CN105319198A - Gasoline benzene content prediction method based on Raman spectrum analysis technology - Google Patents
Gasoline benzene content prediction method based on Raman spectrum analysis technology Download PDFInfo
- Publication number
- CN105319198A CN105319198A CN201410336402.9A CN201410336402A CN105319198A CN 105319198 A CN105319198 A CN 105319198A CN 201410336402 A CN201410336402 A CN 201410336402A CN 105319198 A CN105319198 A CN 105319198A
- Authority
- CN
- China
- Prior art keywords
- benzene
- training sample
- spectral coverage
- raman spectrum
- gasoline
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
The invention discloses a gasoline benzene content prediction method based on a Raman spectrum analysis technology. According to the method, the training samples are collected at first, then the Raman spectrums of the training samples are measured, then the measured Raman spectrums are preprocessed, and the characteristic spectrum segments of benzene are selected. In this area, the interference peaks generated by other components in gasoline will overlap on the characteristic peak of benzene, and thus the model prediction precision is reduced. So a spectral analysis technology is adopted to fit the interference peaks so as to deduct the interference peaks. Finally, a multiple regression method is used to establish a regression model of the benzene characteristic spectrum segment, from which the interference peaks have been removed, and the volume content of benzene. The disclosed method utilizes a spectral analysis technology to effectively solve the problem that in a Raman spectrum, the signals of different components overlap with each other, so the precision of analysis model is prominently improved. Moreover, the provided method has the advantages that the required training sample is few, the extrapolation of the model is strong, the analysis time is short, and the prediction result is precise; and thus the method is suitable for the online quantitative analysis of gasoline benzene content.
Description
Technical field
The present invention relates to component chemical product quantitative test field, particularly utilize Raman spectrum to carry out accurate quantitative analyses to the benzene composition in gasoline.
Background technology
Benzene is considered to a kind of carcinogen, it is present in the chemical products such as gasoline, along with evaporation and/or the rough burning of the chemical products such as gasoline, benzene can cause severe contamination to air, therefore carries out to the benzene content in gasoline the important component part that strict control is the tail gas pollution of motor-driven vehicle comprehensive regulation.China starts from Dec 31st, 2009 quality index (GB17930-2006) implementing in full motor petrol state III, benzene content wherein in strict regulations gasoline must not higher than 1% volume, and the benzene content that regulation adopts vapor-phase chromatography to carry out in gasoline in industry standard SH/T0713-2002 is simultaneously measured.
The benzene content that application vapor-phase chromatography detects in gasoline mainly comprises following step:
(1) spectrogram being equipped with each composition of twin columns gas phase look systematic survey of transfer valve and flame ionization detector is adopted: the sample to be tested by certain volume and containing quantitative internal standard compound (as butanone) injects the pre-column that contains polar stationary phase; After C9 and the non-aromatics composition lighter than C9 flow out from pre-column, through drain emptying; Before benzene flows out, pre-column is placed in blowback state, retained fraction also imports the analytical column (WCOT) containing non-polar stationary phase; Benzene, toluene and internal standard compound flow out chromatographic column according to boiling point order and detect with flame ionization detector;
(2) peak area of benzene, toluene and internal standard compound is recorded;
(3) due to the response of flame ionization detector and the concentration linear proportional relation of each component, so can with reference to the content of benzene composition in internal standard compound calculation sample.
Although vapor-phase chromatography has higher analysis precision, the method has that analysis time is long, experimentation cost is high and the defect such as sample pretreatment program is loaded down with trivial details, and is not suitable for on-line measurement.Further, vapor-phase chromatography only can be used for the benzene content detecting regular gasoline and contain in ethers oxygenated compound gasoline, if containing alcohol constituents in gasoline, then measurement result can be interfered.
And based on Raman spectroscopy detection method because have sample preparation program simple, the advantages such as nondestructive real-time detection and simple installation can be carried out to sample, get the attention and successful Application at oil product component quantifying analysis field.But owing to comprising Various Complex potpourri in gasoline, and benzene is compared with other compositions, and its content is usually lower, the characteristic peak corresponding to benzene easily the spectrum peak that produces by other compositions cover, cause the precise decreasing of analytical model.
For addressing this problem, in prior art, usually improved the precision of prediction of analytical model by the method setting up Quantifying model on a large amount of basis of collecting training sample.But the collection work due to training sample needs a large amount of time and economic expense, and the extrapolation of its regression model set up also is difficult to ensure, once the composition of measurement environment and measuring object changes, probably need again to collect sample and carry out Model Reconstruction, these all add difficulty to practical application.
Summary of the invention
The object of the present invention is to provide a kind of benzene content in gasoline method for quick predicting based on Raman spectrum analytic technique with high precision and good extrapolation.
For achieving the above object, the invention provides a kind of benzene content in gasoline Forecasting Methodology based on Raman spectrum analytic technique, comprising: step S1, collect training sample, the benzene volume content of described training sample is known; Step S2, measures the Raman spectrum of described training sample; Step S3, carries out pre-service to measuring the described Raman spectrum obtained; Step S4, intercepting spectral coverage is obtained from Raman spectrum after pretreatment, this intercepting spectral coverage comprises the wave number at the characteristic peak place of benzene, and applies the described intercepting spectral coverage of each described training sample of Raman spectrum analytic technique decomposition, obtains characteristic peak and the Interference Peaks of the benzene of this training sample; Step S5, deducts respective described Interference Peaks, obtains the feature spectral coverage of the benzene of each sample in described training sample from the described intercepting spectral coverage of each sample described training sample; Step S6, utilizes the feature spectral coverage of the described benzene of described training sample to set up the regression model between the feature spectral coverage of described benzene and benzene volume content; And step S7, process from described step S2 to described step S5 is carried out to sample to be tested, obtain the feature spectral coverage of the described benzene after the described Interference Peaks of deduction of described sample to be tested, using the input of the described regression model that this feature spectral coverage is set up in described step S6, obtain the predicted value of the volume content of the described benzene in described sample to be tested.
Preferably, the benzene volume content of described training sample can be measured by standard analytical process.
Preferably, the described Raman spectrum obtained measurement carries out pre-service and can comprise smothing filtering, baseline correction and stable hydrocarbon normalization.
Preferably, described Raman spectrum analytic technique can be applied in described step S4 and decompose the described intercepting spectral coverage of each described training sample and comprise described intercepting spectral coverage is decomposed into by the stacking pattern at two Voigt peaks, that is:
Wherein, V
i(v, φ
i) be Voigt function, v represents wave number, and wave number is at 996cm
-1neighbouring spectrum peak is considered to the characteristic peak of benzene, is designated as V
b(v, φ
b), and another spectrum peak is considered to Interference Peaks, is designated as V
o(v, φ
o), r (v) represents regression criterion.
Wherein establish A
bv () represents the feature spectral coverage of the benzene after the described Interference Peaks of deduction, then have:
A
B(v)＝A(v)-V
O(v,φ
O)。
Preferably, can be any one in arithmetic of linearity regression, partial least square method, principal component regression algorithm and algorithm of support vector machine for setting up the algorithm of regression model in step S6.
Preferably, if adopt partial least square method to set up described regression model, then leaving-one method can be adopted to determine required main cause subnumber, and cross validation mean square deviation can be adopted as the Performance Evaluating Indexes of described regression model.
Method provided by the present invention, not only apply statistical regression and pay close attention to the physical significance of Raman spectrum, the process of removing Interference Peaks is carried out to studied Raman spectrum, make characteristic spectrum intensity and the benzene volume content (%v/v of obtained benzene, namely the number percent of volume shared by the benzene in unit volume gasoline) between the linear relation of relation character, the sample size needed in the process of establishing of regression model is less, and according to theory deduction (there is linear relationship between the raman spectrum strength of certain composition and its concentration) with test and confirm that this regression model has good extrapolation and degree of precision.
Accompanying drawing explanation
Fig. 1 gives the process flow diagram of the benzene content in gasoline Forecasting Methodology based on Raman spectrum analytic technique according to an embodiment of the invention;
Fig. 2 gives the original Raman spectrum spectrogram of 22 training samples adopting Raman spectrometer measurement to obtain;
Fig. 3 gives the Raman spectrum shown in Fig. 2 spectrum spectrogram after pretreatment;
Fig. 4 gives the Raman spectrum shown in Fig. 3 by the wave-number range after intercepting at 980 ~ 1020 (unit cm
-1) Raman spectrum spectrogram;
The spectral resolution result that Fig. 5 obtains after giving and resolving the spectrum of in the Raman spectrum shown in Fig. 4;
Fig. 6 gives the feature spectral coverage of the benzene obtained after all Raman spectrums shown in Fig. 4 remove respective Interference Peaks;
Fig. 7 gives the performance of the regression model obtained under different main cause subnumber when applying partial least square method (PLS) modeling; And
Fig. 8 gives the comparison diagram between predicted value and standard analysis value obtained according to set up regression model.
Embodiment
The invention provides a kind of benzene content in gasoline Forecasting Methodology based on Raman spectrum analytic technique, comprising: step S1, collect training sample, the benzene volume content of described training sample is known; Step S2, measures the Raman spectrum of described training sample; Step S3, carries out pre-service to measuring the described Raman spectrum obtained; Step S4, intercepting spectral coverage is obtained from Raman spectrum after pretreatment, this intercepting spectral coverage comprises the wave number at the characteristic peak place of benzene, and applies the described intercepting spectral coverage of each described training sample of Raman spectrum analytic technique decomposition, obtains characteristic peak and the Interference Peaks of the benzene of this training sample; Step S5, deducts respective described Interference Peaks, obtains the feature spectral coverage of the benzene of each sample in described training sample from the described intercepting spectral coverage of each sample described training sample; Step S6, utilizes the feature spectral coverage of the described benzene of described training sample to set up the regression model between the feature spectral coverage of described benzene and benzene volume content; And step S7, process from described step S2 to described step S5 is carried out to sample to be tested, obtain the feature spectral coverage of the described benzene after the described Interference Peaks of deduction of described sample to be tested, using the input of the described regression model that the sample value of this feature spectral coverage is set up in described step S6, obtain the predicted value of the volume content of the described benzene in described sample to be tested.
Fig. 1 shows the process flow diagram of the benzene content in gasoline Forecasting Methodology based on Raman spectrum analytic technique according to an embodiment of the invention.
Step S1, collects training sample.The training sample of gasoline sample as modeling measuring the volume content of its benzene through the standard analytical process vapor-phase chromatography of regulation (namely in SH/T0713-2002 " motor petrol and aviation gasoline in Benzene and Toluene content determination (vapor-phase chromatography) ") can be selected.In present embodiment, from the different batches gasoline of certain oil refining field domestic, choose 22 gasoline samples respectively, the benzene volume content in 22 selected gasoline samples is comparatively evenly distributed in the scope of 0.5% ~ 1.0%.Step S2, can measure the Raman spectrum of each training sample.In present embodiment, OceanOpticsQE65000 grating spectrograph can be adopted as measuring Raman spectrometer, its optical resolution is 6cm
-1; Centre wavelength can be adopted to be that the laser instrument of 785nm is as excitation source; Sensing range is set to 0 ~ 2100cm
-1.Carry out three continuous coverages for each sample, each integral time can be set to 10 seconds, using the Raman spectrum of the arithmetic mean of three measurement results as this sample.The original Raman spectrum of 22 samples as shown in Figure 2.
Step S3, can carry out pre-service to each Raman spectrum recorded.According to the difference of measuring condition, the pre-treatment step adopted also is not quite similar.In present embodiment, in order to eliminate the impact of the disturbing factors such as measurement noises, background fluorescence and excitation source Strength Changes as far as possible, the pre-service that Raman spectrum carries out is comprised:
(1) smothing filtering: specifically can adopt Savitzky-Golay convolution smooth filtering method, filter window width can be set to 5cm
-1;
(2) baseline correction: can adopt iteration polynomials baseline correction method, the degree of polynomial is set as 1, by carrying out iteration optimization to the parameter of linear baseline, obtains matching baseline; Then from the Raman spectrum after smoothing processing, matching baseline is deducted, to eliminate fluorescence background interference;
(3) stable hydrocarbon normalization: with the maximum Raman peaks (1448cm in stable hydrocarbon Raman spectrum in present embodiment
-1near) as a reference pointly spectrum after baseline calibration to be normalized, to eliminate the impact of excitation source strength fluctuation on Raman spectrum.
The Raman spectrum spectrogram after pretreatment of whole 22 training samples can refer to shown in Fig. 3.
Step S4, intercepting spectral coverage is obtained from Raman spectrum after pretreatment, this intercepting spectral coverage comprises the wave number at the characteristic peak place of benzene, and applies the intercepting spectral coverage that Raman spectrum analytic technique decomposes each training sample, obtains characteristic peak and the Interference Peaks of the benzene of this training sample.Because the characteristic peak of known benzene is at wave number 996cm
-1near, the intercepting scope selected in present embodiment is that wave number is from 980cm
-1to 1020cm
-1.The factor such as computation complexity and performance requirement can be considered, intercepting scope is adjusted.Fig. 4 is the spectrogram of the intercepting spectral coverage of the Raman spectrum of 22 training samples.Fig. 5 is the Raman spectrum analysis result of the intercepting spectral coverage of one of them sample.The Raman spectrum analytic technique adopted in present embodiment comprises:
(1) analytic model defining Raman spectrum is:
Wherein v represents wave number, and L represents the number at the independence spectrum peak that spectral signal comprises; V
i(v, φ
i) be Voigt function, it is defined as follows:
[α, ω, γ, θ] be the parameter phi that can be used for representing Voigt Profile, wherein the implication of each element is: α (peak height), ω (center at peak), γ (half-breadth at peak) and θ (Gauss-Lorentz coefficient); And r (v) represents regression criterion;
(2) characteristic peak of known benzene is at wave number 996cm
-1near, the Raman spectrum of each training sample near this wave number is made up of two independent spectrum peaks, according to spectrum resolution model, is expressed as the stacking pattern at two Voigt peaks, namely
Nonlinear Least-Square Algorithm can be utilized to carry out optimizing, to obtain the parameter [φ of two independent spectral functions by the parameter of gradient method to two independent spectrum peaks
1 t, φ
2 t]
t.Optimization aim in present embodiment minimizes for making regression criterion signal, namely
wherein Φ=[φ
1 t, φ
2 t]
t.
According to priori, resolve in two the independent spectrum peaks obtained, wave number is at 996cm
-1neighbouring spectrum peak is the characteristic peak of benzene, and the peak on the left of namely in Fig. 5, can be designated as V
b(v, φ
b); Another spectrum peak composing right side in peak and Fig. 5 is considered to Interference Peaks, can be designated as V
o(v, φ
o).
Step S5, deducts respective Interference Peaks from the intercepting spectral coverage of each sample, obtains the feature spectral coverage of the benzene of each sample.Interference Peaks is deducted, if A from intercepting spectral coverage
bv () represents the feature spectral coverage of the benzene after deduction Interference Peaks, then have:
A
B(v)＝A(v)-V
O(v,φ
O)。
The spectrogram of the feature spectral coverage of the benzene that Fig. 6 obtains after giving the deduction Interference Peaks of 22 samples.
Step S6, sets up the regression model between the feature spectral coverage of benzene and benzene volume content.Using the input of the sample value on the feature spectral coverage of benzene as this regression model in present embodiment, using the volume content of benzene as the predicted value of this regression model.Particularly, 1cm can be spaced apart with wave number
-1step-length choose sample point from the feature spectral coverage of benzene, can at 980cm corresponding to each sample
-1to 1020cm
-1scope in select 41 sample points, the intensity of the Raman spectrum after the standardization (this standardization is completed by the pre-treatment step of step S3) corresponding to each sample point is the sample value of this point.Arithmetic of linearity regression, partial least square method (PartialLeastSquares can be adopted, be called for short PLS), the method such as principal component regression algorithm or algorithm of support vector machine sets up regression model, these methods are all methods conventional in regression modeling.Preferably, partial least square method modeling is adopted in present embodiment, PLS is a kind of multivariate statistics data analysing method be widely used in recent years, and the method integrates the advantage of these three kinds of analytical approachs of principal component analysis (PCA), canonical correlation analysis and multiple linear regression analysis.When application PLS sets up regression model, first to determine main cause subnumber required in modeling process.This main cause subnumber is determined with leaving-one method in present embodiment, with cross validation mean square deviation (SEVC, Standarderrorofcross-validation) as the Performance Evaluating Indexes of regression model, namely set main cause subnumber as m, it is that the 1st sample is to the 22nd sample that all 22 samples are sorted, each with 21 samples of all the other except i-th sample for known sample application partial least square method sets up a regression model
for the model predication value obtained after predicting i-th sample with this model, and y
ibe the standard analysis value (the benzene volume content namely obtained in step sl) of i-th sample, this process repeats 22 times, chooses different sample i at every turn, then when main cause subnumber is m, the cross validation mean square deviation of this regression model is:
Wherein N is training sample sum, gets 22 in present embodiment.
Fig. 7 shows the model performance (SECV index) when applying partial least square method modeling corresponding to different main cause subnumber.As can be seen from Figure 7, when main cause subnumber is 3, SECV=3.8380 × 10
-3, this regression model has reached higher precision, therefore, in present embodiment, the value of main cause subnumber can be set as 3.And then using whole 22 training samples as known sample, the regression model between the feature spectral coverage sample point of adopt partial least square method modeling to obtain volume content that main cause subnumber is the benzene of 3 and benzene.Because partial least square method is the very ripe method be widely used, so repeat no more the detailed step of PLS modeling, can see " partial least-square regression method and application thereof ", author Wang Huiwen.
Step S7, process as above-mentioned steps S2 to step S5 is carried out to sample to be tested, obtain the feature spectral coverage of the benzene after deducting Interference Peaks, using the input of the model that the sample value of this feature spectral coverage is set up in step S6, calculate the volume content predicted value of the benzene in sample to be tested through model.
Except above-mentioned 22 gasoline samples, inventor also have chosen the test samples of 6 gasoline samples as above-mentioned model.The benzene volume content of three in these 6 samples is in the scope of 0.5% ~ 1%, and the benzene volume content of other three is respectively 0.3%, 1.5% and 1.8%.Fig. 8 shows the standard analysis value of these 6 test samples and the comparison diagram of model predication value, can find out that the predicated error of the method is very little intuitively.Model evaluation parameter can be set: the upper deviation, lower deviation, square error (SEP, StandartErrorofPrediction) and multiple correlation coefficient (R
2), each parameter is defined as follows:
Wherein
be the benzene volume content predicted value that i-th test samples obtains after step 7, and y
ibe the standard analysis value of i-th test samples,
for the arithmetic mean of standard analysis value, M is the number of test samples, i=1 ..., 6.Be 0.0309% to obtaining the upper deviation after above-mentioned 6 test samples analyses, lower deviation is-0.0186%, SEP=0.0094%, R
2=0.9882.Can find out, the inventive method within training sample set coverage and outside unknown sample all have higher precision of prediction, the extrapolation of model is stronger.
In order to further illustrate the stability of the inventive method, from above-mentioned 28 gasoline samples altogether, randomly drawing 6 as test samples, remaining 22 as training sample.Above-mentioned experiment is repeated 10 times, and the property indices of experimental result is as shown in table 1, and the multiple correlation coefficient of standard analysis value and model prediction is all more than 0.98, and the average of multiple correlation coefficient reaches 0.9880, and maximum square error is 0.0168%.
Table 1. Performance Evaluation table
In the present invention, the name of various element is only the object in order to example, has identical or similar characteristics and function element and may have different titles in reality.
Although disclose embodiments of the present invention for purpose of explanation, it should be understood that as the present invention is not limited thereto, and it will be understood by those skilled in the art that various do not depart from the scope and spirit of the present invention amendment, increase and to substitute be possible.Correspondingly, any and all modifications, change or equivalent arrangement should take into account scope of the present invention, and detailed scope of the present invention is open by the claims by enclosing.
Claims (6)
1., based on a benzene content in gasoline Forecasting Methodology for Raman spectrum analytic technique, comprising:
Step S1, collect training sample, the benzene volume content of described training sample is known;
Step S2, measures the Raman spectrum of described training sample;
Step S3, carries out pre-service to measuring the described Raman spectrum obtained;
Step S4, intercepting spectral coverage is obtained from Raman spectrum after pretreatment, this intercepting spectral coverage comprises the wave number at the characteristic peak place of benzene, and applies the described intercepting spectral coverage of each described training sample of Raman spectrum analytic technique decomposition, obtains characteristic peak and the Interference Peaks of the benzene of this training sample;
Step S5, deducts respective described Interference Peaks, obtains the feature spectral coverage of the benzene of each training sample in described training sample from the described intercepting spectral coverage of each training sample described training sample;
Step S6, utilizes the feature spectral coverage of the described benzene of described training sample to set up the regression model between the feature spectral coverage of described benzene and benzene volume content; And
Step S7, process from described step S2 to described step S5 is carried out to sample to be tested, obtain the feature spectral coverage of the described benzene after the described Interference Peaks of deduction of described sample to be tested, using the input of the described regression model that this feature spectral coverage is set up in described step S6, obtain the predicted value of the volume content of the described benzene in described sample to be tested.
2. benzene content in gasoline Forecasting Methodology according to claim 1, the benzene volume content of wherein said training sample is measured by standard analytical process.
3. benzene content in gasoline Forecasting Methodology according to claim 1, the wherein said described Raman spectrum obtained measurement carries out pre-service and comprises smothing filtering, baseline correction and stable hydrocarbon normalization.
4. benzene content in gasoline Forecasting Methodology according to claim 1, apply described Raman spectrum analytic technique in wherein said step S4 to decompose the described intercepting spectral coverage of each described training sample and comprise described intercepting spectral coverage is decomposed into by the stacking pattern at two Voigt peaks, that is:
Wherein, V
i(v, φ
i) be Voigt function, v represents wave number, and r (v) represents regression criterion, and wave number is at 996cm
-1neighbouring spectrum peak is considered to the characteristic peak of benzene, is designated as V
b(v, φ
b), and another spectrum peak is considered to Interference Peaks, is designated as V
o(v, φ
o).
5. benzene content in gasoline Forecasting Methodology according to claim 4, wherein establishes A
bv () represents the feature spectral coverage of the benzene after the described Interference Peaks of deduction, then have:
A
B(v)＝A(v)-V
O(v,φ
O)。
6. benzene content in gasoline Forecasting Methodology according to claim 1 is any one in arithmetic of linearity regression, partial least square method, principal component regression algorithm and algorithm of support vector machine for setting up the algorithm of regression model in wherein said step S6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410336402.9A CN105319198B (en) | 2014-07-15 | 2014-07-15 | Benzene content in gasoline Forecasting Methodology based on Raman spectrum analytic technique |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410336402.9A CN105319198B (en) | 2014-07-15 | 2014-07-15 | Benzene content in gasoline Forecasting Methodology based on Raman spectrum analytic technique |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105319198A true CN105319198A (en) | 2016-02-10 |
| CN105319198B CN105319198B (en) | 2018-05-08 |
Family
ID=55247060
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410336402.9A Active CN105319198B (en) | 2014-07-15 | 2014-07-15 | Benzene content in gasoline Forecasting Methodology based on Raman spectrum analytic technique |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105319198B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106290294A (en) * | 2016-07-25 | 2017-01-04 | 浙江大学 | A kind of method utilizing copolymerization Jiao's microscopic Raman detection chlorophyll constituents of tea b content |
| CN106706546A (en) * | 2016-12-28 | 2017-05-24 | 中山市腾创贸易有限公司 | Analysis method for artificial intelligence learning materials on basis of infrared and Raman spectrum data |
| CN107818298A (en) * | 2017-09-27 | 2018-03-20 | 厦门大学 | General Raman spectral characteristics extracting method for machine learning material recognition |
| CN107843581A (en) * | 2017-10-20 | 2018-03-27 | 国家电网公司 | Raman spectrum based on Voigt peak type fitting algorithms quantitatively detects SF6Feature decomposition gas SO2The method of content |
| CN108444976A (en) * | 2018-04-26 | 2018-08-24 | 浙江大学 | A kind of heating value of natural gas measurement method based on Raman spectrum |
| CN108918502A (en) * | 2018-07-25 | 2018-11-30 | 安徽大学 | The detection method of fenifrothion pesticide residue in a kind of corn based on SERS |
| CN108982468A (en) * | 2018-07-13 | 2018-12-11 | 浙江大学 | A kind of Raman analysis method of trace impurity in p-chlorotoluene |
| CN108982470A (en) * | 2018-08-01 | 2018-12-11 | 中国石油化工集团公司 | A kind of method with LR laser raman method measurement kerosene performance |
| CN109030449A (en) * | 2018-04-25 | 2018-12-18 | 中国民航科学技术研究院 | A kind of lubricating oil and mixture ratio of fuel to oil rapid detection method |
| CN112304922A (en) * | 2020-10-29 | 2021-02-02 | 辽宁石油化工大学 | Method for quantitatively analyzing crude oil by Raman spectrum based on partial least square method |
| CN112362636A (en) * | 2020-11-04 | 2021-02-12 | 西安建筑科技大学 | Chlorite mineral species identification method based on Raman spectrum |
| CN112782146A (en) * | 2019-11-11 | 2021-05-11 | 中国石油天然气股份有限公司 | Gasoline olefin content analysis method based on Raman spectrum |
| WO2021149762A1 (en) * | 2020-01-24 | 2021-07-29 | パナソニックＩｐマネジメント株式会社 | Oil state diagnostic method and oil state diagnostic device |
| CN113655050A (en) * | 2021-08-17 | 2021-11-16 | 南京富岛信息工程有限公司 | Method for improving Raman spectrum detection limit of trace crude oil in light oil |
| CN113670894A (en) * | 2021-10-22 | 2021-11-19 | 港湾之星健康生物(深圳)有限公司 | Method for quantitatively monitoring substance content by Raman spectrum |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101403696A (en) * | 2008-10-21 | 2009-04-08 | 浙江大学 | Method for measuring gasoline olefin content based on Raman spectrum |
| CN101498667A (en) * | 2009-02-16 | 2009-08-05 | 浙江大学 | Method for detecting ethylene or ethylene propylene rubber content in ethylene-propylene copolymerization polypropylene |
| CN102230888A (en) * | 2011-06-16 | 2011-11-02 | 浙江大学 | Method for detecting content of plasticizing agent |
| CN102564965A (en) * | 2011-12-31 | 2012-07-11 | 聚光科技(杭州)股份有限公司 | Detecting method based on spectroscopy detection technology |
-
2014
- 2014-07-15 CN CN201410336402.9A patent/CN105319198B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101403696A (en) * | 2008-10-21 | 2009-04-08 | 浙江大学 | Method for measuring gasoline olefin content based on Raman spectrum |
| CN101498667A (en) * | 2009-02-16 | 2009-08-05 | 浙江大学 | Method for detecting ethylene or ethylene propylene rubber content in ethylene-propylene copolymerization polypropylene |
| CN102230888A (en) * | 2011-06-16 | 2011-11-02 | 浙江大学 | Method for detecting content of plasticizing agent |
| CN102564965A (en) * | 2011-12-31 | 2012-07-11 | 聚光科技(杭州)股份有限公司 | Detecting method based on spectroscopy detection technology |
Non-Patent Citations (3)
| Title |
|---|
| 姚捷 等: "基于拉曼特征峰的甲醇汽油甲醇含量测定", 《光散射学报》 * |
| 李津蓉: "拉曼光谱的数学解析及其在定量分析中的应用", 《中国博士学位论文全文数据库 工程科技I辑（月刊）》 * |
| 李津蓉等: "基于Voigt峰的快速光谱建模算法", 《光谱学与光谱分析》 * |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106290294A (en) * | 2016-07-25 | 2017-01-04 | 浙江大学 | A kind of method utilizing copolymerization Jiao's microscopic Raman detection chlorophyll constituents of tea b content |
| CN106290294B (en) * | 2016-07-25 | 2018-12-28 | 浙江大学 | A method of chlorophyll constituents of tea b content is detected using burnt microscopic Raman is copolymerized |
| CN106706546A (en) * | 2016-12-28 | 2017-05-24 | 中山市腾创贸易有限公司 | Analysis method for artificial intelligence learning materials on basis of infrared and Raman spectrum data |
| CN107818298A (en) * | 2017-09-27 | 2018-03-20 | 厦门大学 | General Raman spectral characteristics extracting method for machine learning material recognition |
| CN107818298B (en) * | 2017-09-27 | 2021-03-09 | 厦门大学 | General Raman spectrum feature extraction method for machine learning substance identification algorithm |
| CN107843581A (en) * | 2017-10-20 | 2018-03-27 | 国家电网公司 | Raman spectrum based on Voigt peak type fitting algorithms quantitatively detects SF6Feature decomposition gas SO2The method of content |
| CN109030449A (en) * | 2018-04-25 | 2018-12-18 | 中国民航科学技术研究院 | A kind of lubricating oil and mixture ratio of fuel to oil rapid detection method |
| CN108444976A (en) * | 2018-04-26 | 2018-08-24 | 浙江大学 | A kind of heating value of natural gas measurement method based on Raman spectrum |
| CN108444976B (en) * | 2018-04-26 | 2019-10-22 | 浙江大学 | A kind of heating value of natural gas measurement method based on Raman spectrum |
| CN108982468B (en) * | 2018-07-13 | 2020-02-28 | 浙江大学 | Raman analysis method for trace impurities in p-chlorotoluene |
| CN108982468A (en) * | 2018-07-13 | 2018-12-11 | 浙江大学 | A kind of Raman analysis method of trace impurity in p-chlorotoluene |
| CN108918502A (en) * | 2018-07-25 | 2018-11-30 | 安徽大学 | The detection method of fenifrothion pesticide residue in a kind of corn based on SERS |
| CN108982470A (en) * | 2018-08-01 | 2018-12-11 | 中国石油化工集团公司 | A kind of method with LR laser raman method measurement kerosene performance |
| CN112782146A (en) * | 2019-11-11 | 2021-05-11 | 中国石油天然气股份有限公司 | Gasoline olefin content analysis method based on Raman spectrum |
| CN112782146B (en) * | 2019-11-11 | 2022-11-04 | 中国石油天然气股份有限公司 | Gasoline olefin content analysis method based on Raman spectrum |
| WO2021149762A1 (en) * | 2020-01-24 | 2021-07-29 | パナソニックＩｐマネジメント株式会社 | Oil state diagnostic method and oil state diagnostic device |
| CN112304922A (en) * | 2020-10-29 | 2021-02-02 | 辽宁石油化工大学 | Method for quantitatively analyzing crude oil by Raman spectrum based on partial least square method |
| CN112362636A (en) * | 2020-11-04 | 2021-02-12 | 西安建筑科技大学 | Chlorite mineral species identification method based on Raman spectrum |
| CN113655050A (en) * | 2021-08-17 | 2021-11-16 | 南京富岛信息工程有限公司 | Method for improving Raman spectrum detection limit of trace crude oil in light oil |
| CN113655050B (en) * | 2021-08-17 | 2024-04-26 | 南京富岛信息工程有限公司 | Method for improving Raman spectrum detection limit of trace crude oil in light oil |
| CN113670894A (en) * | 2021-10-22 | 2021-11-19 | 港湾之星健康生物(深圳)有限公司 | Method for quantitatively monitoring substance content by Raman spectrum |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105319198B (en) | 2018-05-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105319198A (en) | Gasoline benzene content prediction method based on Raman spectrum analysis technology | |
| CN101403696A (en) | Method for measuring gasoline olefin content based on Raman spectrum | |
| CN109983338B (en) | Measurement of Hydrocarbon Fuel gas composition and Properties | |
| EP0877923B1 (en) | Mutivariate sample analysis | |
| EP2710353B1 (en) | SPECTROSCOPIC APPARATUS AND METHOD of DETERMINING COMPONENTS PRESENT IN A SAMPLE | |
| CN1982872B (en) | Near-infrared diffuse reflection spectral method for fastly inspecting drop water content | |
| US20040033617A1 (en) | Topological near infrared analysis modeling of petroleum refinery products | |
| CN112782146B (en) | Gasoline olefin content analysis method based on Raman spectrum | |
| JP2005512051A (en) | A method of analyzing unknowns as a blend of known substances calculated to match specific analytical data and predicting the properties of unknowns based on the calculated blend | |
| EP0663997B1 (en) | Spectroscopic instrument calibration | |
| CN101413884B (en) | Near-infrared spectrum analyzer and method for correcting resolution | |
| EP0851219B1 (en) | Comparison of data obtained from measurement instruments | |
| US7973926B1 (en) | Methods for determining olefin concentrations in olefin-containing fuels | |
| CN107966499B (en) | Method for predicting crude oil carbon number distribution by near infrared spectrum | |
| Nishii et al. | Selection of the NIR region for a regression model of the ethanol concentration in fermentation process by an online NIR and mid-IR dual-region spectrometer and 2D heterospectral correlation spectroscopy | |
| JP2003520347A (en) | Method and apparatus for online analysis of solvent mixtures | |
| EP3861320B1 (en) | Systems and methods for implicit chemical resolution of vacuum gas oils and fit quality determination | |
| US20230102813A1 (en) | Open-loop/closed-loop process control on the basis of a spectroscopic determination of undetermined substance concentrations | |
| Baeten et al. | Olive oil analysis by infrared and Raman spectroscopy: methodologies and applications | |
| CN111650179A (en) | Raman spectrum quantitative analysis method for three characteristic components in lavender essential oil | |
| CN111965166A (en) | Rapid measurement method for biomass briquette characteristic index | |
| CN111829976A (en) | Method for predicting composition of gasoline fraction hydrocarbon group of crude oil by near infrared spectrum | |
| Pereira et al. | A flow system for generation of concentration perturbation in two-dimensional correlation near-infrared spectroscopy: application to variable selection in multivariate calibration | |
| Macian et al. | Application assessment of UV–vis and NIR spectroscopy for the quantification of fuel dilution problems on used engine oils | |
| Pavoni et al. | FT‐IR Spectroscopy and Chemometrics as a Useful Approach for Determining Chemical‐Physical Properties of Gasoline, by Minimizing Analytical Times and Sample Handling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: Jiaochuan streets in Zhenhai District 315207 Zhejiang city of Ningbo Province Applicant after: Sinopec Corp. Applicant after: Zhejiang University Address before: 100728 Beijing, Chaoyangmen, North Street, No. 22, No. Applicant before: Sinopec Corp. Applicant before: Zhejiang University |
|
| COR | Change of bibliographic data | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |