A LIBS spectrum correction method and system based on weight fusion prediction

Abstract

The application discloses a LIBS spectrum correction method and system based on weight fusion prediction, relates to the technical field of material analysis, and comprises the following steps: collecting LIBS spectrum data and plasma sound signal data of a sample to be measured; selecting a LIBS spectrum line and a plasma sound signal feature of a metallurgical element to be measured; performing Pearson linear correlation analysis and Spearman rank nonlinear correlation analysis; calculating linear weight components and nonlinear weight components; fusing the linear weight components and the nonlinear weight components; performing weighted feature fusion on the plasma sound signal feature; distributing weights according to a signal-to-noise ratio; fusing the LIBS spectrum of the metallurgical element to be measured and the fused plasma sound signal feature vector; and performing plasma total sound energy correction on the LIBS spectrum of the metallurgical element to be measured in weight fusion prediction. The method can significantly reduce the uncertainty of the original LIBS spectrum, and improve the quantitative analysis capability and reliability of LIBS.

Description

Technical Field

[0001] This application relates to the field of materials analysis technology, and in particular to a method for spectral correction in materials content analysis using LIBS technology. Background Technology

[0002] In recent years, laser-induced breakdown spectroscopy (LIBS) has attracted increasing attention from scholars in the field of materials analysis. Due to its advantages such as speed, non-destructive testing, and the ability to simultaneously detect multiple elements, this technique has been widely applied in materials analysis, environmental monitoring, metallurgical process control, and other fields. However, during the detection process, LIBS spectral signals are often subject to various random influences, including matrix effects of the material itself, laser energy fluctuations, plasma instability, and environmental factors, leading to poor spectral stability and high uncertainty. The uncertainty of LIBS spectra is typically assessed using the relative standard deviation (RSD). This poor spectral stability and high uncertainty not only affect the stable correspondence between spectral line intensity and the content of the analyte but also reduce the accuracy and long-term repeatability of quantitative analysis results.

[0003] In LIBS quantitative analysis, the external standard method is suitable for samples with very similar matrices between standards, simple compositions, and low concentrations of the analyte. However, the external standard method typically has low quantitative accuracy. Therefore, introducing a LIBS spectral line of the matrix element as a correction line can improve quantitative accuracy; this method of introducing an internal standard line is called the internal standard method. However, in recent years, with the innovation of LIBS technology, methods using external reference signals to correct LIBS spectra have gradually emerged, but research on this method is still relatively limited.

[0004] Plasma acoustic signals, as an external reference signal, originate from the rapid attenuation of shock waves during plasma expansion. These signals contain rich physicochemical information about the material itself. Furthermore, plasma acoustic signals exhibit higher stability compared to plasma spectral signals, making their use for LIBS spectrum correction possible. However, existing research on LIBS spectrum correction using plasma acoustic signals primarily focuses on single acoustic signal features (sound pressure peak, acoustic energy, attenuation slope, etc.), neglecting the weighted fusion prediction and correction of LIBS spectra using multiple acoustic signal features. Therefore, proposing a method for LIBS spectrum correction using weighted fusion prediction to reduce LIBS spectral uncertainty while further improving quantitative analysis performance is highly significant. Summary of the Invention

[0005] This application provides a LIBS spectral correction method and system based on weighted fusion prediction to solve the problem that the prior art does not involve weighted fusion prediction of multi-sound signal features to correct LIBS spectra.

[0006] On the one hand, embodiments of this application provide a LIBS spectral correction method based on weighted fusion prediction, including:

[0007] Simultaneously acquire LIBS spectral data and plasma acoustic signal data of the sample to be tested;

[0008] LIBS spectral lines and plasma acoustic signal characteristics of the metallurgical elements to be tested were selected from LIBS spectral data and plasma acoustic signal data, respectively.

[0009] Pearson linear correlation analysis and Spearman rank nonlinear correlation analysis were performed on the LIBS spectral lines and plasma acoustic signal characteristics of the metallurgical elements to be tested, and the Pearson linear correlation coefficient and Spearman rank nonlinear correlation coefficient were obtained respectively.

[0010] The linear weight components and nonlinear weight components are calculated based on the Pearson linear correlation coefficient and the Spearman rank nonlinear correlation coefficient, respectively.

[0011] The linear and nonlinear weight components are fused to obtain the fused weight.

[0012] Multiple plasma acoustic signal features are weighted and fused according to fusion weights to obtain the fused plasma acoustic signal feature vector.

[0013] The LIBS spectral lines and plasma acoustic signal features of the metallurgical elements to be tested are weighted according to the signal-to-noise ratio to obtain the LIBS spectral weights and plasma acoustic signal feature weights of the metallurgical elements to be tested, respectively.

[0014] The LIBS spectrum of the metallurgical element to be tested and the plasma acoustic signal feature vector after fusion are fused according to the LIBS spectral weights and plasma acoustic signal feature weights of the metallurgical element to be tested, so as to obtain the weighted fusion predicted LIBS spectrum of the metallurgical element to be tested.

[0015] The LIBS spectra of the metallurgical elements to be measured, predicted by weighted fusion, are corrected by total plasma acoustic energy to obtain the corrected LIBS spectra.

[0016] On the other hand, embodiments of this application also provide a LIBS spectral correction system based on weighted fusion prediction, including:

[0017] The data acquisition module is used to simultaneously acquire LIBS spectral data and plasma acoustic signal data of the sample under test;

[0018] The feature selection module is used to select the LIBS spectral lines and plasma acoustic signal features of the metallurgical element to be measured from the LIBS spectral data and plasma acoustic signal data, respectively.

[0019] The correlation analysis module is used to perform Pearson linear correlation analysis and Spearman rank nonlinear correlation analysis on the LIBS spectral lines and plasma acoustic signal characteristics of the metallurgical elements to be tested, and to obtain the Pearson linear correlation coefficient and the Spearman rank nonlinear correlation coefficient, respectively.

[0020] The weight component calculation module is used to calculate linear weight components and nonlinear weight components based on Pearson linear correlation coefficient and Spearman rank nonlinear correlation coefficient, respectively.

[0021] The weight fusion module is used to fuse linear weight components and nonlinear weight components to obtain fused weights;

[0022] The feature fusion module is used to perform weighted feature fusion on multiple plasma acoustic signal features according to the fusion weight, so as to obtain the fused plasma acoustic signal feature vector.

[0023] The weight allocation module is used to allocate weights to the LIBS spectral lines and plasma acoustic signal features of the metallurgical elements to be tested according to the signal-to-noise ratio, so as to obtain the LIBS spectral weights and plasma acoustic signal feature weights of the metallurgical elements to be tested respectively.

[0024] The spectral fusion module is used to fuse the LIBS spectrum of the metallurgical element to be tested and the fused plasma acoustic signal feature vector according to the LIBS spectral weights and plasma acoustic signal feature weights of the metallurgical element to be tested, so as to obtain the weighted fusion predicted LIBS spectrum of the metallurgical element to be tested.

[0025] The spectral correction module is used to perform plasma total acoustic energy correction on the LIBS spectra of the metallurgical elements to be measured predicted by weighted fusion, so as to obtain the corrected LIBS spectra.

[0026] On the other hand, embodiments of this application also provide a computer storage medium storing a plurality of computer instructions for causing a computer to execute the above-described method.

[0027] The LIBS spectral correction method and system based on weighted fusion prediction disclosed in this application have the following advantages:

[0028] It can significantly reduce the uncertainty of the original LIBS spectrum, significantly improve the quantitative analysis capability and reliability of LIBS, and at the same time, provide a certain methodological basis for LIBS spectral correction based on external reference signals. Attached Figure Description

[0029] 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 drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a flowchart of the LIBS spectral correction method based on weighted fusion prediction provided in the embodiments of this application.

[0031] Figure 2 This is a LIBS spectral weighting diagram of the metallurgical element Zr and the weighting distribution of various plasma acoustic signal features provided in the embodiments of this application.

[0032] Figure 3 This is a LIBS spectral weighting diagram of the metallurgical element Mo and the weighting distribution of various plasma acoustic signal features provided in the embodiments of this application.

[0033] Figure 4 This is a comparison of RSD before and after LIBS spectral weighted fusion prediction correction for the metallurgical element Zr, provided in an embodiment of this application.

[0034] Figure 5 This is a comparison of RSD before and after LIBS spectral weighted fusion prediction correction for the metallurgical element Mo, provided in the embodiments of this application.

[0035] Figure 6 This is a calibration curve of the metallurgical element Zr before correction, provided in an embodiment of this application.

[0036] Figure 7 This is a calibrated curve of the metallurgical element Zr provided in the embodiments of this application.

[0037] Figure 8 This is a calibration curve of the metallurgical element Mo before correction, provided in the embodiments of this application.

[0038] Figure 9 This is a calibrated curve of the metallurgical element Mo provided in the embodiments of this application. Detailed Implementation

[0039] 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 without creative effort are within the scope of protection of this application.

[0040] Figure 1A flowchart illustrating a LIBS spectral correction method based on weighted fusion prediction provided in this application embodiment. This application embodiment provides a LIBS spectral correction method based on weighted fusion prediction, including:

[0041] S100 simultaneously acquires LIBS spectral data and plasma acoustic signal data of the sample under test.

[0042] For example, under standard laboratory conditions, a laser-induced plasma spectroscopy-acoustic signal synchronous acquisition system is used to acquire multiple sets of raw LIBS spectral data and raw plasma acoustic signal data at multiple different locations on the sample under test. The raw LIBS spectral data contains spectral line data from multiple samples under test. The multiple sets of raw LIBS spectral data and raw plasma acoustic signal data are then averaged to obtain the LIBS spectral data and plasma acoustic signal data, respectively.

[0043] Furthermore, the sample to be tested is pretreated by grinding, polishing, ultrasonic cleaning and alcohol cleaning to ensure that the surface of the sample is relatively clean and flat before testing. Then, raw LIBS spectral data and raw plasma acoustic signal data are collected from the pretreated sample.

[0044] S110, select the LIBS spectral lines and plasma acoustic signal characteristics of the metallurgical element to be tested from the LIBS spectral data and plasma acoustic signal data, respectively.

[0045] For example, the LIBS spectral lines of the metallurgical element to be tested are selected based on pure elemental spectral lines and the NIST (National Institute of Standards and Technology) database. The plasma acoustic signal characteristics include maximum peak value, minimum peak value, peak-to-valley difference, total acoustic energy, root mean square value, mean, skewness, kurtosis, waveform factor, peak factor, impulse factor, margin factor, spectral peak value, center frequency, harmonic components, and spectral entropy.

[0046] S120, Pearson linear correlation analysis and Spearman rank nonlinear correlation analysis were performed on the LIBS spectral lines and plasma acoustic signal characteristics of the metallurgical elements to be tested, and the Pearson linear correlation coefficient and Spearman rank nonlinear correlation coefficient were obtained respectively.

[0047] For example, the Pearson linear correlation coefficient is denoted as Spearman's rank nonlinear correlation coefficient is denoted as ,in Indicates the first k The Pearson linear correlation coefficient between the plasma acoustic signal characteristics and the LIBS spectral lines of the metallurgical element to be measured. Indicates the first kThe Pierman rank nonlinear correlation coefficient between the plasma acoustic signal characteristics and the LIBS spectral lines of the metallurgical element to be measured.

[0048] S130, calculate the linear weight component and the nonlinear weight component based on the Pearson linear correlation coefficient and the Spearman rank nonlinear correlation coefficient, respectively.

[0049] For example, the linear weight component calculation process is as follows:

[0050]

[0051] in, For the first k Linear weighted components of plasma acoustic signal characteristics, This represents the total number of types of plasma acoustic signal characteristics. It is a very small positive value, and its purpose is to prevent division by zero. This indicates taking the absolute value.

[0052] Furthermore, the calculation process for the nonlinear weight components is as follows:

[0053]

[0054] in, For the first k Nonlinear weighted components of plasma acoustic signal characteristics.

[0055] S140 merges the linear and nonlinear weight components to obtain the fused weight.

[0056] For example, the weight fusion process is as follows:

[0057]

[0058] in, To integrate weights, These are the weighting coefficients used in the fusion process.

[0059] S150, weighted feature fusion is performed on multiple plasma acoustic signal features according to the fusion weight to obtain the fused plasma acoustic signal feature vector.

[0060] For example, after obtaining the fusion weights, the fusion weights are normalized, and multiple plasma acoustic signal features are weighted and fused according to the normalized fusion weights.

[0061] Specifically, the normalization process is as follows:

[0062]

[0063] in, These are the normalized fusion weights. They must satisfy:

[0064]

[0065] Furthermore, Z-score normalization is performed on multiple sets of plasma acoustic signal features of the sample to be tested to obtain normalized plasma acoustic signal features. Then, weighted feature fusion is performed on multiple normalized plasma acoustic signal features according to the normalized fusion weights.

[0066] The specific processing procedure is as follows:

[0067]

[0068] in, Characteristics of plasma acoustic signals The mean value of the plasma acoustic signal characteristics. The standard deviation of the plasma acoustic signal characteristics. The characteristics of the standardized plasma acoustic signal.

[0069] Therefore, the eigenvector of the fused plasma acoustic signal is represented as:

[0070]

[0071] in, This represents the feature vector of the fused plasma acoustic signal.

[0072] Furthermore, the fused plasma acoustic signal feature vector is adjusted to the same statistical scale as the LIBS spectral line of the metallurgical element to be measured, and the fused plasma acoustic signal feature is obtained. The fused LIBS spectrum and plasma acoustic signal feature of the metallurgical element to be measured are fused according to the LIBS spectral weight of the metallurgical element to be measured and the feature weight of the plasma acoustic signal.

[0073] Specifically, the eigenvectors of the fused plasma acoustic signal are adjusted to the same statistical scale as the LIBS spectral lines of the metallurgical element to be measured. The specific process is as follows:

[0074]

[0075] in, The standard deviation of the LIBS spectrum of the metallurgical element to be measured is given. The mean value of the LIBS spectrum of the metallurgical element to be measured is given. This refers to the characteristics of plasma acoustic signal fusion.

[0076] S160: The LIBS spectral lines and plasma acoustic signal characteristics of the metallurgical element to be tested are weighted according to the signal-to-noise ratio to obtain the LIBS spectral weights and plasma acoustic signal characteristic weights of the metallurgical element to be tested.

[0077] For example, the unrestricted weighting of the LIBS spectrum of the metallurgical element to be measured. The calculation process is as follows:

[0078]

[0079] in, The signal-to-noise ratio of the LIBS spectrum of the metallurgical element to be measured. The signal-to-noise ratio (SNR) is a characteristic of plasma acoustic signal fusion. It is a very small positive value.

[0080] To ensure the effectiveness of the fusion, the LIBS spectral weights of the metallurgical elements to be measured are determined. Restrictions were imposed:

[0081]

[0082] Furthermore, the feature weights of the plasma acoustic signal are calculated. :

[0083]

[0084] The following constraints must be satisfied:

[0085] S170, based on the LIBS spectral weights of the metallurgical element to be tested and the characteristic weights of the plasma acoustic signal, the LIBS spectrum of the metallurgical element to be tested and the fused plasma acoustic signal characteristic vector are fused to obtain the weighted fusion predicted LIBS spectrum of the metallurgical element to be tested.

[0086] For example, the fusion formula is as follows:

[0087]

[0088] in, Y The LIBS spectrum of the metallurgical element to be measured. The LIBS spectra of the metallurgical elements to be measured are predicted by weighted fusion.

[0089] S180, the total acoustic energy of plasma is corrected for the LIBS spectra of the metallurgical elements to be measured predicted by weighted fusion to obtain the corrected LIBS spectra.

[0090] For example, the correction process is as follows:

[0091]

[0092] in, The LIBS spectrum of the metallurgical element to be measured is corrected. This represents the total energy of the plasma acoustic signal.

[0093] After obtaining the corrected LIBS spectrum, a calibration curve was established by combining the metallurgical element content of the sample with the corrected LIBS spectrum. The calibration curve was then used to perform quantitative analysis of the metallurgical element content of the sample. Specifically, the calibration curve was established with the nominal content of the metallurgical element in the sample as the abscissa and the corresponding corrected LIBS spectral intensity as the ordinate.

[0094] This application also provides a LIBS spectral correction system based on weighted fusion prediction, including:

[0095] The data acquisition module is used to simultaneously acquire LIBS spectral data and plasma acoustic signal data of the sample under test;

[0096] The feature selection module is used to select the LIBS spectral lines and plasma acoustic signal features of the metallurgical element to be measured from the LIBS spectral data and plasma acoustic signal data, respectively.

[0097] The correlation analysis module is used to perform Pearson linear correlation analysis and Spearman rank nonlinear correlation analysis on the LIBS spectral lines and plasma acoustic signal characteristics of the metallurgical elements to be tested, and to obtain the Pearson linear correlation coefficient and the Spearman rank nonlinear correlation coefficient, respectively.

[0098] The weight component calculation module is used to calculate linear weight components and nonlinear weight components based on Pearson linear correlation coefficient and Spearman rank nonlinear correlation coefficient, respectively.

[0099] The weight fusion module is used to fuse linear weight components and nonlinear weight components to obtain fused weights;

[0100] The feature fusion module is used to perform weighted feature fusion on multiple plasma acoustic signal features according to the fusion weight, so as to obtain the fused plasma acoustic signal feature vector.

[0101] The weight allocation module is used to allocate weights to the LIBS spectral lines and plasma acoustic signal features of the metallurgical elements to be tested according to the signal-to-noise ratio, so as to obtain the LIBS spectral weights and plasma acoustic signal feature weights of the metallurgical elements to be tested respectively.

[0102] The spectral fusion module is used to fuse the LIBS spectrum of the metallurgical element to be tested and the fused plasma acoustic signal feature vector according to the LIBS spectral weights and plasma acoustic signal feature weights of the metallurgical element to be tested, so as to obtain the weighted fusion predicted LIBS spectrum of the metallurgical element to be tested.

[0103] The spectral correction module is used to perform plasma total acoustic energy correction on the LIBS spectra of the metallurgical elements to be measured predicted by weighted fusion, so as to obtain the corrected LIBS spectra.

[0104] This application also provides a computer storage medium storing a plurality of computer instructions for causing a computer to execute the above-described method.

[0105] Example

[0106] A set of standard titanium alloy samples (S1~S5) were used as the implementation objects. The method was verified by selecting metallurgical components Zr and Mo respectively. Two spectral lines were selected for analysis: Zr II at 343.82 nm and Mo I at 379.83 nm. The specific steps are as follows:

[0107] 1. Obtain a set of titanium alloy standard samples as test samples. The test samples are cylinders with a diameter of 35 mm and a height of 40 mm.

[0108] 2. The samples to be tested are pretreated in sequence. The pretreatment process includes grinding, polishing, ultrasonic cleaning and alcohol cleaning to ensure that the surface of the sample is relatively clean and flat before testing.

[0109] 3. Under standard laboratory conditions, a laser-induced plasma spectroscopy-acoustic signal synchronous acquisition system was used to detect the test samples. Multiple sets of raw LIBS spectral data and raw plasma acoustic signal data were acquired from various locations on each test sample. The raw LIBS spectral data and raw plasma acoustic signal data were acquired and recorded synchronously. Specifically, 20 different locations were detected for each test sample, and 10 sets of raw LIBS spectral data and raw plasma acoustic signal data were acquired from the same location on each test sample, resulting in a total of 200 sets of raw LIBS spectral data and raw plasma acoustic signal data for each test sample. Each set of LIBS spectral data and plasma acoustic signal data was obtained by averaging multiple sets of raw LIBS spectral data and raw plasma acoustic signal data.

[0110] 4. Select LIBS spectral lines and plasma acoustic signal characteristics of the metallurgical elements to be tested from the LIBS spectral data and plasma acoustic signal data respectively. The LIBS spectral lines of the metallurgical elements to be tested are selected by combining pure elemental spectral lines and the NIST database; the plasma acoustic signal characteristics include maximum peak value, minimum peak value, peak-to-valley difference, total acoustic energy, root mean square value, mean, skewness, kurtosis, waveform factor, peak factor, impulse factor, margin factor, spectral peak value, center frequency, harmonic components, and spectral entropy.

[0111] 5. Pearson linear correlation analysis and Spearman rank nonlinear correlation analysis were performed on the LIBS spectral lines and plasma acoustic signal characteristics of the metallurgical elements to be tested, respectively.

[0112] 6. Calculate the linear and nonlinear weight components separately, then perform weight fusion and weight normalization. The results are as follows: Figure 2 and 3 As shown.

[0113] 7. The plasma acoustic signal features obtained from different locations of the same sample are subjected to Z-score normalization, and then weighted feature fusion is performed.

[0114] 8. Adjust the normalized and fused plasma acoustic signal characteristics to the same statistical scale as the LIBS spectral lines of the metallurgical element to be measured.

[0115] 9. Perform weight allocation and impose weight restrictions on the LIBS spectra of the metallurgical elements to be measured.

[0116] 10. Perform weighted calculation of plasma acoustic signal fusion features and weighted fusion prediction of LIBS spectra of the metallurgical elements to be measured.

[0117] 11. Perform total acoustic energy correction on the LIBS spectra of the metallurgical elements to be measured predicted by weighted fusion, and finally establish calibration curves for quantitative analysis.

[0118] 12. Sample S3 was used as the validation set. The specific quantitative results are shown in Table 1, or refer to [the table below]. Figure 4 and Figure 5 .

[0119] Table 1 Comparison of quantitative results before and after correction

[0120]

[0121] Through observation Figure 6-9 As can be seen, the coefficient of determination R² of the calibration curve was significantly improved, indicating that the performance of the calibration curve after correction is better. Meanwhile, as shown in Table 1, when using sample S3 for verification, the metallurgical components Zr and Mo contents were 0.972 wt.% and 1.645 wt.% before correction, and 1.17 wt.% and 2.647 wt.% after correction. The predicted results are getting closer and closer to the nominal contents, indicating a significant improvement in quantitative accuracy.

[0122] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0123] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A LIBS spectral correction method based on weighted fusion prediction, characterized in that, include: Simultaneously acquire LIBS spectral data and plasma acoustic signal data of the sample to be tested; LIBS spectral lines and plasma acoustic signal characteristics of the metallurgical element to be tested are selected from the LIBS spectral data and the plasma acoustic signal data, respectively. Pearson linear correlation analysis and Spearman rank nonlinear correlation analysis were performed on the LIBS spectral lines of the metallurgical element to be tested and the plasma acoustic signal characteristics to obtain the Pearson linear correlation coefficient and the Spearman rank nonlinear correlation coefficient, respectively. The linear weight component and the nonlinear weight component are calculated based on the Pearson linear correlation coefficient and the Spearman rank nonlinear correlation coefficient, respectively. The linear weight component and the nonlinear weight component are fused to obtain the fused weight; The multiple plasma acoustic signal features are weighted and fused according to the fusion weight to obtain the fused plasma acoustic signal feature vector; The LIBS spectral lines of the metallurgical element to be tested and the plasma acoustic signal features are weighted according to the signal-to-noise ratio to obtain the LIBS spectral weights and plasma acoustic signal feature weights of the metallurgical element to be tested, respectively. The LIBS spectrum of the metallurgical element to be tested and the plasma acoustic signal feature weights are fused according to the LIBS spectral weights of the metallurgical element to be tested and the fused plasma acoustic signal feature vector to obtain the weighted fusion predicted LIBS spectrum of the metallurgical element to be tested. The LIBS spectra of the metallurgical elements to be measured, predicted by weighted fusion, are corrected by total plasma acoustic energy to obtain the corrected LIBS spectra.

2. The LIBS spectral correction method based on weighted fusion prediction according to claim 1, characterized in that, Multiple sets of raw LIBS spectral data and raw plasma acoustic signal data are acquired at multiple different locations on the sample under test using a laser-induced plasma spectral-acoustic signal synchronous acquisition system. The raw LIBS spectral data and raw plasma acoustic signal data are averaged to obtain the LIBS spectral data and the plasma acoustic signal data, respectively.

3. The LIBS spectral correction method based on weighted fusion prediction according to claim 2, characterized in that, The sample to be tested is pretreated by grinding, polishing, ultrasonic cleaning and alcohol cleaning, and then the raw LIBS spectral data and the raw plasma acoustic signal data are collected from the pretreated sample.

4. The LIBS spectral correction method based on weighted fusion prediction according to claim 1, characterized in that, The LIBS spectral lines of the metallurgical elements to be tested are selected based on pure elemental spectral lines and the NIST database. The plasma acoustic signal characteristics include maximum peak value, minimum peak value, peak-to-valley difference, total acoustic energy, root mean square value, mean, skewness, kurtosis, waveform factor, peak factor, impulse factor, margin factor, spectral peak value, center frequency, harmonic components, and spectral entropy.

5. The LIBS spectral correction method based on weighted fusion prediction according to claim 1, characterized in that, After obtaining the fusion weights, the fusion weights are normalized, and multiple plasma acoustic signal features are weighted and fused according to the normalized fusion weights.

6. The LIBS spectral correction method based on weighted fusion prediction according to claim 5, characterized in that, The plasma acoustic signal features of the sample to be tested are Z-score normalized to obtain normalized plasma acoustic signal features. The normalized plasma acoustic signal features are then weighted and fused according to the normalized fusion weights.

7. The LIBS spectral correction method based on weighted fusion prediction according to claim 1, characterized in that, The fused plasma acoustic signal feature vector is adjusted to the same statistical scale as the LIBS spectral line of the metallurgical element to be tested, and the plasma acoustic signal fusion feature is obtained. The LIBS spectrum of the metallurgical element to be tested and the plasma acoustic signal feature fusion feature are then fused according to the LIBS spectral weight of the metallurgical element to be tested and the plasma acoustic signal feature weight.

8. The LIBS spectral correction method based on weighted fusion prediction according to claim 1, characterized in that, A calibration curve was established by combining the metallurgical element content of the sample to be tested with the corrected LIBS spectrum, and the metallurgical element content of the sample to be tested was quantitatively analyzed using the calibration curve.

9. A LIBS spectral correction system based on weighted fusion prediction, characterized in that, The system employs the LIBS spectral correction method based on weighted fusion prediction as described in any one of claims 1-8, and the system comprises: The data acquisition module is used to simultaneously acquire LIBS spectral data and plasma acoustic signal data of the sample under test; The feature selection module is used to select LIBS spectral lines and plasma acoustic signal features of the metallurgical element to be tested from the LIBS spectral data and the plasma acoustic signal data, respectively. The correlation analysis module is used to perform Pearson linear correlation analysis and Spearman rank nonlinear correlation analysis on the LIBS spectral lines of the metallurgical element to be tested and the plasma acoustic signal characteristics, and to obtain the Pearson linear correlation coefficient and the Spearman rank nonlinear correlation coefficient, respectively. The weight component calculation module is used to calculate the linear weight component and the nonlinear weight component based on the Pearson linear correlation coefficient and the Spearman rank nonlinear correlation coefficient, respectively. The weight fusion module is used to fuse the linear weight components and the nonlinear weight components to obtain the fused weights; The feature fusion module is used to perform weighted feature fusion on multiple plasma acoustic signal features according to the fusion weight, so as to obtain a fused plasma acoustic signal feature vector. The weight allocation module is used to allocate weights to the LIBS spectral lines of the metallurgical element to be tested and the plasma acoustic signal features according to the signal-to-noise ratio, so as to obtain the LIBS spectral weights and plasma acoustic signal feature weights of the metallurgical element to be tested, respectively. The spectral fusion module is used to fuse the LIBS spectrum of the metallurgical element to be tested and the fused plasma acoustic signal feature vector according to the LIBS spectral weight of the metallurgical element to be tested and the plasma acoustic signal feature weight, so as to obtain the weighted fusion predicted LIBS spectrum of the metallurgical element to be tested. The spectral correction module is used to perform plasma total acoustic energy correction on the LIBS spectra of the metallurgical elements to be measured predicted by weighted fusion, so as to obtain the corrected LIBS spectra.

10. A computer storage medium, characterized in that, The computer storage medium stores a plurality of computer instructions, which are used to cause the computer to perform the method described in any one of claims 1-8.

Images