Method for determining hydrocarbon components in mixed aromatic hydrocarbon by using comprehensive two-dimensional gas chromatography-mass spectrometry

Abstract

The invention provides a method for determining hydrocarbon components in mixed aromatic hydrocarbons by using comprehensive two-dimensional gas chromatography-mass spectrometry. The method comprises the following steps: optimizing a target function to make a chromatographic peak signal obtained by a separation model accord with a chromatographic separation process mechanism, fully utilizing complementary information of chromatography and mass spectrometry, improving separation precision, and guiding analysis of complex aliasing peaks; analyzing overlapped chromatographic peaks determined by principal component analysis by adopting the established separation model to obtain a chromatographic peak signal representing a single component after separation, and matching the chromatographic peak signal with a corresponding mass spectrum; in combination with a single chromatographic peak signal and a mass spectrum obtained through separation, according to characteristic fragments and chromatographic retention time in the mass spectrum, component types corresponding to chromatographic peaks are judged, and characterization of the composition of the mixed aromatic hydrocarbon is achieved; a separated single chromatographic peak is subjected to quantitative analysis, the content of each component is calculated by adopting an external standard method in contrast with a standard curve, a quantitative result of the mixed aromatic hydrocarbon sample is obtained, and the measurement precision of the mass fraction of each hydrocarbon component is improved.

Description

Technical Field

[0001] The invention relates to the field of detection technology, and in particular to a method for determining hydrocarbon components in mixed aromatics by utilizing comprehensive two-dimensional gas chromatography-mass spectrometry. Background Art

[0002] When using comprehensive two-dimensional gas chromatography-mass spectrometry to analyze mixed aromatic samples from Fischer-Tropsch synthesis, qualitative and quantitative analysis is difficult due to the complex sample matrix, numerous components, and serious retention time overlap. Especially when the sample contains homologues with similar structures, their retention behaviors on non-polar and polar chromatographic columns are similar, and the chromatographic peaks on the two-dimensional chromatogram are easy to overlap, making it impossible to achieve reliable qualitative and accurate quantification. In addition, when the retention time of certain trace components partially overlaps with that of a large number of components, the small peak is likely to be masked by the large peak and cannot be detected, affecting the sensitivity and detection limit of the analysis. Therefore, how to optimize the chromatographic separation conditions, maximize the peak capacity of the two-dimensional chromatogram, and reduce peak overlap is a key technical problem that needs to be solved when analyzing complex mixture samples by comprehensive two-dimensional gas chromatography. Summary of the invention

[0003] The invention provides a method for determining hydrocarbon components in mixed aromatics by using comprehensive two-dimensional gas chromatography-mass spectrometry, aiming at improving the determination accuracy of the mass fraction of each hydrocarbon component.

[0004] The present invention provides a method for determining hydrocarbon components in mixed aromatics by comprehensive two-dimensional gas chromatography-mass spectrometry, comprising the following steps:

[0005] Obtain comprehensive two-dimensional gas chromatography-mass spectrometry data of Fischer-Tropsch synthesis mixed aromatics samples, and establish a two-dimensional retention time matrix of sample components, where each matrix element is the retention time value of a component on a non-polar and polar chromatographic column;

[0006] A hierarchical clustering algorithm was used to cluster the retention time matrix, and homologues with similar structures were grouped into one category according to the Euclidean distance, their distribution areas on the two-dimensional chromatogram were determined, and the areas where the chromatographic peaks overlapped were found;

[0007] For the overlapping regions of chromatographic peaks determined by cluster analysis, principal component analysis was used to extract characteristic information of overlapping peaks from aspects such as peak shape, peak width, retention time, and mass spectrum fragmentation, and a multidimensional characteristic matrix of overlapping chromatographic peaks was established.

[0008] Based on the feature matrix extracted by principal component analysis, an independent component analysis method was used to establish a separation model for overlapping chromatographic peaks by minimizing statistical independence, and an objective function was constructed to measure the physicochemical significance of the separated chromatographic peak signals;

[0009] By optimizing the objective function, the chromatographic peak signals obtained by the separation model conform to the mechanism of the chromatographic separation process, making full use of the complementary information of chromatography and mass spectrometry to improve the separation accuracy and guide the resolution of complex overlapping peaks.

[0010] The established separation model is used to analyze the overlapping chromatographic peaks determined by principal component analysis, and the chromatographic peak signals representing single components after separation are obtained and matched with the corresponding mass spectra.

[0011] Combining the separated single chromatographic peak signals and mass spectra, based on the characteristic fragments and chromatographic retention times in the mass spectra, the types of components corresponding to the chromatographic peaks are judged to realize the characterization of the composition of mixed aromatics.

[0012] Quantitative analysis is carried out on the separated single chromatographic peaks. The external standard method is used to calculate the content of each component by referring to the standard curve, and the quantitative results of the mixed aromatic hydrocarbon sample are obtained, providing composition information feedback for process optimization.

[0013] The technical solution provided by the embodiments of the present invention may include the following beneficial effects:

[0014] The present invention discloses a method for analyzing the components of Fischer-Tropsch synthesis mixed aromatics based on comprehensive two-dimensional gas chromatography-mass spectrometry. This method first establishes a two-dimensional retention time matrix of sample components, and determines the chromatographic peak overlapping region through hierarchical clustering algorithm. For the overlapping region, principal component analysis is used to extract characteristic information and establish a multi-dimensional characteristic matrix. Then, the independent component analysis method is used to construct an overlapping chromatographic peak separation model, and the resolution of complex overlapping peaks is realized by optimizing the objective function. Finally, combining the separated single chromatographic peak signals and mass spectra, the component types are judged and quantitative analysis is carried out. The present invention makes full use of the complementary information of chromatography and mass spectrometry, improves the separation accuracy and characterization ability of overlapping components in complex mixtures, provides accurate composition information feedback for Fischer-Tropsch synthesis process optimization, and improves the measurement accuracy of the mass fraction of each hydrocarbon component. Brief Description of the Drawings

[0015] Figure 1 It is a flow chart of a method for determining hydrocarbon components in mixed aromatics by comprehensive two-dimensional gas chromatography-mass spectrometry of the present invention. Detailed Embodiments

[0016] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be described in detail below with reference to the drawings and specific embodiments.

[0017] As Figure 1 shown, a method for determining hydrocarbon components in mixed aromatics by comprehensive two-dimensional gas chromatography-mass spectrometry in this embodiment may specifically include:

[0018] Step S101: Obtain the comprehensive two-dimensional gas chromatography-mass spectrometry (GC×GC-MS) data of the Fischer-Tropsch synthesis mixed aromatic hydrocarbon sample, and establish a two-dimensional retention time matrix of the sample components, where each matrix element is the retention time value of a component on non-polar and polar chromatographic columns.

[0019] Obtain a Fischer-Tropsch synthesis mixed aromatic hydrocarbon sample, analyze the sample using comprehensive two-dimensional gas chromatography-mass spectrometry technology to obtain the two-dimensional chromatogram data of the sample. According to the comprehensive two-dimensional gas chromatography-mass spectrometry analysis results, determine the retention time values of each component in the mixed aromatic hydrocarbon sample on non-polar and polar chromatographic columns. Analyze the retention time values of each component through a clustering algorithm, and classify the components into different categories based on the similarity of retention times. For each category of components, establish a two-dimensional matrix of the retention times of this category of components on non-polar and polar chromatographic columns respectively. Analyze each two-dimensional retention time matrix, and use the support vector machine algorithm to determine whether there are abnormal data points in the matrix. If there are abnormal points, remove them. Integrate the two-dimensional retention time matrices of all categories of components to construct an overall two-dimensional retention time matrix model of the mixed aromatic hydrocarbon sample. Use the principal component analysis method to perform dimensionality reduction on the overall two-dimensional retention time matrix to obtain a low-dimensional matrix that can reflect the characteristics of the components of the mixed aromatic hydrocarbon sample, laying a foundation for subsequent qualitative and quantitative analysis.

[0020] Exemplarily, obtain a Fischer-Tropsch synthesis mixed aromatic hydrocarbon sample, analyze the sample using an Agilent 7890B comprehensive two-dimensional gas chromatograph and an Agilent 7000D triple quadrupole mass spectrometer to obtain the two-dimensional chromatogram data of the sample. According to the comprehensive two-dimensional gas chromatography-mass spectrometry analysis results, determine the retention time values of each component in the mixed aromatic hydrocarbon sample on a DB-5MS non-polar chromatographic column and a DB-17HT polar chromatographic column. Analyze the retention time values of each component through the K-means clustering algorithm, and classify the components into 8 different categories based on the Euclidean distance similarity measure. For each category of components, establish a two-dimensional matrix of the retention times of this category of components on non-polar and polar chromatographic columns respectively. Analyze each two-dimensional retention time matrix, and use the one-class SVM support vector machine algorithm to determine whether there are abnormal data points in the matrix. If there are abnormal points with a Mahalanobis distance greater than 5, remove them. Integrate the two-dimensional retention time matrices of all categories of components to construct a 100×100 overall two-dimensional retention time matrix model of the mixed aromatic hydrocarbon sample. Use the principal component analysis (PCA) method to perform dimensionality reduction on the overall two-dimensional retention time matrix. By solving the eigenvalues and eigenvectors of the covariance matrix, select the first 3 principal components with a contribution rate greater than 85% to obtain a 3×100 low-dimensional matrix that can reflect the characteristics of the components of the mixed aromatic hydrocarbon sample, realizing data compression and noise reduction, and laying a foundation for subsequent qualitative and quantitative analysis.

[0021] Step S102: Use the hierarchical clustering algorithm to perform clustering analysis on the retention time matrix. Classify homologues with similar structures into one category according to the Euclidean distance, determine their distribution regions on the two-dimensional chromatogram, and find the regions where chromatographic peaks overlap.

[0022] Obtain the retention time matrix data, preprocess the data in the matrix, remove noise and outliers to ensure the accuracy and reliability of the data. According to the preprocessed retention time matrix data, use the hierarchical clustering algorithm for clustering analysis. By calculating the Euclidean distance between data points, classify homologues with similar structures into one category. According to the results of the clustering analysis, determine the distribution regions of each category of homologues on the two-dimensional chromatogram, and obtain the distribution positions and ranges of different categories of homologues. Analyze the distribution of chromatographic peaks on the two-dimensional chromatogram, find the regions where chromatographic peaks overlap, and judge whether there are cases where chromatographic peaks of multiple homologues appear at the same position. If there are regions where chromatographic peaks overlap, further distinguish and identify different homologues by analyzing characteristics such as the chromatographic peak shape and retention time in the overlapping region. According to the analysis results of the chromatographic peak overlapping region, optimize and adjust the two-dimensional chromatogram, reduce the chromatographic peak overlapping situation by changing chromatographic conditions or selecting a more suitable chromatographic column. Compare the optimized two-dimensional chromatogram with the original chromatogram, evaluate the optimization effect, determine the final results of homologue separation and identification, and obtain accurate and reliable qualitative and quantitative analysis results.

[0023] Exemplarily, assume that we obtain a two-dimensional gas chromatography retention time matrix data containing 100 compounds, where the retention time range of the first dimension is 1 - 50 minutes and the retention time range of the second dimension is 1 - 10 seconds. First, preprocess the original data. Use the box plot method to identify outliers. For example, if the retention time of a certain compound in the first dimension is 60 minutes, which exceeds the maximum value of 50 minutes, it is regarded as an outlier and replaced with the median. Then, calculate the signal-to-noise ratio of each data point, and chromatographic peaks with a signal-to-noise ratio lower than 3 are regarded as noise and removed. Next, perform standardization processing on the data, converting the retention time data of the two dimensions of each compound into a standard normal distribution with a mean of 0 and a standard deviation of 1. Apply the hierarchical clustering algorithm to the preprocessed data. First, regard each compound as a separate class, and then calculate the Euclidean distance between every two compounds. For example, the Euclidean distance between compound A and compound B is sqrt((A's retention time in the first dimension - B's retention time in the first dimension)^2+(A's retention time in the second dimension - B's retention time in the second dimension)^2)=sqrt((10 - 12)^2+(3 - 4)^2). According to the calculated distance matrix, merge the two closest classes into a new class, and repeat this process until all compounds are merged into one class. The entire process can be plotted as a dendrogram. Cut the dendrogram into different clusters according to a preset threshold (e.g., 5). For example, divide 100 compounds into 5 clusters. According to the clustering results, mark compounds of different classes with different colors on the two-dimensional chromatogram. For example, mark the first class of compounds as red, the second class of compounds as blue, etc., and the distribution areas of different classes of compounds on the two-dimensional chromatogram can be visually seen. For example, the first class of compounds is mainly distributed in the area with a retention time of 10 - 15 minutes in the first dimension and a retention time of 2 - 4 seconds in the second dimension. By observing the two-dimensional chromatogram, it is found that there is chromatographic peak overlap in the area with a retention time of 25 minutes in the first dimension and a retention time of 6 seconds in the second dimension. There are two chromatographic peaks of compounds that almost completely overlap in this area. By analyzing the shapes of the chromatographic peaks in the overlapping area, it is found that the peak height ratio of these two chromatographic peaks is approximately 2:1, and there are slight differences in peak width. Combining with the mass spectrometry data, it can be inferred that the compound with a higher peak height may be a normal alkane, and the compound with a lower peak height may be an isoparaffin. To improve the situation of chromatographic peak overlap, it can be tried to reduce the column temperature gradient, changing the original temperature increase of 5℃ per minute to 3℃ per minute, or replace it with a more polar chromatographic column. For example, replace the DB-5 column with a DB-17 column. After optimization, run the sample again to obtain a new two-dimensional chromatogram. Compare the new two-dimensional chromatogram with the original chromatogram, and it is found that the two originally overlapping chromatographic peaks have been basically separated, the peak height ratio is approximately 5:1, and the difference in peak width is more obvious. Therefore, it can be considered that the optimized chromatographic conditions can better separate the target compounds and obtain more accurate qualitative and quantitative analysis results. Verify the stability and reproducibility of the optimized chromatographic conditions through repeated experiments to ensure the reliability of the results.For a more comprehensive evaluation, the optimization effect can be quantified by calculating the resolution of chromatographic peaks. For example, the resolution before optimization was 8, and after optimization it was 5, indicating a significant improvement in resolution and proving the effectiveness of the optimization scheme. Finally, based on the optimized two-dimensional chromatogram, accurate qualitative and quantitative analysis of homologues in the sample can be performed. For example, it can be determined that the contents of n-alkanes and isoalkanes in the sample are 10% and 5% respectively.

[0024] Step S103: For the chromatographic peak overlapping regions determined by cluster analysis, use the principal component analysis method to extract the characteristic information of the overlapping peaks from aspects such as peak shape, peak width, retention time, and mass spectrometry fragments, and establish a multi-dimensional characteristic matrix of the overlapping chromatographic peaks.

[0025] Using the cluster analysis method, cluster the chromatographic peaks in the chromatogram, identify the chromatographic peak regions with overlap, and determine the target regions that need to be separated and resolved. For the overlapping chromatographic peak regions determined by cluster analysis, extract one-dimensional characteristics such as peak shape, peak width, and retention time, as well as two-dimensional characteristics such as mass spectrometry fragment information, to construct a multi-dimensional characteristic matrix of the overlapping chromatographic peaks. According to the constructed multi-dimensional characteristic matrix, use the principal component analysis method to map the high-dimensional characteristic data to a low-dimensional space, extract the main characteristic information of the overlapping chromatographic peaks, and reduce the characteristic dimension. Through the principal component score matrix obtained by principal component analysis, represent the characteristics of the overlapping chromatographic peaks, highlight the differences between different overlapping peaks, and provide a basis for subsequent peak separation. According to the results of principal component analysis, use the multi-way curve resolution algorithm to resolve the overlapping chromatographic peaks, estimate parameters such as the peak shape and peak area of each overlapping peak, and achieve quantitative analysis of the overlapping peaks. Combining the mass spectrometry fragment information, identify the components of each separated chromatographic peak, infer the substance components contained in the overlapping peaks, and complete the qualitative analysis. Integrate the results of qualitative and quantitative analysis to obtain the complete analysis results of the overlapping chromatographic peaks, including information such as the retention time, peak area, and substance components of each overlapping peak, providing data support for subsequent sample analysis.

[0026] Exemplarily, assume there is a set of gas chromatography-mass spectrometry (GC-MS) data. First, data preprocessing is performed, including baseline correction, noise reduction, etc. Peak detection is carried out on the preprocessed chromatogram, and 100 chromatographic peaks are detected. The K-means clustering algorithm is used, with the number of clusters set to 5. Clustering is performed based on the retention time of the chromatographic peaks (e.g., 12 minutes, 15 minutes, 18 minutes, etc.) and the peak width (e.g., 1 minute, 2 minutes, 3 minutes, etc.). It is found that there is an obvious overlapping phenomenon among the chromatographic peaks in the third cluster (including 15 peaks), and this region is determined as the target region. For these 15 overlapping peaks, the peak height of each peak (e.g., 1000, 1500, 2000, etc.), the half-peak width (e.g., 15 minutes, 18 minutes, 22 minutes, etc.), and the abundances of 10 main fragments in the mass spectrum corresponding to each peak (e.g., the abundance of ion m / z43 is 5, the abundance of ion m / z57 is 8, etc.) are extracted to construct a multi-dimensional feature matrix with 15 rows (corresponding to 15 peaks) and 13 columns (corresponding to 3 one-dimensional features and 10 two-dimensional features). After standardizing this matrix, principal component analysis (PCA) is applied. The contribution rates of the first two principal components are calculated to be 60% and 25% respectively, and the total contribution rate reaches 85%. Therefore, the first two principal components are selected. The original data is projected onto the low-dimensional space composed of these two principal components to obtain the principal component scores of each overlapping peak. For example, the score of the first overlapping peak is (3, -5), and the score of the second overlapping peak is (-8, 9), etc. According to the principal component scores, different sub-peaks in the overlapping peak group can be clearly distinguished. For example, peaks with close scores may belong to the same substance. Using multivariate curve resolution-alternating least squares (MCR-ALS), with the number of components set to 3, the Gaussian function is used as the initial peak shape, and the principal component scores obtained by principal component analysis are used as the initial concentration matrix. The peak shape and the concentration matrix are iteratively optimized until convergence. Finally, the analytical results of 3 components are obtained. The peak center of the first component is at 15 minutes, and the peak area is 3000. The peak center of the second component is at 18 minutes, and the peak area is 2500. The peak center of the third component is at 11 minutes, and the peak area is 2000. Combining the mass spectra of each component and comparing with the standard spectral library, the first component is identified as n-hexane (matching degree 95%), the second component is identified as toluene (matching degree 92%), and the third component is identified as xylene (matching degree 90%). Finally, the complete analytical results of this overlapping region are obtained: it contains three components, n-hexane, toluene, and xylene, with retention times of 15 minutes, 18 minutes, and 11 minutes respectively, and peak areas of 3000, 2500, and 2000 respectively.

[0027] Step S104, according to the feature matrix extracted by principal component analysis, an independent component analysis method is used to establish a separation model for overlapping chromatographic peaks by minimizing statistical independence, and an objective function for measuring the physical and chemical significance of the separated chromatographic peak signals is constructed.

[0028] Perform data dimensionality reduction processing on the initial feature matrix obtained according to principal component analysis to obtain a first feature matrix with low dimensions. Process the first feature matrix using the independent component analysis method to obtain multiple mutually independent second independent components, and determine the combination with the minimum statistical independence. Construct a separation model for overlapping chromatographic peaks based on the second independent components to obtain the first signals of each separated single chromatographic peak. Construct an objective function through the first signals to obtain the peak area corresponding to each single chromatographic peak, and judge the magnitude relationship between the peak area and a preset threshold. If the peak area is greater than the preset threshold, calculate the precision of the second signal of each chromatographic peak based on the peak area. Obtain the recovery rate of the second signal of each chromatographic peak to obtain the relationship between the recovery rate and a preset range. Based on the peak area, precision, and recovery rate, use the weighted average method to obtain the comprehensive evaluation index of each chromatographic peak, and judge the relationship between the comprehensive evaluation index and the set index.

[0029] Exemplarily, assume that the initial feature matrix obtained through principal component analysis is a matrix with 100 rows and 50 columns, representing 50 original variables of 100 samples. When performing data dimensionality reduction, the cumulative contribution rate threshold of the principal components is set to 95%. By calculating the eigenvalues and eigenvectors, it is determined that the cumulative contribution rate of the first 10 principal components reaches 96%, meeting the threshold requirement. Therefore, these 10 principal components are selected to form the first feature matrix, whose dimension is 100 rows and 10 columns, achieving data dimensionality reduction. The first feature matrix is processed using the fast independent component analysis algorithm, and the number of independent components is set to 8. The random weight matrix is initialized, and through iterative optimization to maximize non-Gaussianity. After 100 iterations, the algorithm converges, and 8 mutually independent second independent components are obtained. Calculate the mutual information between every two independent components, and it is found that the mutual information between the 3rd and 5th independent components is the smallest, which is 0.5, and it is determined as the combination with the smallest statistical independence. Based on these 8 independent components, a separation model for overlapping chromatographic peaks is constructed. Assume that each independent component corresponds to a single chromatographic peak. Through linear combination, the 8 independent components are multiplied by the corresponding mixing coefficients to reconstruct the original overlapping chromatographic signal, obtaining the first signal of 8 separated single chromatographic peaks, and the length of each first signal is 1000 data points. To improve the accuracy of the separation model, prior information is introduced. For example, it is known that the peak shape of each chromatographic peak is a Gaussian function, and the peak width is between 5 and 2 minutes. A target function is constructed through the first signal. The target function is the sum of the squares of the differences between the sum of the peak areas of all single chromatographic peaks and the total peak area obtained from actual measurement, plus the sum of the squares of the differences between the peak shape of each single chromatographic peak and the Gaussian function, and the sum of the squares of the differences between the peak width and the prior range. The gradient descent method is used for optimization, and the learning rate is set to 0.1. After 500 iterations, the target function converges, and the peak areas corresponding to each single chromatographic peak are obtained, which are 2, 5, 8, 1, 5, 9, 0, 8 respectively. Judge the size relationship between the peak area and the preset threshold of 0, and it is found that the peak areas of the 3rd and 6th chromatographic peaks are less than the threshold, so the subsequent precision and recovery rate calculations are not performed. For the 6 chromatographic peaks with peak areas greater than 0, calculate the precision of the second signal of each chromatographic peak. Assume that the second signal of each chromatographic peak is obtained through 5 repeated measurements. Calculate the standard deviation of the peak area of each measurement, and then divide it by the average peak area to obtain the relative standard deviation, that is, the precision, which are 1%, 5%, 8%, 9%, 3%, 5% respectively. Obtain the recovery rate of the second signal of each chromatographic peak. Assume that a standard product with a known concentration is added to the sample, and then chromatographic analysis is performed to calculate the recovery rates, which are 98%, 102%, 95%, 105%, 99%, 101% respectively. Judge whether the recovery rates are within the preset range of 95% - 105%, and they are all within the range.According to the peak area, precision, and recovery rate, the comprehensive evaluation index of each chromatographic peak was obtained by the weighted average method. The set weights were 5, 3, and 2 respectively, and the calculated comprehensive evaluation indexes were 56, 47, 91, 68, 95, and 89. By judging the relationship between the comprehensive evaluation index and the set index 0, it was found that the comprehensive evaluation indexes of the second and third chromatographic peaks were greater than the set index, indicating that the separation effects of these two chromatographic peaks were good.

[0030] Step S105, by optimizing the objective function, the chromatographic peak signals obtained by the separation model conform to the mechanism of the chromatographic separation process, making full use of the complementary information of chromatography and mass spectrometry to improve the separation precision and guiding the resolution of complex overlapping peaks.

[0031] Collect chromatographic and mass spectrometry data to obtain a set of original signals containing multiple chromatographic peaks. The principal component analysis method was used to perform data dimensionality reduction on the original signals to obtain the dimensionality-reduced data. Based on the dimensionality-reduced data, an initial separation model was constructed to obtain the initially separated chromatographic peaks and the corresponding mass spectrometry information. Through the initially separated chromatographic peaks and mass spectrometry information, an objective function was constructed, including the peak shape parameters of the chromatographic peaks and the similarity measure of the mass spectrometry. If the peak shape of the chromatographic peak conforms to the Gaussian distribution, the objective function was optimized. The differential evolution algorithm was used to optimize the objective function to obtain the optimized separation model parameters. According to the optimized separation model parameters, the chromatographic peaks were re-separated to obtain the optimized chromatographic peak separation results and mass spectrometry information. It was judged whether the optimized chromatographic peak separation results met the preset quality control standards. If they met the preset quality control standards, the final chromatographic peak separation results and the corresponding mass spectrometry information were output.

[0032] Exemplarily, the collected chromatographic mass spectrometry data is a three-dimensional matrix, for example, with a size of 1000 (scanning time points) x 50 (mass-to-charge ratio channels) x 20 (number of samples), assuming the number of samples is 1. The original signal contains 5 overlapping chromatographic peaks. First, the three-dimensional data matrix is unfolded into a two-dimensional matrix along the sample dimension, and the size becomes 1000 x 50. Then, principal component analysis is performed on the two-dimensional matrix to calculate the eigenvalues and eigenvectors of the covariance matrix. The eigenvectors corresponding to the first 2 largest eigenvalues are selected as the principal components, and the original data is projected onto the two-dimensional space formed by these two principal components to obtain the data after dimensionality reduction, with the size becoming 1000 x 2. According to the data after dimensionality reduction, it can be observed that the data points are roughly distributed in 5 clusters, corresponding to 5 chromatographic peaks. Using the K-means clustering algorithm, the data points are divided into 5 categories, with each category corresponding to a chromatographic peak, obtaining the initially separated chromatographic peaks, and each chromatographic peak contains a set of time points. According to the time points of each chromatographic peak, the corresponding mass spectrometry data is extracted from the original two-dimensional data matrix to obtain the mass spectrometry information of each chromatographic peak. Assume that the retention time of the first initially separated chromatographic peak is between 200 and 300 scanning time points, the peak shape is approximately Gaussian distribution, its peak center is at the 250th scanning time point, the standard deviation is 20 scanning time points, and the corresponding mass spectrometry has high responses at mass-to-charge ratios of 100, 150, and 200. A target function is constructed, including the peak shape parameters (peak center, standard deviation) of the chromatographic peak and the mass spectrometry similarity measure. For example, for the peak shape parameters, the mean square error can be used to measure the difference between the fitted Gaussian distribution and the actual peak shape; for the mass spectrometry similarity, the cosine similarity can be used to measure the similarity between the mass spectrometry at different time points and the mass spectrometry at the peak center. The two metrics are weighted and summed to obtain the target function. Assume the weights are 6 and 4 respectively. The differential evolution algorithm is used to optimize the target function. The population size is set to 50, the number of iterations is 100, the crossover probability is 7, and the mutation factor is 5. In each iteration, according to the mutation, crossover, and selection operations of the differential evolution algorithm, the parameter values (peak center, standard deviation, mass spectrometry weights) of each individual in the population are updated. After 100 iterations, the optimized separation model parameters are obtained. For example, the peak center of the first optimized chromatographic peak becomes the 245th scanning time point, and the standard deviation becomes 18 scanning time points. According to the optimized model parameters, the probability of each time point belonging to each chromatographic peak is recalculated to obtain the optimized chromatographic peak separation result. For example, after optimization, the 295th scanning time point, which originally belonged to the first chromatographic peak, now belongs to the second chromatographic peak with a probability of 8. According to the optimized chromatographic peak assignment result, the mass spectrometry information of each chromatographic peak is recalculated. Quality control standards are set, for example, requiring that the peak area of each chromatographic peak is greater than 1000, the peak shape symmetry factor is between 8 and 2, and the signal-to-noise ratio of the mass spectrometry is greater than 10.If all of the five optimized chromatographic peaks meet these criteria, the final chromatographic peak separation results are output, including the retention time range, peak area, peak shape parameters, etc. of each chromatographic peak, as well as the corresponding mass spectrometry information. If all of the optimized chromatographic peaks do not meet the quality control criteria, the weights of the objective function or the parameters of the differential evolution algorithm need to be adjusted and the optimization is performed again until the criteria are met.

[0033] Step S106: Analyze the overlapping chromatographic peaks determined by principal component analysis using the established separation model to obtain the chromatographic peak signals representing single components after separation, and match them with the corresponding mass spectrometry diagrams.

[0034] Obtain a mixture sample, collect the original chromatographic-mass spectrometry data, perform mass spectrometry detection after separation by a chromatographic column to obtain the total ion current chromatogram and the mass spectrometry diagram corresponding to each time point. According to the total ion current chromatogram, use the derivative method or the inflection point detection method to determine the initial peak boundaries and preliminarily judge the number of overlapping peaks. Use the moving window factor analysis method or the interactive iterative target transformation factor analysis method to determine the number of principal components in each overlapping region. Through principal component analysis of each overlapping region, obtain the quantity information of the single components existing in each region. According to the principal component analysis results, construct a separation model, where the separation model includes a Gaussian function model, and use the least squares method to perform curve fitting on each overlapping region to obtain the peak shape parameters of each single component. Through iterative optimization of each overlapping peak region, use the Gauss-Newton method or the Levenberg-Marquardt method for iterative optimization to obtain the peak area of each single-component chromatographic peak. According to the peak areas of each single component, calculate the signal intensity of each single-component chromatographic peak and determine the relative content of different components. Compare the retention time of each single-component chromatographic peak obtained by separation with the retention time of the reference substance in the mass spectrometry library. If the difference between the two retention times is less than the preset threshold of 1 minute, the component is preliminarily determined. Through the mass spectrometry matching algorithm, use the mass spectrometry library to calculate the similarity of the mass spectrometry diagrams at each time point. If the similarity is greater than 8, it is determined that the component exists.

[0035] Exemplarily, assume that a mixture sample is obtained and analyzed by a gas chromatography-mass spectrometry (GC-MS) instrument, and an original data file is collected. After being processed by software, a total ion chromatogram (TIC) is obtained. By observing the TIC graph, it is found that there are obvious overlapping peaks between the retention times of 10 minutes and 15 minutes. First, apply the first derivative method to this region, calculate the difference in signal intensity between each data point and the next data point, and divide it by the time difference to obtain the first derivative curve. In the first derivative curve, positive peaks represent the rising edges in the original TIC graph, negative peaks represent the falling edges, and zero points correspond to the peak vertices or valley points in the original TIC graph. By observing the first derivative curve, it is found that there are at least 3 overlapping peaks in this region, and the number of overlapping peaks is initially judged to be 3. Then, use the moving window factor analysis method to further determine the number of principal components in the overlapping peak region. Set the window size to 5 minutes and the step size to 1 minute, and move the window within the range of 10 minutes to 15 minutes. Perform singular value decomposition (SVD) on the data matrix within each window to obtain singular values. According to the magnitudes of the singular values, select the eigenvectors corresponding to the first few larger singular values as the principal components. Usually, select the principal components with a cumulative contribution rate greater than 95%. Through analysis, it is found that the cumulative contribution rates of the first three principal components in each window are all greater than 95%, so it is determined that the number of principal components in this overlapping region is 3, that is, there are 3 single components. According to the principal component analysis results, a separation model is constructed. Since chromatographic peaks are usually approximately Gaussian distributed, a Gaussian function is selected as the separation model. Apply the least squares method to perform curve fitting on the overlapping region from 10 minutes to 15 minutes. The initial parameters are set as follows: the peak centers of the three peaks are located at 11 minutes, 15 minutes, and 14 minutes respectively, the peak widths are all 5 minutes, and the peak heights are estimated according to the intensity at this position in the TIC graph. Through iterative optimization by the least squares method, the peak shape parameters of each peak are obtained. For example, the peak centers are 12 minutes, 16 minutes, and 18 minutes respectively, the peak widths are 4 minutes, 6 minutes, and 5 minutes respectively, and the peak heights are 1000, 1500, and 800 respectively. Next, perform iterative optimization. Apply the Levenberg-Marquardt method to further optimize the parameters of each peak to minimize the sum of the squared residuals between the fitted curve and the original TIC graph. After multiple iterations, the optimized peak areas are 450, 680, and 320 respectively. Calculate the signal intensity based on the peak areas, and then determine the relative contents of each component. Assume that the response factors of the three components are the same, determine their relative contents, and finally perform component identification. Compare the retention times of 12 minutes, 16 minutes, and 18 minutes of the three separated peaks with the retention times in the standard substance library. Assume that the retention time of substance A in the library is 12.5 minutes, the retention time of substance B is 15.5 minutes, and the retention time of substance C is 17.5 minutes, all of which are less than the preset threshold of 1 minute. It is preliminarily determined that these three peaks correspond to substances A, B, and C respectively. Then, extract the mass spectra at the time points corresponding to these three peaks and calculate the similarity with the mass spectra in the standard substance library.Suppose the similarities between the mass spectra of the three calculated peaks and the standard mass spectra of substances A, B, and C are 92, 89, and 95 respectively, all greater than 8. Then it is finally determined that these three peaks represent substances A, B, and C respectively. To ensure the reliability of the results, data quality assessment can also be carried out through software. By checking indicators such as signal-to-noise ratio and resolution, ensure that the data quality meets the requirements, and by checking the fitting residual plot, ensure the accuracy of the fitting results.

[0036] Step S107: Combine the separated single chromatographic peak signals and mass spectra, and judge the component types corresponding to the chromatographic peaks according to the characteristic fragments and chromatographic retention times in the mass spectra, so as to realize the characterization of the composition of mixed aromatics.

[0037] Obtain the original chromatographic signal of the mixed aromatic hydrocarbon sample, perform baseline correction on the original chromatographic signal to obtain the chromatographic signal with background noise removed. Perform peak identification based on the chromatographic signal with background noise removed to determine the retention time corresponding to each chromatographic peak, and obtain the peak area corresponding to each chromatographic peak. Collect the mass spectrometry data corresponding to each chromatographic peak, perform mass number calibration on the mass spectrometry data to obtain the mass spectrum with accurate mass numbers. Analyze the mass spectrum with accurate mass numbers, identify the characteristic fragment ions, and obtain the characteristic fragment information corresponding to each chromatographic peak. Compare the characteristic fragment information corresponding to each chromatographic peak with the standard database. If the matching degree is higher than the set value, the component type is preliminarily determined. Combine the retention time information corresponding to each chromatographic peak. If the deviation between the standard retention time and the measured retention time of a certain component type is less than the predetermined range, further confirm the component type of the chromatographic peak. Calculate the relative content of each component in the mixed aromatics according to the peak area corresponding to each chromatographic peak and the confirmed component type information, and obtain the composition information of the mixed aromatics.

[0038] Exemplarily, assume that a mixed aromatics sample is analyzed by gas chromatography - mass spectrometry (GC - MS). There is an obvious baseline drift in the original chromatographic signal collected by the instrument. The wavelet transform method is used for baseline correction. The db4 wavelet basis function is selected for 5 - layer wavelet decomposition. Then, the high - frequency coefficients are processed by thresholding. The high - frequency coefficients less than the set threshold of 5 are set to zero. Finally, wavelet reconstruction is performed to obtain the chromatographic signal after baseline correction. The corrected chromatographic signal is smoothed by Gaussian filtering. The filter window size is set to 5 and the standard deviation is set to 1. Then, the first - derivative method is used for peak identification. The minimum peak height for peak identification is set to 1 and the minimum peak width is set to 3 seconds. The starting point and ending point of each peak are determined. The peak area of each peak is integrated, and the retention time corresponding to each peak is recorded. For example, the retention time of a chromatographic peak is 13.4 minutes and the peak area is 12345. The mass spectrometry data corresponding to 13.4 minutes is collected. The polynomial fitting method is used for mass number calibration. Five standard substance ion peaks with known mass numbers are selected for cubic polynomial fitting to obtain the mass number calibration curve. The mass - to - charge ratio data collected by the instrument is substituted into the calibration curve to obtain the accurate mass number. For example, the accurate mass number of a certain ion peak is 70469. The mass spectrometry diagram is analyzed, and the ion peaks with intensity greater than 5% are used as characteristic fragment ions. For example, characteristic fragment ions with mass numbers of 50391, 60469, and 70391 are obtained. These characteristic fragment information is compared with the NIST standard spectral library, and the component with the highest matching degree is retrieved as benzene, and the matching degree is 95%. Therefore, it can be confirmed that the component corresponding to this chromatographic peak is benzene. According to the same method, the components corresponding to other chromatographic peaks can be confirmed. For example, the chromatographic peak with a retention time of 16.7 minutes is confirmed as toluene, and the peak area is 23456. According to the peak areas of all the confirmed components and their corresponding response factors, the response factors can be obtained through standard sample tests. For example, the response factor of benzene is 2 and the response factor of toluene is 1. Calculate the relative content of each component in the mixed aromatics, so as to obtain the composition information of the mixed aromatics.

[0039] Step S108, perform quantitative analysis on the separated single chromatographic peak to obtain the mass fractions of monomeric compounds and component compounds.

[0040] Here, it specifically includes:

[0041] In the formula: i—the mass fraction of a certain monomer or a certain component in the sample, %(m / m);

[0042] Ai—the peak area value of a certain monomer or a certain component in the sample;

[0043] ∑Ai—the sum of the peak area values of all components in the sample.

[0044] The present invention discloses a method for analyzing the components of Fischer-Tropsch synthesis mixed aromatics based on comprehensive two-dimensional gas chromatography-mass spectrometry. This method first establishes a two-dimensional retention time matrix of sample components, and determines the chromatographic peak overlapping region through hierarchical clustering algorithm. For the overlapping region, principal component analysis is used to extract characteristic information and establish a multi-dimensional characteristic matrix. Then, an independent component analysis method is used to construct a separation model for overlapping chromatographic peaks, and the complex overlapping peaks are resolved by optimizing the objective function. Finally, combined with the separated single chromatographic peak signal and mass spectrum, the component type is judged and quantitative analysis is carried out. The present invention makes full use of the complementary information of chromatography and mass spectrometry, improves the separation precision and characterization ability of overlapping components in complex mixtures, provides accurate composition information feedback for optimizing the Fischer-Tropsch synthesis process, and improves the measurement precision of the mass fraction of each hydrocarbon component.

[0045] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and all should be covered within the protection scope of the present invention.

Claims

1. A method for determining hydrocarbon components in mixed aromatics using comprehensive two-dimensional gas chromatography-mass spectrometry, characterized in that: The method comprises: Obtain comprehensive two-dimensional gas chromatography-mass spectrometry data of Fischer-Tropsch synthesis mixed aromatics samples, and establish a two-dimensional retention time matrix of sample components, where each matrix element is the retention time value of a component on a non-polar and polar chromatographic column; A hierarchical clustering algorithm was used to cluster the retention time matrix, and homologues with similar structures were grouped into one category according to the Euclidean distance, their distribution areas on the two-dimensional chromatogram were determined, and the areas where the chromatographic peaks overlapped were found; For the overlapping regions of chromatographic peaks determined by cluster analysis, principal component analysis was used to extract characteristic information of overlapping peaks from the aspects of peak shape, peak width, retention time and mass spectrum fragmentation, and a multidimensional characteristic matrix of overlapping chromatographic peaks was established. Based on the feature matrix extracted by principal component analysis, an independent component analysis method was used to establish a separation model for overlapping chromatographic peaks by minimizing statistical independence, and an objective function was constructed to measure the physicochemical significance of the separated chromatographic peak signals; By optimizing the objective function, the chromatographic peak signal obtained by the separation model is made to conform to the mechanism of the chromatographic separation process, and the complementary information of chromatography and mass spectrometry is fully utilized to improve the separation accuracy and guide the analysis of complex aliasing peaks. The established separation model is used to analyze the overlapping chromatographic peaks determined by principal component analysis, and the chromatographic peak signals representing the single components after separation are obtained, which are matched with the corresponding mass spectra; Combining the single chromatographic peak signal obtained by separation with the mass spectrum, the component type corresponding to the chromatographic peak is determined according to the characteristic fragments and chromatographic retention time in the mass spectrum, thereby achieving the characterization of the composition of mixed aromatics; The separated single chromatographic peak is quantitatively analyzed to obtain the mass fractions of monomer compounds and component compounds.

2. The method according to claim 1, characterized in that The method comprises obtaining comprehensive two-dimensional gas chromatography-mass spectrometry data of the Fischer-Tropsch synthesis mixed aromatics sample, establishing a two-dimensional retention time matrix of the sample components, wherein each matrix element is a retention time value of a component on a non-polar and polar chromatographic column, including: Obtaining a Fischer-Tropsch synthesis mixed aromatics sample, analyzing the sample using comprehensive two-dimensional gas chromatography-mass spectrometry technology, and obtaining two-dimensional chromatogram data of the sample; According to the comprehensive two-dimensional gas chromatography-mass spectrometry analysis results, the retention time values ​​of each component in the mixed aromatic hydrocarbon sample on the non-polar chromatographic column and the polar chromatographic column were determined; The retention time values ​​of each component were analyzed by clustering algorithm, and the components were divided into different categories according to the similarity of retention time; For each type of component, a two-dimensional matrix of the retention time of the component on a non-polar chromatographic column and a polar chromatographic column is established respectively; Each two-dimensional retention time matrix is ​​analyzed, and the support vector machine algorithm is used to determine whether there are abnormal data points in the matrix. If there are abnormal points, they are removed; The two-dimensional retention time matrix of all categories of components was integrated to construct an overall two-dimensional retention time matrix model for mixed aromatic samples; The principal component analysis method was used to reduce the dimensionality of the overall two-dimensional retention time matrix to obtain a low-dimensional matrix that can reflect the component characteristics of the mixed aromatic hydrocarbon samples, laying the foundation for subsequent qualitative and quantitative analysis.

3. The method according to claim 1, characterized in that The hierarchical clustering algorithm is used to perform cluster analysis on the retention time matrix, and homologues with similar structures are classified into one category according to the Euclidean distance, and their distribution areas on the two-dimensional chromatogram are determined to find the areas where the chromatographic peaks overlap, including: Obtain retention time matrix data, pre-process the data in the matrix, remove noise and outliers, and ensure the accuracy and reliability of the data; Based on the preprocessed retention time matrix data, a hierarchical clustering algorithm was used for cluster analysis, and homologues with similar structures were grouped into one category by calculating the Euclidean distance between data points; According to the results of cluster analysis, the distribution area of ​​each type of homologues on the two-dimensional chromatogram was determined, and the distribution position and range of different types of homologues were obtained; Analyze the distribution of chromatographic peaks on the two-dimensional chromatogram, find out the area where chromatographic peaks overlap, and determine whether there are multiple chromatographic peaks of homologues appearing in the same position; If there is an area of ​​overlapping chromatographic peaks, the different homologues are further distinguished and identified by analyzing the characteristics of the chromatographic peak shapes and retention times in the overlapping area; According to the analysis results of the overlapping areas of chromatographic peaks, the two-dimensional chromatogram is optimized and adjusted to reduce the overlapping of chromatographic peaks by changing the chromatographic conditions or selecting a more suitable chromatographic column; The optimized two-dimensional chromatogram was compared with the original chromatogram to evaluate the optimization effect, determine the final homologue separation and identification results, and obtain accurate and reliable qualitative and quantitative analysis results.

4. The method according to claim 1, characterized in that: The method uses a principal component analysis method to extract characteristic information of overlapping peaks from peak shape, peak width, retention time, and mass spectrum fragmentation in the overlapping chromatographic peaks determined by cluster analysis, and establishes a multidimensional characteristic matrix of overlapping chromatographic peaks, including: Cluster analysis method is used to cluster the chromatographic peaks in the chromatogram, identify the overlapping chromatographic peak areas, and determine the target areas that need to be separated and analyzed; For the overlapping chromatographic peak regions determined by cluster analysis, the one-dimensional features of peak shape, peak width and retention time, as well as the two-dimensional features of mass spectrum fragment information were extracted to construct a multidimensional feature matrix of overlapping chromatographic peaks. According to the constructed multi-dimensional feature matrix, the principal component analysis method is used to map the high-dimensional feature data into a low-dimensional space, extract the main feature information of overlapping chromatographic peaks, and reduce the feature dimension; The principal component score matrix obtained through principal component analysis is used to characterize overlapping chromatographic peaks, highlight the differences between different overlapping peaks, and provide a basis for subsequent peak separation; According to the results of principal component analysis, the multivariate curve resolution algorithm is used to analyze the overlapping chromatographic peaks, estimate the peak shape and peak area parameters of each overlapping peak, and realize the quantitative analysis of overlapping peaks; Combined with the mass spectrometry fragment information, the components of each chromatographic peak obtained by separation are identified, the components of the substances contained in the overlapping peaks are inferred, and the qualitative analysis is completed; The results of qualitative and quantitative analysis are combined to obtain complete analytical results of overlapping chromatographic peaks, including the retention time, peak area and material composition information of each overlapping peak.

5. The method according to claim 1, characterized in that The feature matrix extracted by principal component analysis is used to establish a separation model for overlapping chromatographic peaks by minimizing statistical independence using an independent component analysis method, and to construct an objective function for measuring the physicochemical significance of the separated chromatographic peak signals, including: Perform data dimensionality reduction processing based on the initial feature matrix obtained by principal component analysis to obtain a low-dimensional first feature matrix; The first feature matrix is ​​processed by independent component analysis method to obtain multiple second independent components that are independent of each other, and the combination with minimum statistical independence is determined; Constructing a separation model of overlapping chromatographic peaks according to the second independent component to obtain the first signal of each single chromatographic peak after separation; The objective function is constructed by using the first signal to obtain the peak area corresponding to each single chromatographic peak, and to determine the magnitude relationship between the peak area and a preset threshold value; If the peak area is greater than a preset threshold, the precision of the second signal of each chromatographic peak is calculated based on the peak area; Obtaining the recovery rate of the second signal of each chromatographic peak, and obtaining the relationship between the recovery rate and the preset range; According to the peak area, precision and recovery rate, the weighted average method was used to obtain the comprehensive evaluation index of each chromatographic peak, and the relationship between the comprehensive evaluation index and the set index was determined.

6. The method according to claim 1, characterized in that The method optimizes the objective function so that the chromatographic peak signal obtained by the separation model conforms to the mechanism of the chromatographic separation process, makes full use of the complementary information of the chromatogram and the mass spectrum, improves the separation accuracy, and is used to guide the analysis of complex aliasing peaks, including: Collecting chromatographic mass spectrometry data to obtain a set of original signals including multiple chromatographic peaks; The principal component analysis method is used to perform data dimension reduction processing on the original signal to obtain the reduced-dimensional data; According to the data after dimension reduction, an initial separation model is constructed to obtain the initial separated chromatographic peaks and corresponding mass spectrum information; The objective function is constructed through the initially separated chromatographic peak and mass spectrum information, including the peak shape parameters of the chromatographic peak and the similarity measurement of the mass spectrum. If the peak shape of the chromatographic peak conforms to the Gaussian distribution, the objective function is optimized; The differential evolution algorithm is used to optimize the objective function and obtain the optimized separation model parameters; Re-separate the chromatographic peaks according to the optimized separation model parameters to obtain optimized chromatographic peak separation results and mass spectrum information; Determine whether the optimized chromatographic peak separation result meets the preset quality control standard. If it meets the preset quality control standard, output the final chromatographic peak separation result and the corresponding mass spectrum information.

7. The method according to claim 1, characterized in that The method uses the established separation model to analyze the overlapping chromatographic peaks determined by the principal component analysis to obtain the chromatographic peak signal representing a single component after separation, and matches it with the corresponding mass spectrum, including: Obtain a mixture sample, collect original chromatography-mass spectrometry data, perform mass spectrometry detection after separation by a chromatographic column, and obtain a total ion current chromatogram and a mass spectrum corresponding to each time point; Based on the total ion current chromatogram, the initial peak boundary is determined by using the derivative method or the inflection point detection method, and the number of overlapping peaks is preliminarily determined; The number of principal components in each overlapping region is determined by using the moving window factor analysis method or the interactive iterative target transformation factor analysis method. By performing principal component analysis on each overlapping region, the quantitative information of the single component existing in each region is obtained. According to the results of principal component analysis, a separation model is constructed, which includes a Gaussian function model. The least square method is used to perform curve fitting on each overlapping area to obtain the peak shape parameters of each single component. The peak area of ​​each single component chromatographic peak is obtained by iteratively optimizing each overlapping peak area using the Gauss-Newton method or the Levenberg-Marquardt method; According to the peak area of ​​each single component, the signal intensity of each single component chromatographic peak is calculated to determine the relative content of different components; The retention time of each single component chromatographic peak separated is compared with the retention time of the standard substance in the mass spectrum library. If the difference between the two retention times is less than the preset threshold of 1 minute, the component is preliminarily determined; The mass spectrum matching algorithm is used to calculate the similarity of the mass spectra at each time point using the mass spectrum library. If the similarity is greater than 8, the component is confirmed to exist.

8. The method according to claim 1, characterized in that The single chromatographic peak signal obtained by combining the separation and the mass spectrum, judging the component type corresponding to the chromatographic peak according to the characteristic fragments and chromatographic retention time in the mass spectrum, and realizing the characterization of the mixed aromatic hydrocarbon composition, comprises: Acquire the original chromatographic signal of the mixed aromatic hydrocarbon sample, perform baseline correction on the original chromatographic signal, and obtain the chromatographic signal with background noise removed; Perform peak identification based on the chromatographic signal after background noise removal, determine the retention time corresponding to each chromatographic peak, and obtain the peak area corresponding to each chromatographic peak; Collecting mass spectrum data corresponding to each chromatographic peak, calibrating the mass spectrum data to obtain a mass spectrum with an accurate mass number; Analyze the mass spectrum with accurate mass number, identify characteristic fragment ions, and obtain characteristic fragment information corresponding to each chromatographic peak; The characteristic fragment information corresponding to each chromatographic peak is compared with the standard database. If the matching degree is higher than the set value, the component type is preliminarily determined; Combined with the retention time information corresponding to each chromatographic peak, if the deviation between the standard retention time of a component type and the measured retention time is less than a predetermined range, the component type of the chromatographic peak is further confirmed; According to the peak area corresponding to each chromatographic peak and the confirmed component type information, the relative content of each component in the mixed aromatics is calculated to obtain the composition information of the mixed aromatics.

9. The method according to claim 1, characterized in that: Quantitative analysis is performed on the separated single chromatographic peak to obtain the mass fractions of monomer compounds and component compounds, including: Where: i —Mass fraction of a monomer or component in the sample, % (m / m); A i —The peak area value of a monomer or component in the sample; ∑A i —The sum of the peak area values ​​of all components in the sample.

Images