peak wavelength K*Nsample*Ntypeⅹ1 =L 1200ⅹ1 ; Finally, the correlation coefficient C between any predicted label value column vector and all other column vectors is calculated using the Pearson correlation coefficient formula i , and add up the results to get the cumulative correlation coefficient column vector C 282ⅹ1 ,like Figure 3 As shown in the figure, for spectral peaks of different wavelengths, the Pearson correlation coefficient between any column vector and all other vectors is calculated , and accumulate .

[0033] S105, arranging the cumulative correlation coefficients in ascending order, and selecting the spectrum peaks corresponding to the cumulative correlation coefficients with a preset number of first order as the analysis line clusters; After arranging the cumulative correlation coefficients in ascending order, the smaller the cumulative correlation coefficients are, the smaller the correlation of the spectrum peaks is, and the larger the cumulative correlation coefficients are, the greater the correlation of the spectrum peaks is. Therefore, the spectrum peaks corresponding to the cumulative correlation coefficients ranked in the front can be extracted as analysis line clusters.

[0034] Preferably, a relationship curve between the number of spectrum peak wavelengths and the classification scanning accuracy is constructed, and the number of wavelengths corresponding to when the relationship curve reaches a maximum value or a stable value is obtained, and the number of wavelengths is used as the preset number.

[0035] For example, the accuracy of each scan classification is recorded as CA i , after scanning, the classification accuracy column vector CA can be obtained 282ⅹ1 ; Establish the number of wavelengths and CA i The change curve of Figure 4As shown, the horizontal axis represents the number of wavelengths, and the vertical axis represents the classification accuracy. In this example, when N max = 91 hours CA i The value reaches a maximum value of 94.8%. 282ⅹ1 The wavelengths corresponding to the first 91 cumulative correlation coefficients are selected to form the final analysis line cluster.

[0036] S105, extracting predetermined features of the analysis line cluster from the original spectrum, and constructing a multi-spectral peak classification model through a machine learning algorithm based on the predetermined features of the analysis line cluster; The analysis line cluster refers to a line cluster composed of multiple analysis lines, that is, the spectrum line selected for qualitative and quantitative analysis in atomic emission spectroscopy. By extracting the predetermined features of the analysis line cluster and combining the machine learning algorithm to train the multi-spectral peak classification model, a multi-spectral peak classification model for spectral peak identification is obtained.

[0037] Among them, the machine learning algorithm is a linear discriminant analysis algorithm, a support vector machine algorithm, a neural network algorithm or a K nearest neighbor algorithm.

[0038] S106. Screening the spectral analysis lines to be identified by using a multi-spectral peak classification model.

[0039] The multi-spectral peak classification model is continuously trained by using predetermined features of a preset number of sample analysis line clusters, and the spectral analysis lines to be identified are screened based on the trained multi-spectral peak classification model.

[0040] It can be understood that the laser probe classification method using classification cumulative correlation coefficient scanning line selection (classification correlation coefficient scanning line selection method) is compared with the traditional classification method (wavelet transform line selection direct classification method): using the traditional line selection method, such as direct classification of all wavelengths after wavelet transform peak finding, the classification accuracy of 12 types of igneous rock samples can reach 93.8%, but the number of analysis lines is 282, and the dimensions of training and prediction data are 282ⅹ960 and 282ⅹ240 respectively. Using the method provided in this embodiment, while maintaining the classification accuracy at the original level, the number of analysis lines is only 91, and the dimensions of training and prediction data can also be reduced to 91ⅹ960 and 91ⅹ240 respectively, the data volume is significantly reduced, and the computing efficiency is significantly improved.

[0041] In this embodiment, based on the cumulative correlation coefficient, the corresponding spectral peaks are selected as analysis line clusters, so that the traditional spectral feature vector is further simplified, the dimension of the feature vector is reduced, the amount of data is significantly reduced, and the classification efficiency is improved, and the actual computing efficiency can be significantly improved.

[0042] At the same time, there is no need to perform tedious optimization of many parameters in the wavelet transform process, such as scale parameters, translation parameters, etc. It is only necessary to use the transform to roughly identify the more obvious spectral peaks; the target wavelength can be accurately and quantitatively evaluated, reducing the manual intervention process, thus ensuring the reliability and uniqueness of the line selection process and results; reducing the amount of feature data and reducing the difficulty of modeling.

[0043] It should be understood that the serial numbers of the steps in the above embodiments do not imply a sequence of execution, and the execution sequence of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0044] Figure 5 A schematic diagram of the structure of a spectral analysis line screening system provided by an embodiment of the present invention, the system comprising: A spectrum acquisition module 510 is used to acquire plasma spectra of a preset number of samples based on a laser probe; A spectrum peak identification module 520 is used to use the plasma spectrum as an original spectrum, calculate an average spectrum corresponding to the original spectrum, and perform spectrum peak identification on the average spectrum by wavelet transformation; A spectrum peak classification module 530 is used to extract predetermined features of spectrum peaks from the original spectrum, and classify and scan each spectrum peak based on the predetermined features of the spectrum peaks, wherein the predetermined features are used to characterize the information features of the spectrum peaks; Specifically, the predetermined features of a single wavelength spectrum peak are extracted from the original spectrum to obtain the one-dimensional spectrum predetermined features corresponding to each spectrum peak, and after the one-dimensional spectrum predetermined features are divided into a training set and a test set in proportion, a single spectrum peak classification model is trained and tested, and the single spectrum peak classification model is a linear discriminant analysis model; Based on the trained single-spectral peak classification models, K-fold cross-validation is performed on the spectral data in the test set to obtain the spectral peak classification results of each single-spectral peak classification model for the original spectrum, and each spectral peak category corresponds to a predicted label value.

[0045] Wherein, the predetermined feature is spectral intensity, spectral peak area, spectral peak intensity ratio or spectral peak internal standard intensity value.

[0046] The spectrum peak selection module 540 is used to form a column vector matrix from the classification scanning results, calculate the cumulative correlation coefficient between any column vector in the column vector matrix and all other column vectors, arrange the cumulative correlation coefficients in ascending order, and select the spectrum peaks corresponding to the cumulative correlation coefficients with a preset number of rankings as the analysis line clusters; Specifically, the method of forming a column vector matrix using the classification scanning results and calculating the cumulative correlation coefficient between any column vector and all other column vectors includes: The spectrum peaks of all wavelengths are classified and scanned in turn, and the predicted label values ​​of each round of K-fold cross-validation are vertically spliced ​​to obtain a one-dimensional column vector composed of the predicted label values. The column vector matrix is ​​formed based on the classification scanning results of all wavelengths; Substitute all one-dimensional column vectors into the Pearson correlation coefficient calculation formula in turn, calculate the Pearson correlation coefficient between any column vector and all other column vectors, and add up the obtained Pearson correlation coefficients to obtain the cumulative correlation coefficient corresponding to each spectrum peak.

[0047] Preferably, the step of selecting the spectrum peaks corresponding to the preset number of cumulative correlation coefficients ranked first as the analysis line cluster further comprises: A relationship curve between the number of spectrum peak wavelengths and the classification scanning accuracy is constructed, and the number of wavelengths corresponding to when the relationship curve reaches a maximum value or a stable value is obtained, and the number of wavelengths is used as the preset number.

[0048] A model building module 550 is used to extract predetermined features of the analysis line clusters from the original spectrum, and to build a multi-spectral peak classification model based on the predetermined features of the analysis line clusters through a machine learning algorithm; The analysis line screening module 560 is used to screen the spectral analysis lines to be identified by using a multi-spectral peak classification model.

[0049] Among them, the machine learning algorithm is a linear discriminant analysis algorithm, a support vector machine algorithm, a neural network algorithm or a K-nearest neighbor algorithm.

[0050] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems and modules can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.

[0051] Figure 6 1 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. The electronic device is used for spectral analysis and line screening. Figure 6 As shown, the electronic device 6 of this embodiment includes: a memory 610, a processor 620 and a system bus 630, wherein the memory 610 includes an executable program 6101 stored thereon, and those skilled in the art can understand that Figure 6 The electronic device structure shown in the figure does not constitute a limitation of the electronic device, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0052] Combine the following Figure 6 A detailed introduction to the various components of electronic equipment: The memory 610 can be used to store software programs and modules. The processor 620 executes various functional applications and data processing of the electronic device by running the software programs and modules stored in the memory 610. The memory 610 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device (such as cache data), etc. In addition, the memory 610 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0053] The memory 610 includes an executable program 6101 of the network request method, and the executable program 6101 can be divided into one or more modules / units, which are stored in the memory 610 and executed by the processor 620 to implement spectral analysis line screening, etc. The one or more modules / units can be a series of computer program instruction segments that can complete specific functions, and the instruction segments are used to describe the execution process of the computer program 6101 in the electronic device 6. For example, the computer program 6101 can be divided into functional modules such as a spectrum acquisition module, a spectrum peak recognition module, a spectrum peak classification module, a spectrum peak selection module, a model construction module, and an analysis line screening module.

[0054] The processor 620 is the control center of the electronic device. It uses various interfaces and lines to connect various parts of the entire electronic device. By running or executing software programs and / or modules stored in the memory 610, and calling data stored in the memory 610, it performs various functions of the electronic device and processes data, thereby monitoring the overall status of the electronic device. Optionally, the processor 620 may include one or more processing units; preferably, the processor 620 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, application programs, etc., and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 620.

[0055] The system bus 630 is used to connect the various functional components inside the computer, and can transmit data information, address information, and control information. Its types can be, for example, PCI bus, ISA bus, CAN bus, etc. The instructions of the processor 620 are transmitted to the memory 610 through the bus, and the memory 610 feeds back data to the processor 620. The system bus 630 is responsible for the data and instruction exchange between the processor 620 and the memory 610. Of course, the system bus 630 can also be connected to other devices, such as network interfaces, display devices, etc.

[0056] In the embodiment of the present invention, the executable program executed by the processor 620 included in the electronic device includes: Collect plasma spectra of a preset number of samples based on a laser probe; Taking the plasma spectrum as the original spectrum, calculating the average spectrum corresponding to the original spectrum, and performing spectrum peak identification on the average spectrum by wavelet transformation; Extracting predetermined features of spectral peaks from the original spectrum, and classifying and scanning each spectral peak based on the predetermined features of the spectral peaks, wherein the predetermined features are used to characterize information features of the spectral peaks; The classification scanning results are formed into a column vector matrix, and the cumulative correlation coefficient between any column vector in the column vector matrix and all other column phasors is calculated; Arrange the cumulative correlation coefficients in ascending order, and select the spectrum peaks corresponding to the cumulative correlation coefficients with a preset number of first order as the analysis line clusters; Extracting predetermined features of the analysis line cluster from the original data, and constructing a multi-spectral peak classification model through a machine learning algorithm based on the predetermined features of the analysis line cluster; The spectral analysis lines to be identified are screened through the multi-spectral peak classification model.

[0057] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0058] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0059] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A spectral analysis line screening method, characterized in that: include: Collect plasma spectra of a preset number of samples based on a laser probe; Taking the plasma spectrum as the original spectrum, calculating the average spectrum corresponding to the original spectrum, and performing spectrum peak identification on the average spectrum by wavelet transformation; Extracting predetermined features of spectral peaks from the original spectrum, and classifying and scanning each spectral peak based on the predetermined features of the spectral peaks, wherein the predetermined features are used to characterize information features of the spectral peaks; The classification scanning results are formed into a column vector matrix, and the cumulative correlation coefficient between any column vector in the column vector matrix and all other column phasors is calculated; Arrange the cumulative correlation coefficients in ascending order, and select the spectrum peaks corresponding to the cumulative correlation coefficients with a preset number of first order as the analysis line clusters; Extracting predetermined features of the analysis line clusters from the original spectrum, and constructing a multi-spectral peak classification model through a machine learning algorithm based on the predetermined features of the analysis line clusters; The spectral analysis lines to be identified are screened through the multi-spectral peak classification model.

2. The method according to claim 1, characterized in that The step of extracting predetermined features of spectral peaks from the original spectrum and classifying and scanning each spectral peak based on the predetermined features of the spectral peaks comprises: Extracting predetermined features of a single wavelength spectrum peak from the original spectrum to obtain one-dimensional spectrum predetermined features corresponding to each spectrum peak, dividing the one-dimensional spectrum predetermined features into a training set and a test set in proportion, and training and testing a single spectrum peak classification model, wherein the single spectrum peak classification model is a linear discriminant analysis model; Based on the trained single peak classification models, K-fold cross validation is performed on the spectral data in the test set to obtain the peak classification results of the original spectrum by each single peak classification model, where each peak category corresponds to a predicted label value.

3. The method according to claim 2, characterized in that The predetermined feature is spectral intensity, spectral peak area, spectral peak intensity ratio or spectral peak internal standard intensity value.

4. The method according to claim 2, characterized in that: The classification scanning results are formed into a column vector matrix, and the cumulative correlation coefficient between any column vector in the column vector matrix and all other column vectors is calculated, which includes: The spectrum peaks of all wavelengths are classified and scanned in turn, and the predicted label values ​​of each round of K-fold cross-validation are vertically spliced ​​to obtain a one-dimensional column vector composed of the predicted label values. The column vector matrix is ​​formed based on the classification scanning results of all wavelengths; Substitute all one-dimensional column vectors into the Pearson correlation coefficient calculation formula in turn, calculate the Pearson correlation coefficient between any column vector and all other column vectors, and add up the obtained Pearson correlation coefficients to obtain the cumulative correlation coefficient corresponding to each spectrum peak.

5. The method according to claim 1, characterized in that The step of selecting the spectrum peaks corresponding to the preset number of cumulative correlation coefficients ranked first as the analysis line cluster also includes: A relationship curve between the number of spectrum peak wavelengths and the classification scanning accuracy is constructed, and the number of wavelengths corresponding to when the relationship curve reaches a maximum value or a stable value is obtained, and the number of wavelengths is used as the preset number.

6. The method according to claim 1, characterized in that The machine learning algorithm is a linear discriminant analysis algorithm, a support vector machine algorithm, a neural network algorithm or a K nearest neighbor algorithm.

7. A spectral analysis line screening system, characterized in that: include: A spectrum acquisition module, used for acquiring plasma spectra of a preset number of samples based on a laser probe; A spectrum peak identification module, used to take the plasma spectrum as the original spectrum, calculate the average spectrum corresponding to the original spectrum, and perform spectrum peak identification on the average spectrum by wavelet transformation; A spectrum peak classification module is used to extract predetermined features of spectrum peaks from the original spectrum, and classify and scan each spectrum peak based on the predetermined features of the spectrum peaks, wherein the predetermined features are used to characterize the information features of the spectrum peaks; A spectrum peak selection module is used to form a column vector matrix from the classification scanning results, calculate the cumulative correlation coefficients between any column vector in the column vector matrix and all other column vectors, arrange the cumulative correlation coefficients in ascending order, and select the spectrum peaks corresponding to the cumulative correlation coefficients with a preset number of rankings as analysis line clusters; A model building module is used to extract predetermined features of the analysis line clusters from the original spectrum, and to build a multi-spectral peak classification model based on the predetermined features of the analysis line clusters through a machine learning algorithm; The analysis line screening module is used to screen the spectral analysis lines to be identified through a multi-spectral peak classification model.

8. The system according to claim 7, characterized in that The machine learning algorithm is a linear discriminant analysis algorithm, a support vector machine algorithm, a neural network algorithm or a K nearest neighbor algorithm.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of a spectral analysis line screening method as described in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed, the steps of a spectral analysis line screening method as claimed in any one of claims 1 to 6 are implemented.

Images