Data processing method and system applied to gas chromatographic analysis

Abstract

The invention discloses a data processing method applied to gas chromatographic analysis. The data processing method comprises the following steps: acquiring a sample signal in a gas chromatographic analyzer; decomposing the sample signal into a plurality of intrinsic mode functions (IMF) by adopting an optimized ICEEMDA algorithm; wherein the optimized ICEEMDA algorithm is used for obtaining parameters through a constraint penalty leech optimization algorithm; constructing a comprehensive evaluation index and selecting a corresponding intrinsic mode function IMF for signal reconstruction to obtain a reconstructed signal after noise reduction; identifying a wave crest in the reconstructed signal through a peak detection algorithm, and extracting related parameters of the wave crest by using Gaussian fitting; and generating a corresponding analysis report based on the related parameters of the wave crests. According to the method, ICEEMDA algorithm parameters are optimized, an intrinsic mode function is screened by adopting a comprehensive evaluation index, and a refined peak detection step is adopted, so that the signal processing efficiency of a complex sample, the noise suppression capability and the accuracy of quantitative and qualitative analysis in the field of gas chromatography are improved.

Description

Technical Field

[0001] The present invention relates to the field of gas chromatography analysis, and more specifically to a data processing method and system applied to gas chromatography analysis. Background Art

[0002] Gas chromatography is a chemical analysis method widely used in many fields such as environmental monitoring, food safety, and petrochemicals. It uses the difference in the distribution equilibrium of different substances between the stationary phase and the mobile phase to separate and detect the components in the mixture. Through this process, the compounds in the sample can be qualitatively and quantitatively analyzed, thus providing a scientific basis for environmental protection, food quality control, and quality monitoring in industrial production processes.

[0003] In the field of gas chromatography analysis, data processing relies on manual interpretation and calculation of raw signals, which is not only inefficient but also prone to human errors. The identification of peaks and area calculations during manual processing are often not accurate enough, especially when facing complex samples. This manual-dependent method is difficult to ensure the consistency and accuracy of the results, thus affecting the quality of the final analysis report.

[0004] In addition, although the smoothing technology and filtering methods in the existing processing methods can remove noise to a certain extent, they perform poorly when processing raw signals containing a large amount of noise components. Especially when faced with complex peak structures, these traditional methods cannot effectively distinguish and accurately calculate the peak area, resulting in limited accuracy of quantitative analysis. The presence of noise not only affects the accurate positioning of the peak, but may also lead to inaccurate calculation of the peak area, which in turn affects the correct determination of the concentration of the compound in the sample.

[0005] Existing signal reconstruction technologies lack an effective intrinsic mode function (IMF) selection mechanism. After decomposing the original signal into multiple IMFs, it is a difficult problem to select the part that best represents the characteristics of the original signal from the numerous IMFs for reconstruction. Traditional methods often cannot solve this problem well, resulting in the reconstructed signal may lose important information or fail to fully reduce noise, affecting the quality of subsequent data analysis.

[0006] Therefore, how to design a data processing method for gas chromatography analysis to improve the noise reduction effect and peak recognition accuracy of complex sample signals in the field of gas chromatography analysis is an urgent problem to be solved by technical personnel in this field. Summary of the invention

[0007] In view of this, the present invention provides a data processing method applied to gas chromatography analysis, which improves the efficiency of complex sample signal processing, noise suppression ability and accuracy of quantitative and qualitative analysis in the field of gas chromatography analysis; and provides a more accurate and reliable means of chemical composition analysis for industries such as environmental monitoring, food safety and petrochemical industry.

[0008] In order to achieve the above object, the present invention adopts the following technical solution:

[0009] In a first aspect, the present invention provides a data processing method for gas chromatography analysis, comprising the following steps:

[0010] S1, obtaining a sample signal from a gas chromatograph;

[0011] S2, using the optimized ICEEMDA algorithm to decompose the sample signal into multiple intrinsic mode functions IMF; wherein the optimized ICEEMDA algorithm obtains parameters through a constrained penalty leech optimization algorithm; the parameters include: the optimal number NR of noise enhancement signals and the optimal standard deviation Nstd of Gaussian white noise;

[0012] S3, construct a comprehensive evaluation index to select the corresponding intrinsic mode function IMF for signal reconstruction, and obtain the reconstructed signal after noise reduction;

[0013] S4, identifying the peaks in the reconstructed signal by using a peak detection algorithm, and extracting relevant parameters of the peaks by using Gaussian fitting;

[0014] S5. Generate a corresponding analysis report based on relevant parameters of the peak.

[0015] Furthermore, in S2, the constraint penalty leech optimization algorithm is obtained by improving the leech optimization algorithm, and is combined with the Tent chaotic mapping and the constraint penalty factor.

[0016] Furthermore, the Tent chaotic map is expressed as:

[0017]

[0018] Among them, x n 、x n+1 They represent the state values ​​of the nth step and the n+1th step respectively, and μ represents the control parameter.

[0019] Furthermore, the constraint penalty factor is expressed as:

[0020]

[0021] Among them, epoth represents the current iteration number, and epoths represents the total iteration number.

[0022] Furthermore, the S2 includes:

[0023] S21, add noise to the sample signal x(t) to generate NR noise-enhanced signals x i (t):

[0024] x i (t) = x(t) + Nstdw i (t)

[0025] Among them, w i (t) represents Gaussian white noise, i = 1, 2, ... NR;

[0026] S22, for each noise enhancement signal x i (t) Perform EMD decomposition to obtain the first layer intrinsic mode function c of each noise 1,i (t);

[0027] S23, calculate the first layer intrinsic mode function c 1,i The average value of (t)

[0028]

[0029] S24, based on average Update residual r1(t):

[0030]

[0031] S25, add noise to the residual of the k-1th layer to obtain NR noise-enhanced signals r k-1,i (t):

[0032] r k-1,i (t) = r k-1 (t)+Nstdw i (t)

[0033] S26, for each noise enhancement signal r k-1,i (t) Perform EMD decomposition to obtain the k-th layer intrinsic mode function c of each noise k,i (t);

[0034] S27, calculate the k-th layer intrinsic mode function c k,i The average value of (t)

[0035]

[0036] S28, based on average Update residual r k (t):

[0037]

[0038] S29, until the preset number of layers k is met, stop decomposing and output the intrinsic mode functions IMFs of the 1st to kth layers.

[0039] Furthermore, in S3, the comprehensive evaluation index is expressed as:

[0040]

[0041] Among them, τ represents the correlation coefficient, γ represents the energy ratio, and ω represents the reconstruction error.

[0042] Furthermore, in S4, identifying the peaks in the reconstructed signal by using a peak detection algorithm includes:

[0043] S41, calculating the first-order derivative of the reconstructed signal, and identifying the data point where the sign of the derivative changes;

[0044] S42, based on the data points, determining a data point at which a change in the derivative exceeds a preset slope threshold as an initial peak;

[0045] S43, determining an initial wave peak exceeding a preset amplitude threshold as a wave peak.

[0046] Furthermore, in S4, the relevant parameters of the peaks include: the total number of peaks, the total area of ​​the peaks, the position of each peak, the height of each peak, the width of each peak, the area of ​​each peak and the proportion of the area of ​​each peak in the total area of ​​the peaks.

[0047] Furthermore, in S4, the area of ​​each peak is calculated by using the tangent method:

[0048]

[0049] Wherein, H represents the half height of the peak value, and F represents the full width at half maximum.

[0050] In a second aspect, the present invention provides a data processing system for gas chromatography analysis, comprising:

[0051] Signal acquisition module: used to acquire sample signals from gas chromatograph;

[0052] Signal decomposition module: used to decompose the sample signal into multiple intrinsic mode functions IMF using the optimized ICEEMDA algorithm;

[0053] IMF selection and signal reconstruction module: used to construct comprehensive evaluation indicators to select the corresponding intrinsic mode function IMF for signal reconstruction and obtain the reconstructed signal after noise reduction;

[0054] Peak detection and parameter extraction module: used to identify the peaks in the reconstructed signal through the peak detection algorithm, and extract the relevant parameters of the peaks using Gaussian fitting;

[0055] Report generation module: used to generate corresponding analysis reports based on relevant parameters of the peak.

[0056] It can be seen from the above technical solution that compared with the prior art, the technical solution of the present invention has the following advantages:

[0057] Beneficial effects:

[0058] 1. The parameter acquisition process of the traditional ICEEMDA (Improved Complete Ensemble Empirical Mode Decomposition) algorithm is improved by adopting the constrained penalty leech optimization algorithm. It combines the Tent chaos map and the constraint penalty factor, which makes it possible to search for the optimal solution more effectively when processing nonlinear and non-stationary signals, thereby improving the selection accuracy of key parameters such as NR (optimal number of noise-enhanced signals) and Nstd (optimal standard deviation of Gaussian white noise) in noise-enhanced signals. It not only enhances the effectiveness of decomposition of complex sample signals, but also improves the quality of the intrinsic mode function (IMF), which helps to improve the accuracy of the final analysis results.

[0059] 2. Constructing a comprehensive evaluation index (taking into account factors such as correlation coefficient, energy ratio and reconstruction error) ensures that the reconstructed signal after denoising can not only maintain the main characteristics of the original signal, but also significantly reduce the impact of background noise, thereby providing a clearer and more reliable data basis for subsequent peak detection. This is particularly important in the field of gas chromatography analysis because it directly affects the accuracy of qualitative and quantitative analysis of compounds.

[0060] 3. In the peak recognition stage, by carefully analyzing the changes in the first-order derivative of the reconstructed signal and setting reasonable slope thresholds and amplitude thresholds to filter out false or inconspicuous fluctuations, it can more accurately locate the actual valid peaks. Subsequently, the Gaussian function is used to mathematically model these peaks, and the extraction process of peak parameters, such as peak position, peak width and other information, is further refined, which not only simplifies the subjective bias problem that may be caused by traditional manual threshold setting, but also greatly improves the efficiency and quality of data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0062] Figure 1A flow chart of a data processing method for gas chromatography analysis provided by an embodiment of the present invention;

[0063] Figure 2 A schematic diagram of a process of identifying peaks in a reconstructed signal by using a peak detection algorithm provided in an embodiment of the present invention;

[0064] Figure 3 A schematic diagram of relevant parameters of a peak provided by an embodiment of the present invention;

[0065] Figure 4 A framework diagram of a data processing system for gas chromatography analysis provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0066] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0067] Embodiment 1;

[0068] like Figure 1 As shown, this embodiment provides a data processing method applied to gas chromatography analysis, comprising the following steps:

[0069] S1, obtaining a sample signal from a gas chromatograph;

[0070] S2, using the optimized ICEEMDA algorithm to decompose the sample signal into multiple intrinsic mode functions IMF; wherein the optimized ICEEMDA algorithm obtains parameters through a constrained penalty leech optimization algorithm; the parameters include: the optimal number NR of noise enhancement signals and the optimal standard deviation Nstd of Gaussian white noise;

[0071] S3, construct a comprehensive evaluation index to select the corresponding intrinsic mode function IMF for signal reconstruction, and obtain the reconstructed signal after noise reduction;

[0072] S4, identifying the peaks in the reconstructed signal by using a peak detection algorithm, and extracting relevant parameters of the peaks by using Gaussian fitting;

[0073] S5. Generate a corresponding analysis report based on relevant parameters of the peak.

[0074] This method optimizes the parameters of the ICEEMDA algorithm, uses Tent chaos mapping and constraint penalty factors to improve the accuracy of key parameter selection; uses comprehensive evaluation indicators to screen the intrinsic mode function for signal reconstruction to ensure the quality of the signal after noise reduction; and combines refined peak detection steps to accurately identify and model peaks. It significantly improves the efficiency of complex sample signal processing, noise suppression ability, and accuracy of quantitative and qualitative analysis in the field of gas chromatography analysis, providing a more accurate and reliable means of chemical composition analysis for industries such as environmental monitoring, food safety, and petrochemicals.

[0075] The above steps and related technical features are further described in detail below:

[0076] In this embodiment S1, a sample signal in a gas chromatograph is obtained;

[0077] When performing gas chromatography analysis, the sample to be analyzed must first be introduced into the gas chromatograph through an appropriate injection technique (such as a liquid autosampler or a gas injection valve). After the sample enters the gas chromatography column, it is separated according to the distribution equilibrium differences between the stationary phase and the mobile phase of each component in the sample under the impetus of the carrier gas. As different components flow out of the chromatographic column in turn, they pass through the detector (such as a flame ionization detector FID, a thermal conductivity detector TCD, etc.), during which the chemical signal is converted into an electrical signal. These electrical signals are the original signals, which contain information about the composition of the sample.

[0078] In order to ensure the effectiveness of subsequent data processing, the following points should be noted in this step:

[0079] Ensure that the sample is representative and that necessary pretreatment (such as filtration, dilution, etc.) is performed before injection to reduce interference factors. The gas chromatograph and its detector are calibrated regularly to ensure that their working state is stable so as to obtain reliable data. The working parameters of the gas chromatograph, including temperature program, carrier gas flow rate, etc., are reasonably set to optimize the separation effect. It also includes using a high-precision data acquisition system to record the signal output by the detector to ensure that the quality of the original signal obtained is good for subsequent analysis.

[0080] Through the above steps, high-quality sample signals can be effectively obtained from the gas chromatograph as the basic input for subsequent ICEEMDAN algorithm processing.

[0081] In this embodiment S2, the optimized ICEEMDA algorithm is used to decompose the sample signal into multiple intrinsic mode functions IMFs;

[0082] Specifically, in the leech optimization algorithm, leeches are divided into two types: directional leeches and non-directional leeches. When directional leeches are stimulated, they will swim towards humans, while non-directional leeches will move randomly in the search space. The exploration strategy of directional leeches is that when they feel stimulated, some leeches will swim towards humans at a slight angle. After multiple iterations, some leeches gradually approach humans and enter the development stage. After being stimulated, some non-directional leeches will respond in the wrong direction. As the iterations increase, more and more leeches find or approach humans. Finally, some leeches that successfully find humans and suck blood will execute a re-tracking strategy and throw the leeches attached to their feet back into the rice fields to find humans again.

[0083] The constrained penalty leech optimization algorithm is obtained by improving the leech optimization algorithm. The constrained penalty leech optimization algorithm combines the Tent chaotic mapping and the constraint penalty factor.

[0084] The Tent chaos map is expressed as:

[0085]

[0086] Among them, x n 、x n+1 They represent the state values ​​of the nth step and the n+1th step respectively, and μ represents the control parameter.

[0087] Tent chaotic mapping is a special chaotic sequence generation method, which adjusts the chaotic behavior by controlling the parameter μ, making the movement of leeches in the search space more regular.

[0088] The constraint penalty factor is expressed as:

[0089]

[0090] Among them, epoth represents the current iteration number, and epoths represents the total iteration number.

[0091] The constraint penalty factor is used to adjust the convergence speed and global search ability of leeches during the search process. It decreases with the increase of the number of iterations, thereby avoiding falling into the local optimal solution too early.

[0092] In the specific context of applying the constrained penalty leech optimization algorithm to obtain the parameters of the ICEEMDAN algorithm, "directed leeches" can be understood as search individuals that can move more directly toward the optimal solution, while "undirected leeches" are individuals that randomly wander in the solution space to explore more possibilities. "Searching for humans" can be seen as a process of finding the optimal number of noise enhancement signals NR and the standard deviation of Gaussian white noise Nstd and other key parameters.

[0093] Specifically, in this embodiment, the number of iterations is set to 20, the population size is set to 10, the standard deviation range is [0, 10], the number range is [0, 50], the best standard deviation obtained after ICEEMDAN is 0.007472, and the best number of noise enhancement signals is 5. By continuously iteratively adjusting these parameter values, the ultimate goal is to enable the ICEEMDAN algorithm to achieve the best noise reduction effect and signal decomposition quality when processing gas chromatographic raw signals.

[0094] Furthermore, the optimized ICEEMDA algorithm is used to decompose the sample signal into multiple intrinsic mode functions IMF, including:

[0095] S21, add noise to the sample signal x(t) to generate NR noise-enhanced signals x i (t):

[0096] x i (t) = x(t) + Nstdw i (t)

[0097] Among them, w i (t) represents Gaussian white noise, i = 1, 2, ... NR;

[0098] S22, for each noise enhancement signal x i (t) Perform EMD decomposition to obtain the first layer intrinsic mode function c of each noise 1,i (t);

[0099] S23, calculate the first layer intrinsic mode function c 1,i The average value of (t)

[0100]

[0101] S24, based on average Update residual r1(t):

[0102]

[0103] S25, add noise to the residual of the k-1th layer to obtain NR noise-enhanced signals r k-1,i (t):

[0104] r k-1,i (t) = r k-1 (t)+Nstdw i (t)

[0105] S26, for each noise enhancement signal r k-1,i (t) Perform EMD decomposition to obtain the k-th layer intrinsic mode function c of each noise k,i (t);

[0106] S27, calculate the k-th layer intrinsic mode function c k,i The average value of (t)

[0107]

[0108] S28, based on average Update residual r k (t):

[0109]

[0110] S29, until the preset number of layers k is met, stop decomposing and output the intrinsic mode functions IMFs of the 1st to kth layers.

[0111] Through this step, the optimized ICEEMDA algorithm can gradually extract the various frequency components in the signal, thereby obtaining a series of intrinsic mode functions. These IMFs can correspond to different frequency ranges, which is helpful for further analysis and processing of signals. In gas chromatography analysis, these IMFs can be used to distinguish the response curves of different compounds and improve the accuracy and reliability of analysis.

[0112] In this embodiment S3, a comprehensive evaluation index is constructed to select a corresponding intrinsic mode function IMF to reconstruct the signal and obtain a reconstructed signal after noise reduction;

[0113] The comprehensive evaluation index is expressed as:

[0114]

[0115] Among them, τ represents the correlation coefficient, γ represents the energy ratio, and ω represents the reconstruction error.

[0116] Specifically, the correlation coefficient τ is expressed as:

[0117]

[0118] Among them, x i represents the sample value of the original data, Represents the mean of the original data sample values, y i represents the sample value of the reconstructed data, represents the mean of the reconstructed data sample values, and o represents the number of samples;

[0119] The energy ratio γ is expressed as:

[0120]

[0121] Among them, λ irepresents the i-th eigenvalue of the data, k represents the number of selected features, and l represents the total number of features;

[0122] The reconstruction error ω is expressed as:

[0123]

[0124] Among them, x i Represents the original data, represents the reconstructed data, o represents the number of samples;

[0125] The above correlation coefficient is used to measure the correlation between the original data and the reconstructed data. If the two are highly correlated, it means that the reconstructed signal can better maintain the key characteristics of the original signal. The energy ratio is used to evaluate the energy distribution of the reconstructed data. If the energy of the reconstructed data is concentrated on certain specific IMFs, then these IMFs are more likely to contain the main signal information. The reconstruction error is used to measure the gap between the reconstructed data and the original data. A smaller reconstruction error means a better signal recovery effect.

[0126] The comprehensive evaluation index in this step combines the above three aspects, and selects the optimal IMFs for reconstruction by comparing the correlation, energy distribution and error size of the original data and the reconstructed data. The purpose of this is to remove noise as much as possible while retaining the main features of the signal and improving the accuracy of signal processing.

[0127] In this embodiment S4, the peak in the reconstructed signal is identified by a peak detection algorithm, and the relevant parameters of the peak are extracted by Gaussian fitting;

[0128] like Figure 2 As shown, the peaks in the reconstructed signal are identified by a peak detection algorithm, including:

[0129] S41. Calculate the first-order derivative of the reconstructed signal and identify the data points where the derivative sign changes; the first-order derivative reflects the rate of change of the signal. When the signal changes from rising to falling or from falling to rising, the first-order derivative will change its sign (i.e., from positive to negative or from negative to positive). By detecting these derivative sign change points, the potential peak position can be preliminarily located.

[0130] S42, based on the data points, determine the data points whose derivative changes exceed the preset slope threshold as the initial peak; specifically, in this embodiment, the slope threshold is set to 0.01, and the slope threshold is set to exclude those small fluctuations that are not obvious or are only caused by noise. Only those points with sufficiently large slope changes will be identified as potential peak starting points or end points. This helps to remove false peaks and improve the accuracy of peak detection.

[0131] S43, the initial peak exceeding the preset amplitude threshold is determined as a peak. Specifically, in this embodiment, the amplitude threshold is set to 2, and among all the data points considered as initial peaks, only those points whose peak height exceeds the preset amplitude threshold are retained. This threshold is used to distinguish truly meaningful peaks from background noise or smaller interference signals. Only when the height of the peak is higher than the surrounding area, it is considered to be a valid peak.

[0132] Furthermore, Gaussian fitting is used to extract relevant parameters of the peak, including:

[0133] For each initially confirmed peak, an initial Gaussian function model is set. Some initial estimates need to be defined, which can be based on the results of peak detection, such as the maximum value of the peak as the initial value of the peak height, the position of the peak as the initial value of the approximate position of the peak, and half of the peak half-height width as the initial value of the peak width.

[0134] An optimization algorithm such as nonlinear least squares is used to best match the Gaussian function to the actual waveform. The process aims to find a set of parameters that minimizes the difference between the theoretical curve and the experimental data. The background baseline may also need to be considered in the fitting process and separated from the peak signal to ensure a more accurate fitting result.

[0135] like Figure 3 As shown, the relevant parameters of the peak include:

[0136] Total number of peaks: After fitting, confirm and count the number of all valid peaks.

[0137] Total peak area: obtained by summing the areas of all peaks. The area of ​​a single peak can be calculated using the integral Gaussian function formula.

[0138] The position of each peak is determined by the mean of the fitted Gaussian function.

[0139] The height of each peak: taken directly from the amplitude of the Gaussian function.

[0140] The width of each peak: expressed as half-width, which can be calculated from the standard deviation of the Gaussian function.

[0141] The proportion of the area of ​​each peak in the total area of ​​the peaks: calculate the ratio of the area of ​​each peak to the total area of ​​the peaks and give it as a percentage.

[0142] Furthermore, in addition to calculating the peak area by integrating the Gaussian function formula in the above Gaussian fitting, the tangent method is also used to calculate the area of ​​each peak, which is particularly suitable for processing peaks with relatively regular shapes.

[0143]

[0144] Among them, H represents the half height of the peak, and F represents the half-height full width. Specifically, find the maximum value point of the peak and determine half of the peak height (H); find the points on both sides of the peak that intersect with this half-height point, and the distance between these two points is the half-height full width (F).

[0145] In the gas chromatography analysis of this embodiment, the tangent method is used to calculate the area of ​​each peak, which is an effective supplement to Gaussian fitting. By combining these two methods, the accuracy and reliability of peak area calculation can be ensured while ensuring calculation efficiency.

[0146] In this embodiment S5, a corresponding analysis report is generated based on the relevant parameters of the peak;

[0147] The analysis report includes the following key indicators to facilitate further data interpretation and quantitative analysis of sample composition:

[0148] Total number of peaks: The report clearly lists the total number of independent peaks detected. This value reflects the number of distinguishable components in the sample.

[0149] Total Peak Area: Calculates and reports the sum of all identified peak areas. The total peak area is an important reference value for the total amount of all components in the sample.

[0150] Specific position of each peak: Lists the specific position of each peak on the time axis or retention time (usually in minutes). These positions are critical for qualitative analysis and can be compared with the standard retention times of known compounds to determine the components in the sample.

[0151] Height of Each Peak: Provides the maximum amplitude of each peak. Peak height can be used to estimate the concentration of a single component, especially when comparing different samples under the same conditions.

[0152] The width of each peak: describes the extent to which each peak extends along the horizontal axis (usually time). The peak width can be expressed as the full width at half maximum, which helps to evaluate the chromatographic separation effect; narrower peaks often mean better separation.

[0153] Area of ​​each peak: The area of ​​each peak is calculated by tangent method or Gaussian fitting and listed in the report. The peak area directly reflects the relative content or absolute amount of the component in the sample and is one of the core indicators of quantitative analysis.

[0154] The ratio of each peak area to the total area: Calculate and report the ratio of each peak area to the total peak area (given as a percentage). This helps understand the importance of each component relative to the entire sample, especially in the analysis of mixture composition.

[0155] This embodiment provides a data processing method applied to gas chromatography analysis, aiming to improve the efficiency and accuracy of signal processing through an optimized ICEEMDA algorithm and Gaussian fitting technology.

[0156] The main steps include: first, obtaining high-quality sample signals from the gas chromatograph; then, using the ICEEMDA algorithm optimized by the constrained penalty leech optimization algorithm to decompose the signal into multiple intrinsic mode functions (IMFs), and selecting the optimal IMF for signal reconstruction through comprehensive evaluation indicators to obtain the reconstructed signal after noise reduction; then, using the peak detection algorithm combined with Gaussian fitting to identify and extract the relevant parameters of the peak, such as position, height, width and area; finally, generating a detailed analysis report based on these parameters.

[0157] The whole process not only improves the processing efficiency and noise suppression ability of complex sample signals, but also significantly enhances the accuracy of quantitative and qualitative analysis, providing a more accurate and reliable means of chemical composition analysis for industries such as environmental monitoring, food safety, and petrochemicals. In addition, by combining the tangent method to calculate the peak area, the accuracy and reliability of the peak area calculation are further ensured.

[0158] Embodiment 2;

[0159] like Figure 4 As shown, this embodiment provides a data processing system applied to gas chromatography analysis, comprising:

[0160] Signal acquisition module: used to acquire sample signals from gas chromatograph;

[0161] Signal decomposition module: used to decompose the sample signal into multiple intrinsic mode functions IMF using the optimized ICEEMDA algorithm;

[0162] IMF selection and signal reconstruction module: used to construct comprehensive evaluation indicators to select the corresponding intrinsic mode function IMF for signal reconstruction and obtain the reconstructed signal after noise reduction;

[0163] Peak detection and parameter extraction module: used to identify the peaks in the reconstructed signal through the peak detection algorithm, and extract the relevant parameters of the peaks using Gaussian fitting;

[0164] Report generation module: used to generate corresponding analysis reports based on relevant parameters of the peak.

[0165] The system consists of five key modules, aiming to improve the processing efficiency of complex sample signals, noise suppression capabilities and the accuracy of quantitative and qualitative analysis, providing strong technical support for chemical composition analysis in related industries.

[0166] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0167] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A data processing method for gas chromatography analysis, characterized in that: The following steps are involved: S1, obtaining a sample signal from a gas chromatograph; S2, using the optimized ICEEMDA algorithm to decompose the sample signal into multiple intrinsic mode functions IMF; wherein the optimized ICEEMDA algorithm obtains parameters through a constrained penalty leech optimization algorithm; the parameters include: the optimal number NR of noise enhancement signals and the optimal standard deviation Nstd of Gaussian white noise; S3, construct a comprehensive evaluation index to select the corresponding intrinsic mode function IMF for signal reconstruction, and obtain the reconstructed signal after noise reduction; S4, identifying the peaks in the reconstructed signal by using a peak detection algorithm, and extracting relevant parameters of the peaks by using Gaussian fitting; S5. Generate a corresponding analysis report based on relevant parameters of the peak.

2. A data processing method for gas chromatography analysis according to claim 1, characterized in that: In S2, the constraint penalty leech optimization algorithm is obtained by improving the leech optimization algorithm, and is combined with the Tent chaos map and the constraint penalty factor.

3. A data processing method for gas chromatography analysis according to claim 2, characterized in that: The Tent chaos map is expressed as: Among them, x n 、x n+1 They represent the state values ​​of the nth step and the n+1th step respectively, and μ represents the control parameter.

4. A data processing method for gas chromatography analysis according to claim 2, characterized in that: The constraint penalty factor is expressed as: Among them, epoth represents the current iteration number, and epoths represents the total iteration number.

5. The data processing method for gas chromatography analysis according to claim 1, characterized in that: The S2 comprises: S21, add noise to the sample signal x(t) to generate NR noise-enhanced signals x i (t): x i (t)=x(t)+Nstdw i (t) Among them, w i (t) represents Gaussian white noise, i = 1, 2, ... NR; S22, for each noise enhancement signal x i (t) Perform EMD decomposition to obtain the first layer intrinsic mode function c of each noise 1,i (t); S23, calculate the first layer intrinsic mode function c 1,i The average value of (t) S24, based on average Update residual r1(t): S25, add noise to the residual of the k-1th layer to obtain NR noise-enhanced signals r k-1,i (t): r k-1,i (t)=r k-1 (t)+Nstdw i (t) S26, for each noise enhancement signal r k-1,i (t) Perform EMD decomposition to obtain the k-th layer intrinsic mode function c of each noise k,i (t); S27, calculate the k-th layer intrinsic mode function c k,i The average value of (t) S28, based on average Update residual r k (t): S29, until the preset number of layers k is met, stop decomposing and output the intrinsic mode functions IMFs of the 1st to kth layers.

6. The data processing method for gas chromatography analysis according to claim 1, characterized in that: In S3, the comprehensive evaluation index is expressed as: Among them, τ represents the correlation coefficient, γ represents the energy ratio, and ω represents the reconstruction error.

7. A data processing method for gas chromatography analysis according to claim 1, characterized in that: In S4, identifying the peaks in the reconstructed signal by using a peak detection algorithm includes: S41, calculating the first-order derivative of the reconstructed signal, and identifying the data point where the sign of the derivative changes; S42, based on the data points, determining a data point at which a change in the derivative exceeds a preset slope threshold as an initial peak; S43, determining an initial wave peak exceeding a preset amplitude threshold as a wave peak.

8. The data processing method for gas chromatography analysis according to claim 1, characterized in that: In S4, the relevant parameters of the peaks include: the total number of peaks, the total area of ​​the peaks, the position of each peak, the height of each peak, the width of each peak, the area of ​​each peak and the proportion of the area of ​​each peak in the total area of ​​the peaks.

9. The data processing method for gas chromatography analysis according to claim 1, characterized in that: The S4 also includes calculating the area of ​​each peak using the tangent method: Wherein, H represents the half height of the peak value, and F represents the full width at half maximum.

10. A data processing system for gas chromatography analysis, characterized in that: include: Signal acquisition module: used to acquire sample signals from gas chromatograph; Signal decomposition module: used to decompose the sample signal into multiple intrinsic mode functions IMF using the optimized ICEEMDA algorithm; IMF selection and signal reconstruction module: used to construct comprehensive evaluation indicators to select the corresponding intrinsic mode function IMF for signal reconstruction and obtain the reconstructed signal after noise reduction; Peak detection and parameter extraction module: used to identify the peaks in the reconstructed signal through the peak detection algorithm, and extract the relevant parameters of the peaks using Gaussian fitting; Report generation module: used to generate corresponding analysis reports based on relevant parameters of the peak.

Images