CN110672570A - Tea oil identification method based on three-dimensional fluorescence spectrum of vegetable oil - Google Patents
Tea oil identification method based on three-dimensional fluorescence spectrum of vegetable oil Download PDFInfo
- Publication number
- CN110672570A CN110672570A CN201910966520.0A CN201910966520A CN110672570A CN 110672570 A CN110672570 A CN 110672570A CN 201910966520 A CN201910966520 A CN 201910966520A CN 110672570 A CN110672570 A CN 110672570A
- Authority
- CN
- China
- Prior art keywords
- oil
- emission
- excitation
- fluorescence spectrum
- tea oil
- 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.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
- G01N21/63—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light optically excited
- G01N21/64—Fluorescence; Phosphorescence
Landscapes
- Health & Medical Sciences (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
The invention relates to a tea oil identification method based on a three-dimensional fluorescence spectrum of vegetable oil, which comprises the steps of converting a binary function z = f (excitation and emission) expressing the three-dimensional fluorescence spectrum of each vegetable oil sample into a unitary function z = f (excitation-emission) expressing the three-dimensional fluorescence spectrum of each vegetable oil sample through a software programming method, establishing a training set unitary fluorescence spectrum matrix with one-to-one correspondence between excitation-emission and three-dimensional fluorescence spectrum intensity of each vegetable oil sample according to the unitary function z = f (excitation-emission), introducing data in the training set unitary fluorescence spectrum matrix into data analysis software for analysis and establishing an identification model to obtain a set value range of a vegetable oil grade value, and identifying whether the tea oil is adulterated or not by using the identification model. The tea oil identification model constructed by the invention can identify adulterated tea oil with the additive amount of only 4% and the main component of the adulterant very similar to the tea oil; the research method provides a new way for applying the three-dimensional fluorescence spectrum to quickly and effectively identify the vegetable oil.
Description
Technical Field
The invention belongs to the technical field of oil doping detection, and particularly relates to a tea oil identification method based on a three-dimensional fluorescence spectrum of vegetable oil.
Background
The tea oil contains monounsaturated fatty acid which is necessary for human body and is up to 90 percent, is similar to olive oil which is known as 'golden liquid', has the beauty name of 'east olive oil', is also rich in vitamin E, tea polyphenol, camellin saponin, calcium, iron, zinc and other elements, antioxidant VE and squalene, is woody plant oil which has edible value, medical value and health care value simultaneously, is also one of high-quality edible oil which is recommended by grain and agriculture organization of the United nations, and is increasingly popular in China and some Asian countries.
With the increasing yield and the increasing use amount of the tea oil in recent years, illegal merchants mix low-price oil, illegal cooking oil and the like into high-price tea oil to destroy the quality of the tea oil so as to earn violence, seriously infringe the rights and interests of consumers and also jeopardize the safety of eaters, and the establishment of the efficient tea oil identification method has important significance for preventing the tea oil from being adulterated and maintaining the rights of consumers.
At present, the vegetable oil identification is mainly based on the detection of characteristic components of the vegetable oil, such as fatty acid composition, sterol and tocopherol, and the main detection methods comprise fluorescence spectrum, infrared spectrum, Raman spectrum, chromatogram, nuclear magnetic resonance and the like. The identification of vegetable oils by combining the chemometric method with the above detection method has been developed in recent years.
In various detection and identification methods, the fluorescence spectrum has high sensitivity and selectivity, and the main fluorescent substances of the vegetable oil are unsaturated fatty acid, vitamin E and pigment. The study of the "Synchronous Fluorescence Detection of Oil and improved Oil addition Identification" published in 2015 in Journal of the China core laboratory and Oils Association "shows that the Synchronous Fluorescence spectrograms of various doped special virgin olive Oils are collected and analyzed under the excitation light wave of 250-270 nm by using a Synchronous Fluorescence spectrometer, and the study result shows that most of the doped special virgin olive Oils can be distinguished from the olive Oils according to the Synchronous Fluorescence spectrums, but the Synchronous spectrums of the soybean Oil and the special virgin olive Oils are basically similar, so that the method can be distinguished only when a large amount of soybean Oil is doped in the olive Oils.
The three-dimensional fluorescence spectrum describes the relationship of fluorescence intensity simultaneously changing with excitation wavelength and emission wavelength, so that the fluorescence characteristics of a substance can be completely described. In the study of the paper "research on fluorescence spectroscopy of vegetable oil" published in 2009 on "biological impurities", analytical studies were performed on the simultaneous fluorescence spectra and unit fluorescence spectra of various vegetable oils, and the results of the studies show that: the three-dimensional spectrum has fingerprint identification capability and is more visual than information provided by a synchronous spectrum, but the three-dimensional fluorescence spectrum is related to the vegetable oil product, factors such as vegetable oil processing technology and fruit raw material sources also have great influence on the spectrum, the difference of the three-dimensional fluorescence spectrum of the adulterated tea oil and the tea oil is small, and the tea oil and the adulterated oil are difficult to distinguish visually by observing the three-dimensional fluorescence spectrum.
Disclosure of Invention
The invention aims to provide a tea oil identification method based on a vegetable oil three-dimensional fluorescence spectrum identification model, which comprises the steps of converting a binary function for expressing the three-dimensional fluorescence spectrum of the vegetable oil into a unitary function, establishing a training set unitary fluorescence spectrum matrix in which excitation-emission of each vegetable oil sample corresponds to three-dimensional fluorescence spectrum intensity one by one according to the unitary function, and on a multivariate variable statistical analysis software platform, carrying out discriminant analysis on data in the training set unitary fluorescence spectrum matrix by using orthogonal partial least squares to establish a tea oil identification model for specific and sensitive composite practical application and using the tea oil identification model for tea oil identification.
The technical scheme of the invention is as follows:
a tea oil identification method based on a three-dimensional fluorescence spectrum of vegetable oil is characterized in that a binary function z ═ f (excitation and emission) for expressing the three-dimensional fluorescence spectrum of each vegetable oil sample is converted into a unitary function z ═ f (excitation-emission) for expressing the three-dimensional fluorescence spectrum of each vegetable oil sample through a software programming method, a training set unitary fluorescence spectrum matrix with excitation-emission and three-dimensional fluorescence spectrum intensity corresponding to each vegetable oil sample one to one is established according to the unitary function z ═ f (excitation-emission), data in the training set unitary fluorescence spectrum matrix are led into data analysis software to be analyzed and a tea oil identification model is established, a set value range of vegetable oil grade values is obtained, and identification of true tea oil or adulterated tea oil is realized by using the identification model.
Further, the tea oil identification method based on the three-dimensional fluorescence spectrum of the vegetable oil comprises the following specific steps:
(1) collecting different vegetable oil samples to form a training set, wherein the training set comprises different tea oil samples, soybean oil samples and corn oil samples, and collecting fluorescence spectra of the different vegetable oil samples in the training set to obtain data of excitation positions, emission positions and corresponding fluorescence intensities of each vegetable oil sample expressed by a binary function z ═ f (excitation, emission);
(2) converting the binary function z ═ f (excitation, emission) expressing each vegetable oil sample obtained in the step (1) into a univariate function z ═ f (excitation-emission) capable of expressing the relation between the excitation-emission and the fluorescence intensity of each vegetable oil sample by a software programming method;
(3) collecting data expressing excitation-emission and fluorescence intensity in a univariate function z ═ f (excitation-emission) of the fluorescence spectrum of each vegetable oil sample in a training set sample to form a training set univariate fluorescence spectrum matrix;
(4) importing the data in the unitary fluorescence spectrum matrix of the training set obtained in the step (3) into data analysis software, carrying out statistical analysis on the data by using an orthogonal partial least squares discriminant analysis method in the data analysis software, further constructing a tea oil identification model, obtaining a set value range of different plant oil grades, and taking the set value range as a basis for tea oil identification;
(5) collecting tea oil samples which need to be identified and contain different brands and different production processes and adulterated tea oil samples doped with various low-price oils or mixed oils to form an inspection set, collecting three-dimensional fluorescence spectrum data of the tea oil samples and the adulterated tea oil samples in the inspection set, obtaining a binary function z ═ f (excitation, emission) for expressing the three-dimensional fluorescence spectrum of each tea oil sample, obtaining a univariate function z ═ f (excitation-emission) for expressing the fluorescence spectrum of each tea oil sample in the inspection set through the processing and analysis of the step (2), introducing data of excitation-emission and fluorescence intensity in the univariate function z ═ f (excitation-emission) into the tea oil identification model constructed in the step (4), and identifying the real tea oil or the adulterated tea oil in the inspection set according to the obtained scores.
Further, the software programming method in the step (2) is an SAS version 9.4 software programming method.
Further, the number of elements in the first row of the unitary fluorescence spectrum matrix of the training set in the step (3) is the number of variables expressing the unitary function z ═ f (excitation-emission) of each vegetable oil sample in the training set;
the number of variables of the unary function z (excitation-emission) is equal to the number of excitation positions x the number of emission positions;
excitation position number is excitation wavelength range/excitation slit;
emission position number-emission wavelength range/emission slit.
Further, the data analysis software in the step (4) is Umetrics SIMCA15.0.2 software.
Further, the three-dimensional fluorescence spectrum data acquisition method of each sample in the training set and each sample in the testing set in the steps (1) and (5) is as follows: cleaning a 10mm quartz sample cell by using petroleum ether, putting 1-2mL of sample in the sample cell, and placing the sample in a fluorescence spectrometer to acquire the excitation position, the emission position and the fluorescence intensity of the sample and draw a fluorescence contour map.
Further, the parameters for acquiring the three-dimensional fluorescence spectrum data of each sample in the training set and the test set of the fluorescence spectrometer in the steps (1) and (5) are as follows: the excitation wavelength range is 200-580nm, and the interval is 10 nm; the emission wavelength range is 250-700nm, and the interval is 10 nm; the scanning speed is 2000 nm/min; exciting a slit by 10 nm; an emission slit is 10 nm; the PMT voltage of the photomultiplier tube is 400V; the response time was 0.5 s.
Further, the adulterated tea oil in the step (5) takes real tea oil as main oil, and is respectively mixed with soybean oil, corn oil, cottonseed oil, palm oil, mixed oil 1 obtained by mixing the soybean oil, the palm oil, the cottonseed oil and the corn oil according to the volume ratio of 1:1:1:1, and mixed oil 2 obtained by mixing the soybean oil, the corn oil and the peanut oil according to the volume ratio of 1:1:1, wherein each kind of the adulterated tea oil comprises three different concentrations of 4%, 10% and 16%.
An application of a tea oil identification method based on a three-dimensional fluorescence spectrum of vegetable oil in identifying whether soybean oil, corn oil or palm oil is adulterated or not.
The invention has the following beneficial effects:
1. the invention provides a tea oil identification method based on three-dimensional fluorescence spectrum of vegetable oil, which is suitable for screening true or adulterated tea oil; converting a binary function z ═ f (excitation, emission) capable of expressing the three-dimensional fluorescence spectrum of the vegetable oil sample into a unitary function z ═ f (excitation-emission) expression by an SAS 9.4 version software programming method, simultaneously collecting the unitary function z ═ f (excitation-emission) expressing the three-dimensional fluorescence spectrum of each sample to form a training set unitary fluorescence spectrum matrix with excitation-emission and fluorescence intensity in one-to-one correspondence, analyzing data in the matrix by an orthogonal partial least squares discriminant analysis method in Umetrics SIMCA15.0.2 software, constructing an identification model based on the three-dimensional fluorescence spectrum data of the vegetable oil, dividing and setting score ranges belonging to different vegetable oil types in the identification model, converting the obtained binary function z ═ f (excitation, emission) expressing the three-dimensional light spectrum of the tea oil to be tested into the unitary function z ═ f (excitation-emission) and then introducing the binary function z ═ f (excitation-emission) into the model for identification, judging whether the tea oil is adulterated or not according to whether the obtained score falls within the set value range of the tea oil category score; in the invention, 27 test set samples (containing 7 tea oils with different brands and different production processes, 18 adulterated tea oils prepared in laboratories and 2 fake tea oils purchased in rural supermarkets) are identified, all the obtained identification results are correct, 4% of the addition amount of corn oil, soybean oil and various mixed oils which are extremely similar to the components of the tea oil fatty acids is detected, and a unitary function z ═ f (excitation-emission) is used for replacing a currently universal binary function z ═ f (excitation, emission) to represent three-dimensional fluorescence spectrum data to construct a tea oil identification model, so that the tea oil identification model is simple and convenient and can well reflect the class characteristics of the tea oil.
2. The tea oil identification method based on the three-dimensional fluorescence spectrum of the vegetable oil provided by the invention identifies the real tea oil or the adulterated tea oil by utilizing the identification model, the treatment process is simple, and the parameter R representing the fitting capacity in the finally constructed tea oil identification model20.84; parameter Q representing predictive power2The two parameters are both more than 0.5, which indicates that the model is an excellent identification model and the specificity and the sensitivity meet the practical requirements, and the identification method of the invention provides a new way for identifying the quality of the vegetable oil by using the three-dimensional fluorescence spectrum of the vegetable oil.
Drawings
FIG. 1 is a graph of fluorescence contours of randomly selected 1 tea oil sample (A), 1 soybean oil sample (B) and 1 corn oil sample (C) in a training set sample according to the present invention;
FIG. 2 is a fluorescence contour map of randomly selected 1 tea oil sample (A), 1 adulterated tea oil (B) doped with 4% low-price mixed oil 1 and 1 adulterated tea oil (C) doped with 10% corn oil in the test set sample according to the present invention;
FIG. 3 is a graph of fluorescence contours of 3 samples of tea oil of different brands selected from the training set of the present invention, (A) being Fujian Shenlangxiang tea oil, (B) being Fujian Fengda tea oil and (C) being Guangxi Bama tea oil;
FIG. 4 is a two-dimensional score plot of the OPLS-DA model of the present invention;
FIG. 5 is a histogram of the key parameters of the identification model of the present invention, where A is R2B is Q2。
Detailed Description
The present invention will be described in detail with reference to specific examples.
A tea oil identification method based on a three-dimensional fluorescence spectrum of vegetable oil is characterized in that a binary function z ═ f (excitation and emission) for expressing the three-dimensional fluorescence spectrum of each vegetable oil sample is converted into a unitary function z ═ f (excitation-emission) for expressing the three-dimensional fluorescence spectrum of each vegetable oil sample through a software programming method, a training set unitary fluorescence spectrum matrix with excitation-emission and three-dimensional fluorescence spectrum intensity corresponding to each vegetable oil sample one to one is established according to the unitary function z ═ f (excitation-emission), data in the training set unitary fluorescence spectrum matrix are led into data analysis software for analysis and identification model construction, and identification of true tea oil or fake tea oil is realized by using the identification model.
A tea oil identification method based on a three-dimensional fluorescence spectrum of vegetable oil comprises the following specific steps:
(1) collecting different vegetable oil samples to form a training set, wherein the training set comprises different tea oil samples, soybean oil samples and corn oil samples, and collecting fluorescence spectra of the different vegetable oil samples in the training set to obtain data of excitation positions, emission positions and corresponding fluorescence intensities of each vegetable oil sample expressed by a binary function z ═ f (excitation, emission); the acquisition parameters of the fluorescence spectrometer are as follows: the three-dimensional fluorescence spectrum data acquisition method of the training set sample comprises the following steps: cleaning a 10mm quartz sample cell by using petroleum ether, putting 1-2mL of sample in the sample cell, and placing the sample in a fluorescence spectrometer to acquire an excitation position, an emission position and fluorescence intensity of the sample and draw a fluorescence contour map; wherein the excitation wavelength range is 200-580nm, and the interval is 10 nm; the emission wavelength range is 250-700nm, and the interval is 10 nm; the scanning speed is 2000 nm/min; exciting a slit by 10 nm; an emission slit is 10 nm; the PMT voltage of the photomultiplier tube is 400V; response time 0.5 s;
(2) converting the binary function z ═ f (excitation, emission) expressing each vegetable oil sample obtained in the step (1) into a univariate function z ═ f (excitation-emission) capable of expressing the relationship between excitation-emission and fluorescence intensity of each vegetable oil sample by using a software programming method of version 9.4 of the SAS;
(3) collecting data expressing excitation-emission and fluorescence intensity in a univariate function z ═ f (excitation-emission) of the fluorescence spectrum of each vegetable oil sample in a training set sample to form a training set univariate fluorescence spectrum matrix;
(4) importing the data in the training set unitary fluorescence spectrum matrix obtained in the step (3) into Umetrics SIMCA15.0.2 analysis software, carrying out statistical analysis on the data by using an orthogonal partial least squares discriminant analysis method in the analysis software, further constructing a tea oil identification model, obtaining a set value range of various vegetable oil grades, and taking the set value range as a basis for tea oil identification;
(5) collecting tea oil samples which need to be identified and contain different brands and different production processes and adulterated tea oil samples doped with various low-price oils or mixed oils to form an inspection set, collecting three-dimensional fluorescence spectrum data of the tea oil samples and the adulterated tea oil samples in the inspection set, obtaining a binary function z ═ f (excitation, emission) for expressing the three-dimensional fluorescence spectrum of each tea oil sample, obtaining a univariate function z ═ f (excitation-emission) for expressing the fluorescence spectrum of each tea oil sample in the inspection set through the processing and analysis of the step (2), introducing data of excitation-emission and fluorescence intensity in the univariate function z ═ f (excitation-emission) into the tea oil identification model constructed in the step (4), and identifying the real tea oil or the adulterated tea oil in the inspection set according to the obtained scores.
Further, the number of elements in the first row of the unitary fluorescence spectrum matrix of the training set in the step (3) is the number of variables expressing the unitary function z ═ f (excitation-emission) of each vegetable oil sample in the training set;
the number of variables of the unary function z (excitation-emission) is equal to the number of excitation positions x the number of emission positions;
excitation position number is excitation wavelength range/excitation slit, (580-;
the number of emission positions is equal to the emission wavelength range/emission slit, (700-;
therefore, the number of variables of the unary function z (excitation-emission) is 38 × 45 and 1710, and the number of elements in the first row of the training set unary fluorescence spectrum matrix is 1710.
Selection and construction of training sets in the invention: because the components of various vegetable oils are similar, the prices of soybean oil and corn oil are relatively low in the vegetable oil and are the most possible adulterants in the adulterated tea oil, therefore, the inventor selects three vegetable oils of tea oil, soybean oil and corn oil in the selection of the training set samples, and considering that the components of the vegetable oil are related to the types of the vegetable oil and the planting places of plants and the processing technology of the vegetable oil, the invention selects 55 samples of the soybean oil, the corn oil and the tea oil together to form a training set; 19 soybean oil samples are purchased from six provinces such as Heilongjiang, Shandong, Fujian, Jilin, Shaanxi and Henan, 9 squeezing processes are carried out in the manufacturing process, 10 leaching processes are carried out, 11 samples are marked with transgenes, and 8 samples are marked with non-transgenes; 19 corn oil samples are purchased from six provinces such as Heilongjiang, Guangdong, Fujian, Gansu, Shaanxi and Henan, the manufacturing process is squeezing, 14 samples are marked with non-transgenosis, and 5 samples are not marked with non-transgenosis; 17 tea oil samples are purchased from Fujian, Jiangxi, Guangxi and Zhejiang, and the preparation process is squeezing; the fatty acid compositions of the samples in the 55 training sets are verified to meet the Chinese national standard of corresponding vegetable oil by gas chromatography (GB/T17377-2008).
Selection and construction of test sets in the present invention: 27 concentrated tea oil samples are detected, including tea oil samples purchased by 2 village shops, and the tea oil samples are false tea oil which is determined by gas chromatography that the fatty acid composition of the tea oil samples does not accord with the set value of the tea oil; 7 tea oil samples purchased from a large supermarket and in the same batch with the training set sample in different brands; the 18 home-made adulterated tea oil samples are prepared by respectively doping soybean oil, cottonseed oil, corn oil, palm oil, mixed oil 1 or mixed oil 2 into tea oil serving as main oil, wherein the adulteration concentrations are respectively 4%, 10% and 16%, the mixed oil 1 is prepared by mixing the soybean oil, the palm oil, the cottonseed oil and the corn oil according to a volume ratio of 1:1:1:1, the mixed oil 2 is prepared by mixing the soybean oil, the corn oil and the peanut oil according to a volume ratio of 1:1:1, and each kind of prepared adulterated tea oil comprises three different concentrations of 4%, 10% and 16%.
Referring to the attached drawing 1, in order to train the fluorescence contour line spectrogram of 1 sample tea oil, 1 sample soybean oil and 1 sample corn oil randomly selected from all samples in the set, it can be seen from the attached drawing 1 that the fluorescence spectrograms of the tea oil, the soybean oil and the corn oil samples are similar, and after comparing the three, the tea oil has no obvious spectrogram characteristics, and it can not directly judge whether the sample is the tea oil according to the fluorescence contour line spectrogram.
Referring to the attached figure 2, in order to test the fluorescence contour line spectra of 1 randomly selected tea oil sample, 1 adulterated tea oil doped with 4% of low-price mixed oil 1 and 1 adulterated tea oil doped with 10% of corn oil from all the tea oil samples in the set, it can be seen from the attached figure 2 that the spectrograms of the adulterated tea oil doped with 10% of corn oil and the tea oil are almost the same, and the spectrograms of the adulterated tea oil doped with 4% of mixed oil 1 and the tea oil are different, so that the truth of the tea oil cannot be directly judged according to the fluorescence contour line spectra.
Referring to fig. 3, the fluorescent contour line maps of three brands of tea oil in a tea oil sample are shown, wherein the spectrogram of fujian fengda tea oil (B) has a stronger fluorescent peak at the pigment position than the spectrogram of fujian sheng lang xiang tea oil (a) and guangxiama tea oil (C), so that the fujian sheng lang xiang tea oil is dark in color and unrefined, and the fujian sheng lang xiang tea oil and guangxiama tea oil are lighter in color and refined; as can be seen from the above, the three-dimensional fluorescence spectra of the tea oil samples have different contour maps due to different processing techniques and different plant origins, which cause different compositions of the types of fluorescent substances such as unsaturated fatty acids, tocopherols, pigments, and the like.
The strongest fluorescence peak positions of the tea oil, the corn oil and the soybean oil are shown in table 1, table 1 lists that 3 tea oil samples in 17 tea oil samples in the training set have two stronger fluorescence peaks, the 3 samples are all unrefined tea oil with darker colors, and the other 14 samples are refined tea oil and only have one stronger fluorescence peak.
TABLE 1 strongest fluorescence peak position and intensity value of tea oil, corn oil and soybean oil
In summary, with reference to fig. 1, fig. 2, fig. 3 and table 1, the conclusion can be drawn: the spectrogram difference of the tea oil with the corn oil and the soybean oil is small, and the spectrogram of the tea oil of different processing techniques has larger difference, so if the low-price corn oil, the soybean oil or the mixed oil thereof is mixed in the tea oil, the tea oil identification can not be carried out by only observing the spectrogram.
In the invention, the three-dimensional fluorescence spectrum of each sample in a training set or an inspection set is expressed by a binary function z ═ f (EI, EM), the EI represents an excitation position, the EM represents an emission position, and z represents the fluorescence spectrum intensity at the corresponding position, the small program is run on an SAS software platform, the binary function z ═ f (EI, EM) is converted into a univariate function z ═ f (EI-EM) to be expressed, the univariate function z ═ f (EI-EM) is integrated to form a training set univariate fluorescence spectrum matrix, wherein the first row in the matrix is sequentially EI200-EM250 EI200-EM260I 200-270 EI 270 … … EI200-EM700 EI210-EM260EI210-EM270 EI 270 … … EI210-EM700 EI 250-EM 210-EM270 … … 200-EM700, the last row is sequentially EI 250 EI580, EI 250 EI580-EM260 EI580-EM 580-270-EI 580-EM270 … … 580-EI 580-, the first column is the sample number, and the other cell values are the fluorescence spectrum intensity z at the corresponding EI-EM location for each sample in the training set; and importing a data set in the unitary fluorescence spectrum matrix of the training set into SIMCA15.0.2 software, analyzing the data by adopting an orthogonal partial least squares discriminant analysis method (OPLS-DA), further obtaining the identification scores of different vegetable oil varieties in the training set, and forming an identification model for discriminating the tea oil by the identification scores of the different vegetable oil varieties.
Referring to fig. 4, fig. 4 is a two-dimensional score chart obtained by analyzing data in a training set unitary fluorescence spectrum matrix based on OPLS-DA, and it can be seen from fig. 4 that a tea oil region is far from a soybean oil region and a corn oil region, the tea oil can be well distinguished from the corn oil and the soybean oil, and the corn oil region is very close to the soybean oil region; the scores of the tea oil in the categories of the tea oil in the training set sample are all in the range of 0.70-1.15, so that the score range is used as the set value range of the tea oil, and when the score of the identified tea oil sample is not in the range of 0.70-1.15, the sample is adulterated tea oil.
FIG. 5 is a histogram of key parameters of the identification model obtained by analyzing data in a unitary fluorescence spectrum matrix of a training set by OPLS-DA to construct the identification model, and is used for evaluating the specificity and sensitivity of the model; from FIG. 5, it can be seen that the parameter R representing the fitting ability in the identification model20.84, parameter Q representing predictive power20.72, variable number 2, wherein a represents R2B represents Q2The identification model constructed by the invention is an excellent identification model and the specificity and the sensitivity meet the practical requirements.
The binary function z ═ f (EI, EM) expressing the three-dimensional fluorescence spectrum of each sample in the test set was converted into a univariate function z ═ f (EI-EM) according to the procedure (2) in the example, and the data was imported into an identification model for identification prediction, the prediction results are shown in table 2:
TABLE 2 test set sample tea oil Category score values
| Sample numbering | Score value | The result of the judgment | Sample numbering | Score value | The result of the judgment |
| Mix-C-D-Ⅰ | 0.32 | T | Mix-C-Mix 1-III | 0.38 | T |
| Mix-C-D-Ⅱ | 0.45 | T | Mix-C-Mix 2-I | 0.39 | T |
| Mix-C-D-Ⅲ | 0.25 | T | Mix-C-Mix 2-II | 0.35 | T |
| Mix-C-MZ-Ⅰ | 0.45 | T | Mix-C-Mix 2-III | 0.41 | T |
| Mix-C-MZ-Ⅱ | 0.44 | T | M2018-61-A | 0.45 | T |
| Mix-C-MZ-Ⅲ | 0.40 | T | M2018-62-A | 0.63 | T |
| Mix-C-YM-Ⅰ | 0.29 | T | M2018-1-C | 1.15 | T |
| Mix-C-YM-Ⅱ | 0.25 | T | M2018-2-C | 1.05 | T |
| Mix-C-YM-Ⅲ | 0.14 | T | M2018-3-C | 0.72 | T |
| Mix-C-Z-Ⅰ | 0.37 | T | M2018-4-C | 1.15 | T |
| Mix-C-Z-Ⅱ | 0.16 | T | M2018-5-C | 0.80 | T |
| Mix-C-Z-Ⅲ | 0.05 | T | M2018-6-C | 0.91 | T |
| Mix-C-Mix 1-I | 0.37 | T | M2018-7-C | 0.86 | T |
| Mix-C-Mix 1-II | 0.38 | T |
Note: c in the sample numbers represents tea oil, D soybean oil, MZ cottonseed oil, YM corn oil, Z palm oil, mixed 1 represents mixed oil 1, mixed 2 represents mixed oil 2; in the sample names, I, II and III respectively represent adulteration concentrations of 4 percent, 10 percent and 16 percent; M2018-61-A and M2018-62-A represent 2 tea oil samples purchased from rural small stores; M2018-1-C to M2018-7-C represent tea oil samples purchased from a large supermarket.
In table 2, 7 tea oil samples purchased from a large supermarket represented by M2018-1-C to M2018-7-C are identified, and the score values are all in the range of 0.70-1.15, and the tea oil is judged to be true tea oil;
Mix-C-D-I, Mix-C-D-II and Mix-C-D-III respectively represent the tea oil samples doped with soybean oil with the concentration of 4%, 10% and 16%, and the samples are judged to be the adulterated tea oil after the identification that the score values are not in the range of 0.70-1.15;
Mix-C-MZ-I, Mix-C-MZ-II and Mix-C-MZ-III respectively represent the tea oil samples with the doping concentration of 4%, 10% and 16% cottonseed oil, and the scores are not in the range of 0.70-1.15 after identification, and the tea oil samples are judged to be adulterated tea oil;
Mix-C-YM-I, Mix-C-YM-II and Mix-C-YM-III respectively represent tea oil samples doped with corn oil with concentrations of 4%, 10% and 16%, and the samples are judged to be adulterated tea oil if the scores are not within the range of 0.70-1.15 after identification;
Mix-C-Z-I, Mix-C-Z-II and Mix-C-Z-III respectively represent tea oil samples doped with palm oil with the concentrations of 4%, 10% and 16%, and the samples are judged to be the adulterated tea oil after the identification that the score values are not in the range of 0.70-1.15;
Mix-C-Mix 1-I, Mix-C-Mix 1-II and Mix-C-Mix 1-III respectively represent tea oil samples doped with 4%, 10% and 16% of mixed oil 1 (mixed oil 1 obtained by mixing soybean oil, palm oil, cottonseed oil and corn oil according to the volume ratio of 1:1:1: 1), and the samples are judged to be adulterated tea oil if the scores are not within the range of 0.70-1.15 after identification;
Mix-C-Mix 2-I, Mix-C-Mix 2-II and Mix-C-Mix 2-III represent tea oil samples doped with 4%, 10% and 16% of mixed oil 2 (mixed oil 2 obtained by mixing soybean oil, corn oil and peanut oil according to the volume ratio of 1:1: 1), and the samples are judged to be adulterated tea oil if the scores are not within the range of 0.70-1.15 after identification;
M2018-61-A and M2018-62-A represent 2 tea oil samples purchased from small villages, and the samples are judged to be adulterated tea oil if the scores are not within the range of 0.70-1.15 after identification. In conclusion, the prediction of the test set samples by the identification method of the invention is all correct, which shows that the authenticity of the tea oil can be accurately judged by the method of the invention.
The specificity, sensitivity and level data of the model are shown in Table 3
TABLE 3 OPLS-DA tea oil identification model specificity, sensitivity and levels thereof
| Sample overview | Sensitivity of the composition | Specificity of | Adulteration concentration level (%) |
| Tea oil | --- | 100％(7/7) | --- |
| Soybean oil-doped tea oil | 100％(3/3) | --- | 4 |
| Tea oil blended with cottonseed oil | 100％(3/3) | --- | 4 |
| Tea oil doped with corn oil | 100％(3/3) | --- | 4 |
| Palm oil-blended tea oil | 100％(3/3) | --- | 4 |
| Tea oil doped with |
100％(3/3) | --- | 4 |
| Tea oil doped with |
100％(3/3) | --- | 4 |
| Tea oil purchased from small village shops | 100％(2/2) | --- | --- |
The tea oil identification model idea formed based on the unitary function z ═ f (excitation-emission) data expressing the three-dimensional fluorescence spectra of different vegetable oils can also be used for true and false identification of other vegetable oils.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by the present specification, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (9)
1. A tea oil identification method based on three-dimensional fluorescence spectrum of vegetable oil is characterized by comprising the following steps: transforming a binary function z = f (excitation and emission) expressing the three-dimensional fluorescence spectrum of each vegetable oil sample into a unitary function z = f (excitation-emission) expressing the three-dimensional fluorescence spectrum of each vegetable oil sample by a software programming method, establishing a training set unitary fluorescence spectrum matrix in which excitation-emission of each vegetable oil sample corresponds to three-dimensional fluorescence spectrum intensity one by one according to the unitary function z = f (excitation-emission), importing data in the training set unitary fluorescence spectrum matrix into data analysis software for analysis and constructing a tea oil identification model to obtain a set value range of tea oil category scores, and identifying true tea oil or adulterated tea oil by using the identification model.
2. The method for identifying the tea oil based on the three-dimensional fluorescence spectrum of the vegetable oil as claimed in claim 1, which is characterized by comprising the following specific steps:
(1) collecting different vegetable oil samples to form a training set, wherein the training set comprises different tea oil samples, soybean oil samples and corn oil samples, and collecting fluorescence spectra of the different vegetable oil samples in the training set to obtain data of an excitation position, an emission position and corresponding fluorescence intensity of each vegetable oil sample expressed by a binary function z = f (excitation, emission);
(2) converting the binary function z = f (excitation, emission) expressing each vegetable oil sample obtained in the step (1) into a univariate function z = f (excitation-emission) capable of expressing the relation between the excitation-emission and the fluorescence intensity of each vegetable oil sample by a software programming method;
(3) collecting data expressing excitation-emission and fluorescence intensity in a univariate function z = f (excitation-emission) of the fluorescence spectrum of each vegetable oil sample in a training set sample to form a training set univariate fluorescence spectrum matrix;
(4) importing the data in the unitary fluorescence spectrum matrix of the training set obtained in the step (3) into data analysis software, carrying out statistical analysis on the data by using an orthogonal partial least squares discriminant analysis method in the data analysis software, further constructing a tea oil identification model, obtaining a set value range of different plant oil grades, and taking the set value range as a basis for tea oil identification;
(5) collecting tea oil samples to be identified and containing different brands and different production processes and adulterated tea oil samples doped with various low-price oils or mixed oils to form an inspection set, collecting three-dimensional fluorescence spectrum data of the tea oil samples and the adulterated tea oil samples in the inspection set, obtaining a binary function z = f (excitation, emission) for expressing the three-dimensional fluorescence spectrum of each tea oil sample, obtaining a univariate function z = f (excitation-emission) for expressing the fluorescence spectrum of each tea oil sample in the inspection set through the processing and analysis of the step (2), introducing data of the excitation-emission and the fluorescence intensity in the univariate function z = f (excitation-emission) into the tea oil identification model constructed in the step (4), and identifying the real tea oil or the adulterated tea oil in the inspection set according to the obtained scores.
3. The method for identifying the tea oil based on the three-dimensional fluorescence spectrum of the vegetable oil as claimed in claim 2, wherein the method comprises the following steps: the software programming method in the step (2) is an SAS 9.4 version software programming method.
4. The method for identifying the tea oil based on the three-dimensional fluorescence spectrum of the vegetable oil as claimed in claim 2, wherein the method comprises the following steps: the number of elements in the first row of the unitary fluorescence spectrum matrix of the training set in the step (3) is the variable number of the unitary function z = f (excitation-emission) of each vegetable oil sample in the expression training set;
the number of variables of the univariate function z = f (excitation-emission) = number of excitation positions x number of emission positions;
excitation position number = excitation wavelength range/excitation slit;
number of emission positions = emission wavelength range/emission slit.
5. The method for identifying the tea oil based on the three-dimensional fluorescence spectrum of the vegetable oil as claimed in claim 2, wherein the method comprises the following steps: the data analysis software in the step (4) is Umetrics SIMCA15.0.2 software.
6. The method for identifying the tea oil based on the three-dimensional fluorescence spectrum of the vegetable oil as claimed in claim 2, wherein the method comprises the following steps: the three-dimensional fluorescence spectrum data acquisition method of each sample in the training set and each sample in the testing set in the step (1) and the step (5) is as follows: cleaning a 10mm quartz sample cell by using petroleum ether, putting 1-2mL of sample in the sample cell, and placing the sample in a fluorescence spectrometer to acquire the excitation position, the emission position and the fluorescence intensity of the sample and draw a fluorescence contour map.
7. The method for identifying the camellia oil based on the three-dimensional fluorescence spectrum of the vegetable oil as claimed in claim 6, wherein the method comprises the following steps: the acquisition parameters of the fluorescence spectrometer for acquiring the three-dimensional fluorescence spectrum data of each sample in the training set and the inspection set in the steps (1) and (5) are as follows: the excitation wavelength range is 200-580nm, and the interval is 10 nm; the emission wavelength range is 250-700nm, and the interval is 10 nm; the scanning speed is 2000 nm/min; exciting a slit by 10 nm; an emission slit is 10 nm; the PMT voltage of the photomultiplier tube is 400V; the response time was 0.5 s.
8. The method for identifying the tea oil based on the three-dimensional fluorescence spectrum of the vegetable oil as claimed in claim 2, wherein the method comprises the following steps: in the step (5), the adulterated tea oil takes real tea oil as main oil, soybean oil, corn oil, cottonseed oil and palm oil, mixed oil 1 obtained by mixing the soybean oil, the palm oil, the cottonseed oil and the corn oil according to the volume ratio of 1:1:1:1 and mixed oil 2 obtained by mixing the soybean oil, the corn oil and the peanut oil according to the volume ratio of 1:1:1 are respectively added, and each kind of the prepared adulterated tea oil comprises three different concentrations of 4%, 10% and 16%.
9. Use of the method for identifying tea oil according to claim 1 based on three-dimensional fluorescence spectroscopy of vegetable oil for identifying whether soybean oil, corn oil, cottonseed oil or palm oil is adulterated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910966520.0A CN110672570A (en) | 2019-10-12 | 2019-10-12 | Tea oil identification method based on three-dimensional fluorescence spectrum of vegetable oil |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910966520.0A CN110672570A (en) | 2019-10-12 | 2019-10-12 | Tea oil identification method based on three-dimensional fluorescence spectrum of vegetable oil |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110672570A true CN110672570A (en) | 2020-01-10 |
Family
ID=69081742
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910966520.0A Withdrawn CN110672570A (en) | 2019-10-12 | 2019-10-12 | Tea oil identification method based on three-dimensional fluorescence spectrum of vegetable oil |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110672570A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114018887A (en) * | 2021-11-05 | 2022-02-08 | 国网安徽省电力有限公司电力科学研究院 | Method for rapidly detecting antioxidant content in transformer insulating oil |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102645424A (en) * | 2012-03-02 | 2012-08-22 | 宁夏大学 | Method for rapidly detecting edible oil added with illegal cooking oil |
| CN104502320A (en) * | 2014-12-29 | 2015-04-08 | 江南大学 | Method for identifying strong flavor Baijiu by combining three-dimensional fluorescence spectrum with PCA-SVM |
| CN105628661A (en) * | 2015-12-30 | 2016-06-01 | 中国环境科学研究院 | Method for simultaneously detecting particulate organics and dissolved organics in water body |
| CN106950241A (en) * | 2017-02-27 | 2017-07-14 | 南昌大学 | A kind of method of other adulterated oil species and content in prediction tea oil |
| CN107894408A (en) * | 2017-11-24 | 2018-04-10 | 中国农业科学院油料作物研究所 | A kind of edible oil based near infrared spectrometer is polynary to mix pseudo- discrimination method |
| CN109115744A (en) * | 2018-10-16 | 2019-01-01 | 四川理工学院 | The method that edible oil mixes pseudo- frying oil is quickly detected using three-dimensional fluorescence spectrum technology |
-
2019
- 2019-10-12 CN CN201910966520.0A patent/CN110672570A/en not_active Withdrawn
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102645424A (en) * | 2012-03-02 | 2012-08-22 | 宁夏大学 | Method for rapidly detecting edible oil added with illegal cooking oil |
| CN104502320A (en) * | 2014-12-29 | 2015-04-08 | 江南大学 | Method for identifying strong flavor Baijiu by combining three-dimensional fluorescence spectrum with PCA-SVM |
| CN105628661A (en) * | 2015-12-30 | 2016-06-01 | 中国环境科学研究院 | Method for simultaneously detecting particulate organics and dissolved organics in water body |
| CN106950241A (en) * | 2017-02-27 | 2017-07-14 | 南昌大学 | A kind of method of other adulterated oil species and content in prediction tea oil |
| CN107894408A (en) * | 2017-11-24 | 2018-04-10 | 中国农业科学院油料作物研究所 | A kind of edible oil based near infrared spectrometer is polynary to mix pseudo- discrimination method |
| CN109115744A (en) * | 2018-10-16 | 2019-01-01 | 四川理工学院 | The method that edible oil mixes pseudo- frying oil is quickly detected using three-dimensional fluorescence spectrum technology |
Non-Patent Citations (1)
| Title |
|---|
| 尚云鹏: "三维荧光光谱分类识别方法与实验研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114018887A (en) * | 2021-11-05 | 2022-02-08 | 国网安徽省电力有限公司电力科学研究院 | Method for rapidly detecting antioxidant content in transformer insulating oil |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Šmejkalová et al. | High-power gradient diffusion NMR spectroscopy for the rapid assessment of extra-virgin olive oil adulteration | |
| Aparicio et al. | Authenticity of olive oil: Mapping and comparing official methods and promising alternatives | |
| Brody et al. | A study of amber and copal samples using FT-Raman spectroscopy | |
| Gurdeniz et al. | Detection of adulteration of extra-virgin olive oil by chemometric analysis of mid-infrared spectral data | |
| Diomande et al. | Multi-element, multi-compound isotope profiling as a means to distinguish the geographical and varietal origin of fermented cocoa (Theobroma cacao L.) beans | |
| JP2019095447A (en) | Method for multivariate adulteration identification of edible oil based on near-infrared spectroscopy | |
| CN106950241B (en) | A kind of method of other adulterated oil type and contents in prediction tea oil | |
| Bandoniene et al. | Suitability of elemental fingerprinting for assessing the geographic origin of pumpkin (Cucurbita pepo var. styriaca) seed oil | |
| Aceto et al. | A preliminary study on the authentication and traceability of extra virgin olive oil made from Taggiasca olives by means of trace and ultra-trace elements distribution | |
| CN103792198A (en) | Intermediate infrared-near infrared correlation spectrum discriminating method for melamine-doped milk | |
| CN104237370B (en) | A kind of personation sesame oil method for quick identification for adding sesame oil essence | |
| CN106841083A (en) | Sesame oil quality detecting method based on near-infrared spectrum technique | |
| Dankowska et al. | Application of synchronous fluorescence spectroscopy for determination of extra virgin olive oil adulteration | |
| Tan et al. | Edible oils adulteration: A review on regulatory compliance and its detection technologies | |
| Cao et al. | Combined application of fluorescence spectroscopy and chemometrics analysis in oxidative deterioration of edible oils | |
| CN108519361A (en) | A kind of method that fluorescence spectrum differentiates cochin yellow wingceltis and Ovshinsky yellow wingceltis seeds | |
| Lia et al. | Application of fluorescence spectroscopy and chemometric models for the detection of vegetable oil adulterants in Maltese virgin olive oils | |
| CN112305108A (en) | Camellia seed oil adulteration detection method based on oleic acid/behenic acid and beta-resinol/campesterol ratio | |
| CN110672570A (en) | Tea oil identification method based on three-dimensional fluorescence spectrum of vegetable oil | |
| Mi et al. | Authentication of the geographical origin of sesame seeds based on proximate composition, multi-element and volatile fingerprinting combined with chemometrics | |
| Bao et al. | A comparative evaluation of low-field and high-field NMR untargeted analysis: Authentication of virgin coconut oil adulterated with refined coconut oil as a case study | |
| He et al. | Molecular level characterization of the effect of roasting on the extractable components of glandless cottonseed by Fourier transform ion cyclotron resonance mass spectrometry | |
| CN108982570A (en) | A kind of edible oil quality discrimination method based on nuclear magnetic resonance technique | |
| Zhang et al. | Recent advances in analytical detection of olive oil adulteration | |
| Wang et al. | Authentication of American ginseng (Panax quinquefolius L.) from different origins by linear discriminant analysis of multi-elements |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20200110 |
|
| WW01 | Invention patent application withdrawn after publication |