Method for detecting aniline additives in vehicle gasoline based on near infrared spectrum

Abstract

The invention provides a near infrared spectrum-based method for detecting aniline additives in automotive gasoline, and relates to the technical field of chemical analysis and detection. The method comprises the following steps: firstly, acquiring a concentration prediction interval based on the aspects of gasoline source, refining process, quality standard and the like through big data analysis, secondly, enabling aniline additives in gasoline to be more easily detected through derivatization and extraction processes, and next, extracting key wavelength information according to a near infrared spectroscopic analysis result, the method comprises the following steps: acquiring a first prediction concentration interval of aniline additives in gasoline, further highlighting the significance of the aniline additives in the gasoline, acquiring a first aniline prediction concentration based on big data and a machine learning model, and finally, adjusting the first aniline prediction concentration based on the first prediction concentration interval to obtain a second aniline prediction concentration. According to the technical scheme, accurate detection of aniline additives in gasoline is realized by combining chemical treatment and artificial intelligence technical means, the method can be implemented only by near-infrared detection equipment, and the economical efficiency of the detection method is greatly improved.

Description

Technical Field

[0001] The invention belongs to the technical field of chemical analysis and detection, and in particular relates to a method for detecting aniline additives in motor gasoline based on near infrared spectroscopy. Background Art

[0002] Near-infrared spectroscopy is widely used in the detection of automotive gasoline additives due to its rapid and non-destructive characteristics. However, aniline compounds and hydrocarbons, alcohols and other components in the gasoline matrix are prone to overlap in characteristic peaks in the near-infrared spectrum, and the accuracy of low-concentration aniline compound detection is not high, resulting in insufficient detection stability and low detection accuracy. In the existing technology, it is usually necessary to combine near-infrared and mid-infrared spectroscopy for simultaneous detection, but this detection method requires the purchase of multiple sets of equipment, which is less economical. Summary of the invention

[0003] The purpose of the present invention is to provide a method for detecting aniline additives in automotive gasoline based on near infrared spectroscopy, so as to solve the technical problems of low stability and low accuracy in detecting aniline compounds in gasoline in the prior art.

[0004] The present invention proposes a method for detecting aniline additives in motor gasoline based on near infrared spectroscopy, the method comprising: S1: determining a first predicted concentration range according to first basic information of first gasoline sample data; wherein the first basic information is used to characterize the origin, production process and quality management standard information of the first gasoline sample data; S2: performing a first chemical process on the first gasoline sample data to obtain second gasoline sample data; S3: obtaining first key wavelength information based on the first near infrared spectrum information of the second gasoline sample data; S4: inputting the first key wavelength information into a first aniline concentration prediction model to obtain a first aniline predicted concentration; S5: Based on the first predicted concentration interval, adjusting the first predicted aniline concentration to obtain a second predicted aniline concentration.

[0005] Preferably, the S1 further comprises: S11: Determine first origin information from the first basic information, and determine a first origin label based on the first origin information; S12: Determine a first production and processing profile from the first basic information; S13: determining a first quality management standard according to the first origin information, and performing semantic analysis on the first quality management standard to obtain a first quality management level; S14: Inputting the first origin label, the first production and processing portrait, and the first quality management level into a first predicted concentration range determination model to determine the first predicted concentration range.

[0006] Preferably, the S12 further comprises: S121: Determine a first production and processing vector based on the first basic information; S122: Input the first production and processing vector into a production and processing label determination model to determine the first production and processing portrait.

[0007] Preferably, S2 further comprises the following steps: S21: performing a first chemical preprocessing step on the first gasoline sample data to obtain third gasoline sample data; S22: performing a first extraction and separation step on the third gasoline sample data to obtain second gasoline sample data.

[0008] Preferably, the first chemical pretreatment step is a derivatization reaction.

[0009] Preferably, the S3 further comprises the following steps: S31: Acquire first near infrared spectrum information corresponding to the second gasoline sample data; S32: performing a first preprocessing operation on the first near infrared spectrum information to obtain second near infrared spectrum information; S33: performing a first continuous projection processing operation on the second near-infrared spectrum information to obtain the first key wavelength information.

[0010] Preferably, the first preprocessing operation includes a standard normal variable transformation step and a second-order derivative composite processing step.

[0011] Preferably, the S33 further comprises the following steps: S331: randomly selecting a first initial wavelength point from the second near-infrared spectrum information, and acquiring a first initial wavelength corresponding to the first initial wavelength point; S332: Calculate a first projection vector between each first candidate wavelength and the first initial wavelength, determine the first candidate wavelength with the largest first projection vector as the first target wavelength, and add it to the first wavelength set; S333: If the number of the first target wavelengths in the first wavelength set is equal to a first value, output the first wavelength set as the first key wavelength information; otherwise, continue to execute S332.

[0012] Preferably, the S5 further comprises the following steps: S51: if the first predicted aniline concentration coincides with the first predicted concentration interval, the first predicted aniline concentration is used as the second predicted aniline concentration and the process proceeds to S54; otherwise, a first deviation value is calculated and the process proceeds to S52; S52: If the first deviation value is less than or equal to a first preset value, proceed to S53, otherwise output detection error information; S53: normalizing the first predicted aniline concentration within the first predicted concentration interval according to the first deviation value, and obtaining the second predicted aniline concentration, and proceeding to S54; S54: Output the second aniline predicted concentration.

[0013] Preferably, the first value is 15.

[0014] The present application proposes a method for detecting aniline additives in automotive gasoline based on near-infrared spectroscopy, which relates to the field of chemical analysis and detection technology. First, through big data analysis, the concentration prediction interval is obtained based on the gasoline source, refining process, quality standards, etc., and secondly, through derivatization and extraction processes, the aniline additives in gasoline are easier to detect. Next, the key wavelength information is extracted from the near-infrared spectral analysis results to further highlight the significance of aniline additives in gasoline, and the first aniline predicted concentration is obtained based on big data and machine learning models. Finally, the first aniline predicted concentration is adjusted based on the first predicted concentration interval to obtain the second aniline predicted concentration. The technical solution of the present invention combines chemical treatment and artificial intelligence technology to achieve accurate detection of aniline additives in gasoline, which can be implemented with only near-infrared detection equipment, greatly improving the economy of the detection method. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0016] Figure 1 It is an execution flow chart of the method for detecting aniline additives in motor gasoline based on near infrared spectroscopy in the present invention.

[0017] Figure 2 It is a flow chart for determining the first key wavelength information in the detection method of aniline additives in motor gasoline based on near infrared spectroscopy in the present invention.

[0018] Figure 3 It is a schematic diagram of determining the predicted concentration of the second aniline in the motor gasoline based on the near infrared spectrum in the present invention. DETAILED DESCRIPTION

[0019] 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.

[0020] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments, wherein the illustrative embodiments and descriptions are only used to explain the present invention but are not intended to limit the present invention.

[0021] The following is a detailed description of the method for detecting aniline additives in gasoline based on near infrared spectroscopy of the present invention. Figure 1 shown.

[0022] S1: Determine a first predicted concentration range according to first basic information of first gasoline sample data.

[0023] Aniline substances are usually not intentionally added to gasoline. Their content is mainly affected by the source and origin of gasoline, the production process and raw materials. In addition, it may also be affected by quality management standards. For example, the composition of crude oil from different origins is different. Some crude oils may contain trace amounts of aromatic amine precursors, which may be partially converted into aniline compounds during the refining process. For example, crude oil from certain specific oil regions may have a relatively high nitrogen content, which may contain some nitrogen-containing heterocyclic compounds. These compounds may undergo complex chemical reactions during the refining process to generate trace amounts of aniline substances. For example, if the quality management standards are high, aniline additives will usually be removed more thoroughly during the production and processing process.

[0024] Therefore, in this step, the maximum possible concentration range of the aniline additive is first determined through the first basic information of the first gasoline sample data.

[0025] The first basic information is used to characterize information such as the source, origin, production process and quality management standards of the first gasoline sample data.

[0026] The S1 may specifically include the following sub-steps: S11: Determine first origin information from the first basic information, and determine a first origin label based on the first origin information.

[0027] In order to ensure the safety of gasoline, the information on the extraction, production and processing of gasoline is usually recorded, and traceability information is provided. The first basic information can be obtained based on the traceability information.

[0028] Since gasoline produced from different regions has certain characteristics in composition, in this step, it is preferred to classify and label different regions in advance according to the characteristics of gasoline composition in different regions around the world. Then, the first origin label can be determined through the first origin information.

[0029] The first origin label may include information on the composition characteristics of gasoline, such as whether it contains aromatic amine precursors or nitrogen-containing heterocyclic compounds, and the concentration range of the above substances.

[0030] S12: Determine a first production and processing portrait from the first basic information.

[0031] In the production and processing of gasoline, the main considerations are the impact of refining technology, refining links and blending processes on the concentration of aniline substances in gasoline.

[0032] First, different refineries use different process routes and equipment. Some advanced refining processes can more effectively remove impurities, including possible aniline substances. For example, refineries that use deep hydrocracking and refining processes can better control the impurity content in gasoline and reduce the residual aniline substances. However, some small refineries or refineries with relatively backward processes may be less effective in removing impurities, resulting in a relatively high content of aniline substances in gasoline.

[0033] Secondly, the refining process is crucial to remove impurities. If the refining process is not perfect, it may cause aniline residues. For example, if the refining steps such as desulfurization and denitrification are not thorough enough, trace amounts of aniline may enter the final product.

[0034] Finally, if components containing aniline substances are used during the blending process of oil, or if the mixing ratio of the components is inappropriate, the content of aniline substances may also increase.

[0035] Therefore, the S12 may include the following sub-steps: S121: Determine a first production and processing vector based on the first basic information.

[0036] In this step, it is necessary to determine the oil refining process information, oil refining equipment information, refining process information and blending material information corresponding to the first gasoline sample.

[0037] The above four types of information are sorted into the first production and processing vector in a specified data format. The first production and processing vector is preferably a 1×4 dimensional vector.

[0038] S122: Input the first production and processing vector into a production and processing label determination model to determine the first production and processing portrait.

[0039] The production and processing label determination model is obtained by training a convolutional neural network model, wherein gasoline sample data from different regions are used as training data, and for each piece of sample data, the production and processing vector thereof is manually labeled with a production and processing label. For the labeled sample data, the production and processing vector is used as input data, and the manually labeled production and processing label is used as output data, and the production and processing label determination model is obtained by training.

[0040] The first production and processing portrait obtained contains relevant information for characterizing the production and processing stage of the first gasoline sample, for example, "deep hydrocracking, undesulfurized, and blending component content".

[0041] S13: Determine a first quality management standard according to the first origin information, and perform semantic analysis on the first quality management standard to obtain a first quality management level.

[0042] Generally speaking, the quality management standards for gasoline vary according to the place of origin, so based on the first place of origin information determined in step S11, the corresponding first quality relationship standard can be queried. Next, the first quality management standard can be analyzed by a semantic analysis model in the prior art to obtain the first quality management level corresponding to the first gasoline sample.

[0043] Preferably, the quality management level corresponding to the designated place of origin may also be searched through a pre-established mapping table.

[0044] Preferably, the first quality relationship level can be divided into 3 or 4 levels.

[0045] S14: Inputting the first origin label, the first production and processing portrait, and the first quality management level into a first predicted concentration range determination model to determine the first predicted concentration range.

[0046] Among them, the first predicted concentration interval determination model is obtained by training a machine learning model, and historical gasoline-related data from different regions are selected as sample data. For each sample data, its origin label, production and processing portrait, and quality management level are used as input data, and the predicted concentration interval that meets the confidence requirements is used as output data to train the first predicted concentration interval determination model.

[0047] Among them, the predicted concentration range that meets the confidence requirement is determined in the following way: first, the measured aniline additive concentration value corresponding to each sample data is obtained, and the measured aniline additive concentration value is expanded according to the confidence level of 80%, so as to obtain the predicted concentration range corresponding to the sample data.

[0048] S2: Performing a first chemical process on the first gasoline sample data to obtain second gasoline sample data.

[0049] In near-infrared spectral analysis, the concentration detection of aniline additives in gasoline is often interfered by the overlapping of characteristic peaks, resulting in reduced detection sensitivity and accuracy. To solve the above problems, this step, before analyzing the near-infrared spectrum, first uses derivatization reaction and extraction separation technology to significantly enhance the characteristic absorption or fluorescence signal of aniline compounds, and effectively separates the interfering components, thereby preparing for the subsequent near-infrared spectral analysis.

[0050] Wherein, the first chemical treatment includes two steps, namely a first chemical pretreatment step and a first extraction separation step.

[0051] The S2 comprises the following sub-steps: S21: performing a first chemical preprocessing step on the first gasoline sample data to obtain third gasoline sample data.

[0052] The first chemical pretreatment step adopts a derivatization reaction, specifically, converting the aniline compound into a derivative with a stronger characteristic absorption or fluorescence signal to enhance the detection signal.

[0053] The specific implementation steps are as follows: Reagent selection: One or more derivatization reagents, such as aldehydes, ketones, anhydrides or acylating agents, are selected to react with aniline compounds to generate derivatives with specific spectral characteristics. The selection of reagents should be based on their reactivity with aniline, the stability of the derivatives, and their suitability for near-infrared spectroscopy detection.

[0054] Optimization of reaction conditions: By controlling the reaction temperature (e.g. 40-60°C), reaction time (e.g. 10-30 minutes) and pH value (e.g. neutral or weakly acidic), the derivatization reaction can be carried out efficiently while avoiding the occurrence of side reactions.

[0055] Derivative Validation: The formation of derivatives is verified by high performance liquid chromatography (HPLC) or mass spectrometry (MS) techniques, and the reaction conditions are optimized to ensure the purity and stability of the derivatives.

[0056] S22: performing a first extraction and separation step on the third gasoline sample data to obtain second gasoline sample data.

[0057] The first extraction and separation step uses solid phase extraction (SPE) or liquid-liquid extraction (LLE) technology to separate aniline derivatives from other components to reduce interference and improve detection accuracy.

[0058] The specific implementation steps of solid phase extraction (SPE) are as follows: Selection of extraction materials: Use solid phase extraction columns with specific adsorption characteristics (such as C18, C8 or polar adsorbents) to selectively adsorb aniline derivatives.

[0059] Operation process: Pass the sample through the extraction column and elute with an appropriate eluent (such as methanol, acetonitrile or water) to separate the target derivatives from other interfering components.

[0060] Parameter optimization: Optimize the extraction efficiency by adjusting the flow rate (e.g., 1-5 mL / min), eluent concentration, and volume.

[0061] The specific implementation steps of liquid-liquid extraction (LLE) are as follows: Solvent selection: Choose an organic solvent that is immiscible with the sample matrix (such as hexane, ethyl acetate, or toluene) to achieve selective extraction of aniline derivatives.

[0062] Operational procedure: The sample is mixed with the extraction solvent in a certain ratio (e.g. 1:1 to 1:5), the extraction is promoted by shaking or stirring, and then the organic phase and the aqueous phase are separated.

[0063] Parameter optimization: Optimize the extraction efficiency by adjusting the type, ratio and extraction time of the extraction solvent (e.g. 5-15 minutes).

[0064] Through step S2, the identifiability of aniline compounds in the second gasoline sample data can be significantly enhanced, thereby preparing for the subsequent spectral analysis step.

[0065] S3: Obtaining first key wavelength information based on the first near-infrared spectrum information of the second gasoline sample data.

[0066] Through step S2, the aniline additives in the gasoline sample have been enhanced to a certain extent. In order to further improve the detection accuracy, in this step, the first key wavelength information is extracted from the first near-infrared spectrum information corresponding to the second gasoline sample, and the irrelevant wavelength information in the first near-infrared spectrum information is eliminated, and the first key wavelength information with a higher correlation with the recognition accuracy is retained.

[0067] The S3 includes the following sub-steps: Figure 2 As shown: S31: Acquire first near-infrared spectrum information corresponding to the second gasoline sample data.

[0068] The first near-infrared spectrum information is obtained by a near-infrared spectrum analyzer.

[0069] S32: performing a first preprocessing operation on the first near-infrared spectrum information to obtain second near-infrared spectrum information.

[0070] The first preprocessing operation includes a standard normal variable transformation (SNV) step and a second-order derivative composite processing step.

[0071] The standard normal variable transformation (SNV) step specifically includes: The original near-infrared spectrum matrix Xm×n (m is the number of samples, n is the number of wavelength points) is processed by SNV to eliminate light scattering interference. The value of each position in the matrix after processing is calculated by the following formula: ; In the formula, is the transformation value of the i-th sample at the j-th wavelength point, μ i is the mean of the ith sample spectrum, σ i is the standard deviation, x i,j is the original absorbance value. <i≤m,0<j≤n。

[0072] The second-order derivative composite processing step specifically includes: The Savitzky-Golay filter was used to calculate the second-order derivative of the spectrum after SNV processing, and the parameters were set as follows: Window width: 11 wavelength points (dynamically adjusted according to spectral resolution, ranging from 7 to 15 points); Polynomial order: 3rd order; Derivative order: 2nd order; Eliminate baseline drift and sharpen characteristic peaks through the second-order derivative, and output the preprocessed spectrum matrix X preprocessed .

[0073] S33: performing a first continuous projection processing operation on the second near-infrared spectrum information to obtain the first key wavelength information.

[0074] The first key wavelength information can be obtained by performing continuous projection processing on the second near-infrared spectrum information obtained after preprocessing.

[0075] To avoid overfitting and information loss, the wavelength screening target interval can be set to a specified number. Preferably, the number of wavelength points to be acquired is 15.

[0076] The S33 comprises the following sub-steps: S331: randomly selecting a first initial wavelength point from the second near-infrared spectrum information, and acquiring a first initial wavelength corresponding to the first initial wavelength point.

[0077] The first initial wavelength point is randomly acquired from the second near-infrared spectrum information.

[0078] S332: Calculate a first projection vector between each first candidate wavelength and the first initial wavelength, determine the first candidate wavelength with the largest first projection vector as the first target wavelength, and add it to the first wavelength set.

[0079] In this step, multiple first candidate wavelengths can be randomly determined from the second near-infrared spectrum information, and the first projection vector of each first candidate wavelength and the first initial wavelength is calculated one by one, and the first candidate wavelength with the largest first projection vector is added to the first wavelength set. By repeatedly performing this step, the first wavelength set consisting of a specified number of first target wavelengths can be obtained.

[0080] S333: If the number of the first target wavelengths in the first wavelength set is equal to a first value, output the first wavelength set as the first key wavelength information; otherwise, continue to execute S332.

[0081] The goal of this step is to output a first wavelength set that meets the first numerical requirement and use the first wavelength information as the first key wavelength information. Therefore, when the numerical requirement is not met, step S332 needs to be executed repeatedly.

[0082] Preferably, the first value is 15.

[0083] S4: inputting the first key wavelength information into a first aniline concentration prediction model to obtain a first aniline predicted concentration.

[0084] In S2 and S3, the first key wavelength information for characterizing the content of aniline additives in the first gasoline sample has been determined through chemical treatment and key wavelength extraction, and different wavelength information is correlated with the concentration of aniline additives in gasoline.

[0085] Therefore, in this step, the first key wavelength information obtained in S3 is input into the first aniline concentration prediction model to obtain the first aniline predicted concentration.

[0086] The first aniline concentration prediction model can be obtained by training using a machine learning model known in the prior art, such as a convolutional neural network model. During training, historical gasoline data is used as sample data, that is, for each sample data, its key wavelength information is used as input data, and the measured aniline concentration information is used as output data to train and obtain the first aniline concentration prediction model.

[0087] S5: Based on the first predicted concentration interval, adjusting the first predicted aniline concentration to obtain a second predicted aniline concentration.

[0088] In this step, based on the numerical relationship between the first predicted concentration interval and the first predicted aniline concentration, the second predicted aniline concentration corresponding to the first gasoline sample that is finally output is determined.

[0089] The S5 comprises the following sub-steps: S51: If the first predicted aniline concentration overlaps with the first predicted concentration interval, the first predicted aniline concentration is used as the second predicted aniline concentration and the process proceeds to S54; otherwise, the first deviation value is calculated and the process proceeds to S52.

[0090] If the first predicted aniline concentration coincides with the first predicted concentration interval, it proves that the prediction results in S1 and S4 have a certain consistency, and the first predicted aniline concentration can be directly determined as the second predicted aniline concentration and output.

[0091] If the first predicted aniline concentration does not overlap with the first predicted concentration interval, it proves that there is a certain deviation between the prediction results in S1 and S4, and therefore the final result needs to be adjusted according to the specific situation.

[0092] The first deviation value refers to the numerical deviation between the first predicted aniline concentration and the nearest endpoint of the first predicted concentration interval. For example, if the first predicted concentration interval is 20-40 and the first predicted aniline concentration is 42, then the first deviation value is (42-40)=2.

[0093] S52: If the first deviation value is less than or equal to a first preset value, proceed to S53; otherwise, output detection error information.

[0094] In this step, if the first deviation value is less than or equal to the preset value, the first predicted aniline concentration needs to be adjusted according to the first predicted concentration interval. If the first deviation value is too large, it means that the detection error is too large, and the detection error information is output.

[0095] S53: normalizing the first predicted aniline concentration within the first predicted concentration range according to the first deviation value, obtaining the second predicted aniline concentration, and entering S54.

[0096] In this step, the first predicted aniline concentration and the first predicted concentration range need to be integrated to obtain the second predicted aniline concentration.

[0097] For example, Figure 3 As shown, if the first predicted concentration interval is 20-40, the first aniline predicted concentration is 42, and the first preset value is 10, then the midpoint value of the first predicted concentration interval is 30. Since the first aniline predicted concentration 42 is greater than the upper endpoint value 40 of the first predicted concentration interval, the first aniline predicted concentration can be normalized to the numerical range (30, 40) from the midpoint to the upper endpoint of the first predicted concentration interval according to the first deviation value (42-40)=2, and the second aniline predicted concentration can be determined to be 38.

[0098] S54: Output the second aniline predicted concentration.

[0099] The second predicted aniline concentration is the concentration value of the aniline additive corresponding to the first gasoline sample.

[0100] The present application proposes a method for detecting aniline additives in automotive gasoline based on near-infrared spectroscopy, which relates to the field of chemical analysis and detection technology. First, through big data analysis, the concentration prediction interval is obtained based on the gasoline source, refining process, quality standards, etc., and secondly, through derivatization and extraction processes, the aniline additives in gasoline are easier to detect. Next, the key wavelength information is extracted from the near-infrared spectral analysis results to further highlight the significance of aniline additives in gasoline, and the first aniline predicted concentration is obtained based on big data and machine learning models. Finally, the first aniline predicted concentration is adjusted based on the first predicted concentration interval to obtain the second aniline predicted concentration. The technical solution of the present invention combines chemical treatment and artificial intelligence technology to achieve accurate detection of aniline additives in gasoline, which can be implemented with only near-infrared detection equipment, greatly improving the economy of the detection method.

[0101] The above description is only a preferred embodiment of the present invention, so all equivalent changes or modifications made according to the structure, characteristics and principles described in the scope of the patent application of the present invention are included in the scope of the patent application of the present invention.

Claims

1. A method for detecting aniline additives in motor gasoline based on near infrared spectroscopy, characterized in that: The method includes: S1: determining a first predicted concentration range according to first basic information of first gasoline sample data; wherein the first basic information is used to characterize the origin, production process and quality management standard information of the first gasoline sample data; S2: performing a first chemical process on the first gasoline sample data to obtain second gasoline sample data; S3: obtaining first key wavelength information based on the first near infrared spectrum information of the second gasoline sample data; S4: inputting the first key wavelength information into a first aniline concentration prediction model to obtain a first aniline predicted concentration; S5: Based on the first predicted concentration interval, adjusting the first predicted aniline concentration to obtain a second predicted aniline concentration.

2. The method for detecting aniline additives in motor gasoline based on near infrared spectroscopy according to claim 1, characterized in that: Said S1 further comprises: S11: Determine first origin information from the first basic information, and determine a first origin label based on the first origin information; S12: Determine a first production and processing profile from the first basic information; S13: determining a first quality management standard according to the first origin information, and performing semantic analysis on the first quality management standard to obtain a first quality management level; S14: Inputting the first origin label, the first production and processing portrait, and the first quality management level into a first predicted concentration range determination model to determine the first predicted concentration range.

3. The method for detecting aniline additives in motor gasoline based on near infrared spectroscopy according to claim 2, characterized in that: The S12 further comprises: S121: Determine a first production and processing vector based on the first basic information; S122: Input the first production and processing vector into a production and processing label determination model to determine the first production and processing portrait.

4. The method for detecting aniline additives in motor gasoline based on near infrared spectroscopy according to claim 3, characterized in that: The S2 further comprises the following steps: S21: performing a first chemical preprocessing step on the first gasoline sample data to obtain third gasoline sample data; S22: performing a first extraction and separation step on the third gasoline sample data to obtain second gasoline sample data.

5. The method for detecting aniline additives in motor gasoline based on near infrared spectroscopy according to claim 4, characterized in that: The first chemical pretreatment step is a derivatization reaction.

6. The method for detecting aniline additives in motor gasoline based on near infrared spectroscopy according to claim 5, characterized in that: The S3 further comprises the following steps: S31: Acquire first near infrared spectrum information corresponding to the second gasoline sample data; S32: performing a first preprocessing operation on the first near infrared spectrum information to obtain second near infrared spectrum information; S33: performing a first continuous projection processing operation on the second near-infrared spectrum information to obtain the first key wavelength information.

7. The method for detecting aniline additives in motor gasoline based on near infrared spectroscopy according to claim 6, characterized in that: The first preprocessing operation includes a standard normal variable transformation step and a second-order derivative complex processing step.

8. The method for detecting aniline additives in motor gasoline based on near infrared spectroscopy according to claim 7, characterized in that: The S33 further comprises the following steps: S331: randomly selecting a first initial wavelength point from the second near-infrared spectrum information, and acquiring a first initial wavelength corresponding to the first initial wavelength point; S332: Calculate a first projection vector between each first candidate wavelength and the first initial wavelength, determine the first candidate wavelength with the largest first projection vector as the first target wavelength, and add it to the first wavelength set; S333: If the number of the first target wavelengths in the first wavelength set is equal to a first value, output the first wavelength set as the first key wavelength information; otherwise, continue to execute S332.

9. The method for detecting aniline additives in motor gasoline based on near infrared spectroscopy according to claim 8, characterized in that: The S5 further comprises the following steps: S51: if the first predicted aniline concentration coincides with the first predicted concentration interval, the first predicted aniline concentration is used as the second predicted aniline concentration and the process proceeds to S54; otherwise, a first deviation value is calculated and the process proceeds to S52; S52: If the first deviation value is less than or equal to a first preset value, proceed to S53, otherwise output detection error information; S53: normalizing the first predicted aniline concentration within the first predicted concentration interval according to the first deviation value, and obtaining the second predicted aniline concentration, and proceeding to S54; S54: Output the second aniline predicted concentration.

10. The method for detecting aniline additives in motor gasoline based on near infrared spectroscopy according to claim 9, characterized in that: The first value is 15.

Images