Method for identifying high-temperature Daqu liquor based on color change of TMB (Tetramethylbenzidine) oxidized by porphyrin catalyst pseudoperoxidase activity

Abstract

The invention belongs to the field of white spirit identification, and particularly provides a high-temperature Daqu white spirit identification method based on color change of TMB (Tetramethylbenzidine) oxidized by porphyrin catalyst pseudoperoxidase activity. On the basis that nitrogen atoms in pyrazine compounds in Baijiu compete for metal ions in a porphyrin ring, and a pi-pi stacking effect between an aromatic ring and the porphyrin ring, the quasi-peroxidase activity of a porphyrin catalyst can be inhibited, so that TMB color change and visual signal amplification are realized; and finally, realizing rapid identification of the high-temperature Daqu liquor by combining a random forest algorithm and a classification algorithm. The method provides a new strategy for instrument-free rapid identification of high-temperature Daqu liquor.

Description

Technical Field

[0001] The invention belongs to the technical field of liquor identification, and specifically relates to a high-temperature Daqu liquor identification method based on porphyrin catalyst pseudo-peroxidase active oxidation TMB color change, which can be used for identifying liquors with the same aroma. Background Art

[0002] Microbial solid (semi-solid) fermentation, distillation and other processes have created different styles of liquor. More than 70% of my country's precious liquors, such as sauce-flavored, mixed-flavored, sesame-flavored and some strong-flavored liquors, use high-temperature Daqu technology. Liquor brewed by high-temperature Daqu technology with a koji making temperature exceeding 60°C has the characteristics of long brewing cycle, high cost, rich aroma and mellow taste. Due to the high value and good quality of this type of liquor, it is most popular among consumers. Some unscrupulous vendors use high-end bottles to fill low-end liquor to pass off as good ones, which not only infringes on the rights and interests of consumers but also seriously disrupts the order of the liquor market. Therefore, it is of great significance to develop a method for rapid identification of high-temperature Daqu liquors with the same aroma.

[0003] At present, modern analysis of liquor mainly focuses on quality grading, identification of storage years, origin and aroma type. The detection and analysis technologies used mainly rely on large and medium-sized instruments. Although electrochemical methods, mid-infrared spectroscopy and solid phase microextraction-gas chromatography-mass spectrometry have the advantages of accurate separation performance, high precision and high sensitivity, the above technologies still have the limitations of expensive equipment and the need for professional operators. They are difficult to use to accurately distinguish different grades of liquor with the same aroma, and cannot be used for on-site "instrument-free" rapid detection. In order to achieve on-site "instrument-free" identification of liquor, a variety of array sensors based on materials such as pH indicators, quantum dots and nano-precious metals have been developed. They mainly classify and identify the brand and aroma of liquor through comprehensive indicators such as pH, organic acids, carbonyl compounds and oxidative microenvironment in liquor, but they cannot accurately distinguish different grades of liquor with the same aroma. This may be related to the fact that the array sensors developed cannot specifically respond to a single compound and the response indicators are relatively single. At present, the results of colorimetric sensors usually have obvious color changes, and the detection results are visible to the naked eye. Colorimetric sensors have been successfully used in the detection of iron ions in water, humid gas and nitrite. Compared with spectroscopy and chromatography, it has the advantages of low cost, fast response speed and no instrumentation. Therefore, it is of great significance to establish an economical, easy-to-operate and fast colorimetric sensing detection method for the identification of high-temperature Daqu liquor to meet market demand. Summary of the invention

[0004] In order to solve the above problems existing in the prior art, the present invention provides a method for identifying high-temperature Daqu liquor based on the color change of TMB oxidized by porphyrin catalyst pseudoperoxidase activity to construct a multi-channel visual array sensor. Based on the extracted RGB values, an RF model and a DD-SIMCA model are constructed to finally realize the identification of high-temperature Daqu liquor of the same aroma type. The identification method has the characteristics of high sensitivity, fast response speed and low cost.

[0005] The technical solution adopted by the present invention to solve the above-mentioned problems is:

[0006] A method for identifying high-temperature Daqu liquor based on the color change of TMB oxidized by porphyrin catalyst pseudo-peroxidase activity, wherein a color developer and a porphyrin catalyst are added to a known liquor sample to obtain a standard solution, and then an RF model and / or a DD-SIMCA model are constructed based on the RGB values ​​extracted from the standard solution image, and finally the classification and / or authenticity identification of the liquor type to be detected is realized, and the method specifically comprises the following steps:

[0007] Step 1: Solution preparation

[0008] (1) Preparation of porphyrin catalyst solution

[0009] Accurately weigh the porphyrin catalyst, add an organic solvent to it and dissolve it by ultrasonication to obtain a porphyrin catalyst solution for standby use (preferably a concentration of 2×10 -5 ~5×10 -4 mol / L, more preferably 1×10 -4 mol / L),

[0010] (2) Preparation of blank solution

[0011] A blank solution is prepared by mixing TMB solution, hydrogen peroxide (preferably at a concentration of 30 wt%), porphyrin catalyst solution, and sodium acetate buffer (preferably at pH = 4, 1 mol / L) in a certain volume ratio;

[0012] (3) Preparation of standard solution

[0013] The TMB solution, hydrogen peroxide (preferably at a concentration of 30 wt%), porphyrin catalyst solution, sodium acetate buffer and a known liquor sample are mixed in a certain volume ratio to prepare a standard solution;

[0014] Step 2: Construction of sensor array

[0015] Place several 96-well plates in parallel, drip the pre-prepared blank solution into the first column; drip the standard solution into the second column, and finally use a camera to capture the image in the 96-well plate, extract the RGB value of the image, and form the corresponding sensor array;

[0016] Step 3: Build RF model

[0017] According to the sensor array in step 2, an RF model of a known liquor sample is established;

[0018] Step 4: Classify and predict the liquor to be tested and the real liquor

[0019] The TMB solution, hydrogen peroxide (preferably at a concentration of 30 wt%), porphyrin catalyst solution, sodium acetate buffer (preferably at pH = 4, 1 mol / L) and the liquor sample to be tested are mixed in a certain volume ratio to prepare a test solution;

[0020] The test solution is dropped into a 96-well plate, the image in the 96-well plate is captured by a camera, the RGB values ​​of the image are extracted, and then the RF model is used to classify and predict the liquor sample to be tested.

[0021] Preferably, in the method, the authenticity identification of the liquor to be tested is realized by using the independent soft mode algorithm DD-SIMCA of data-driven clustering:

[0022] ①Same as steps 1 and 2 above;

[0023] ② According to the sensor array composed of RGB values ​​in step 2, a DD-SIMCA model of known liquor samples is established;

[0024] ③ Prepare the test solution by referring to the operation of step 4 above, and use the DD-SIMCA model to identify the authenticity of the tested liquor.

[0025] Preferably, the porphyrin catalyst is one or more of tetraphenylporphyrin magnesium, tetraphenylporphyrin zinc, tetraphenylporphyrin, tetraphenylporphyrin copper, tetrakis (3-hydroxyphenyl) porphyrin and tetrakis-p-tolylporphyrin iron;

[0026] More preferably, when the porphyrin catalyst is a plurality of tetraphenylporphyrin magnesium, tetraphenylporphyrin zinc, tetraphenylporphyrin, tetraphenylporphyrin copper, tetrakis (3-hydroxyphenyl) porphyrin and tetrakis-p-tolylporphyrin iron (preferably: a combination of the above six catalysts with the same molar number), the specific operation of step 2 is:

[0027] ① Place several 96-well plates side by side, and drop the prepared blank solution into the first column;

[0028] ② Starting from the second column, the standard solutions prepared from the above-mentioned different single porphyrin catalyst solutions are added dropwise;

[0029] ③ The catalyst-free solution is dripped into each subsequent column. The catalyst-free solution is prepared by mixing TMB solution, hydrogen peroxide (preferably at a concentration of 30 wt%), sodium acetate buffer (preferably pH = 4, 1 mol / L), and a known liquor sample in a certain volume ratio to prepare a catalyst-free solution;

[0030] ④Finally, use a camera to capture the images of all the 96-well plates mentioned above, and then extract the RGB values ​​of the images to form the corresponding sensor array.

[0031] Preferably, the organic solvent is DMF or DMSO.

[0032] Preferably, the volume ratio of TMB solution, hydrogen peroxide, porphyrin catalyst solution, sodium acetate buffer and the liquor to be tested in the test liquid is 1:1:1:6:1; the volume ratio of TMB solution, hydrogen peroxide, porphyrin catalyst solution and sodium acetate buffer in the blank solution is 1:1:1:7; the volume ratio of TMB solution, hydrogen peroxide, sodium acetate buffer and known liquor sample in the catalyst-free solution is 1:1:7:1.

[0033] Preferably, the TMB solution is a 0.03 mol / L TMB solution in DMSO.

[0034] Preferably, the specific operation of establishing the RF model in step 3 is:

[0035] The random forest RF algorithm built by the "TreeBagger" function in Matlab2018b is used to model the RGB data in csv, and "oobvarimp" is set in this function to enable the variable importance evaluation of out-of-bag samples (i.e., samples that have not been sampled to train a certain tree). The number of decision trees nTree is optimized from 1 to 100 according to the RGB data. Based on the optimized nTree, samples accounting for 80% of the total sample volume are randomly selected as training sets, and the remaining samples are used as prediction sets to build a random forest model. Classification and prediction are then performed based on the accuracy of the model's training set and prediction set.

[0036] Preferably, the data-driven clustering independent soft mode algorithm (DD-SIMCA) is finally used to identify the adulteration of liquor:

[0037] Import the RGB data of the sensor array into MATLAB, run DDSGUI in the command window, and input the RGB data of known liquor samples in the pop-up window one by one to obtain the DD-SIMCA identification model diagram of the known liquor samples. When identifying the samples to be tested, after inputting the RGB data of the samples to be tested, the authenticity of the samples to be tested can be identified through the relationship between the threshold values ​​of the liquor samples to be tested and the known liquor samples in the diagram.

[0038] In the above operations, those skilled in the art can reasonably set the number of liquor samples according to needs and basic common sense, on the basis of ensuring the reliability of model establishment and the purpose of the experiment. Generally speaking, when using a single porphyrin catalyst, the sample size needs to be more than 30, and when the sample size is small, each sample has at least 15 parallel samples.

[0039] Compared with the prior art, the advantages and beneficial effects of the present invention are:

[0040] The present invention provides a method for classifying and identifying high-temperature Daqu liquor based on the color change of TMB oxidized by porphyrin catalyst pseudo-peroxidase activity. The method does not require the use of instruments and equipment and sample pretreatment, and only requires the extraction of RGB of the sample color change, and then combined with a chemometric method. The method has the advantages of high sensitivity, fast detection speed, wide application range, and no sample pretreatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 The schematic diagram of the method for identifying liquor by using porphyrin catalyst to oxidize TMB by pseudo-peroxidase activity according to the present invention;

[0042] Figure 2 The visualization result of liquor by the visualization array sensor based on porphyrin catalyst in Example 1;

[0043] Figure 3 The DD-SIMCA results of the visual array sensor in Example 1 for identifying different grades of liquors with the same aroma type: (A) Maotai-aroma type (MT1, MT2); (B) Maotai-aroma type (LJ1); (C) Blended-aroma type (BYB1); (D) Blended-aroma type (KZJ1); (E) Strong-aroma type (LZLJ1); (F) Strong-aroma type (WLY1); (G) Strong-aroma type (YH1), wherein in C and D: YQ1 represents YQJ1, and in E, F, and G: LZ1, LZ2, and LZ3 represent LZLJ1, LZLJ2, and LZLJ3, respectively. DETAILED DESCRIPTION

[0044] The following is an example of identifying different types of high-temperature Daqu liquor, and the technical solution of the present invention is further described in detail in conjunction with the accompanying drawings. The following examples are provided to facilitate a better understanding of the present invention, but are not intended to limit the present invention.

[0045] In the following examples, the liquor samples do not require pretreatment, and only a certain concentration of the test solution needs to be prepared. The test solution is dropped into a 96-well plate, and the color in the 96-well plate is captured with a camera. 10×10 pixels are extracted using Photoshop 2021 software, and then the image information is converted into RGB numerical information using Matlab 2018b. The random forest algorithm (RF) and the data-driven independent soft modeling algorithm (DD-SIMCA) are used in the Matlab (2018b) environment to build a model for rapid identification of high-temperature Daqu liquor.

[0046] Example 1

[0047] In this embodiment, 31 kinds of high-temperature Daqu liquors were selected as test samples (Table 1).

[0048] Table 1 Detailed information and serial number of high-temperature Daqu liquor

[0049]

[0050]

[0051] In this embodiment, the liquor samples were purchased from Wal-Mart (Wuhan, China), and the authenticity of the liquor can be confirmed by the traceability code on the wine box.

[0052] Tetraphenylporphyrin copper (TPP-Cu, 97%) and tetrakis-p-tolylporphyrin iron ((TMPP)FeCl, 95%) were purchased from Shanghai Aladdin Biotechnology Co., Ltd., China. Tetraphenylporphyrin magnesium (TPP-Mg, ≥95%) and tetrakis(3-hydroxyphenyl)porphyrin (3-THPP, 97%) were purchased from Shanghai Yuanye Biotechnology Co., Ltd., China. Tetraphenylporphyrin zinc (TPP-Zn, 98%) and tetraphenylporphyrin (TPP, 97%) were purchased from Shanghai MacLean Biochemical Technology Co., Ltd., China.

[0053] All chemicals were of analytical grade.

[0054] 1. First, prepare the porphyrin catalyst solution: accurately weigh appropriate amounts of TPP-Cu (DMF solution), (TMPP)FeCl (DMSO solution), TPP-Mg (DMF solution), 3-THPP (DMF solution), TPP-Zn (DMF solution), and TPP (DMF solution) to make 1×10 - 3 mol / L solution, ultrasonically dissolve, and dilute 10 times to obtain the porphyrin catalyst solution for standby use.

[0055] Second, the following tests were then carried out using each porphyrin catalyst solution:

[0056] (1) A 0.03 mol / L TMB solution in DMSO, 30 wt% H2O2, a porphyrin catalyst solution, a sodium acetate buffer solution (pH = 4, 1 mol / L, the same below), and a liquor sample were mixed in a volume ratio of 1:1:1:6:1 to prepare a test solution, and 10 parallels were prepared for each liquor sample; a blank control solution was prepared by mixing TMB, 30 wt% H2O2, a porphyrin catalyst solution, and a sodium acetate buffer solution in a volume ratio of 1:1:1:7.

[0057] (2) Take several 96-well plates and place them side by side. In the first column of the first 96-well plate, drip the prepared blank control solution; in the subsequent columns, drip ten parallels of each liquor sample (the volume dripped is consistent with the blank control), and a total of 310 sensing points of the liquor sample are obtained. Use a camera to capture the pictures in the 96-well plate. All photos were taken by a digital camera (Canon 750D). 10×10 pixel pictures were extracted by Photoshop 2021 software, and then the "imread" and other functions in Matlab 2018b were used to batch extract the RGB values ​​corresponding to the 10×10 pixel pictures and convert them into csv format. Then, the random forest (RF) algorithm constructed by the "TreeBagger" function in Matlab2018b is used to model the RGB data in csv, and "oobvarimp" is set in this function to enable variable importance evaluation of out-of-bag samples (i.e., samples that have not been sampled to train a certain tree). The number of decision trees nTree is optimized from 1 to 100 according to the RGB data. Based on the optimized nTree, samples accounting for 80% of the total sample volume (a total of 310 samples, 248 samples were selected this time) are randomly selected as training sets, and the remaining 62 samples are used as prediction sets to build a random forest model, and then classification and prediction are performed according to the accuracy of the model's training set and prediction set.

[0058] TMB (3,3',5,5'-tetramethylbenzidine) is a low-toxic, non-carcinogenic chromogenic substrate. It is colorless in itself, but changes color when oxidized. TMB is often used as a substrate for peroxidase. Under the catalysis of peroxidase, hydrogen peroxide decomposes to produce hydroxyl radicals, which oxidize TMB to produce a soluble blue product. The innovation of the present invention lies in the specificity of the sensing method, that is, the nitrogen atoms of the pyrazine compounds in the liquor can compete for the metal ions in the coordinated porphyrin ring, and the π-π stacking effect between the aromatic ring and the porphyrin ring can inhibit the pseudoperoxidase activity of the porphyrin catalyst, thereby affecting the TMB oxidation reaction and producing a significantly different color change. The schematic diagram is shown in FIG. Figure 1 shown.

[0059] A total of 6 porphyrin catalysts were used to construct the colorimetric array sensor, and the visualization results of each porphyrin catalyst on liquor were obtained. Figure 2 (Wherein, direct sample addition refers to the result obtained by mixing TMB, 30wt% H2O2, liquor sample, and sodium acetate buffer in a volume ratio of 1:1:1:7 as the test solution). Figure 2 It can be seen that different sensor units have certain differences in the color rendering of the same liquor sample, and the same sensor unit also has a certain distinguishing effect on the color rendering of the same-flavor liquor. However, from the RGB values ​​finally extracted from each sensor unit, there are significant differences, so the identification of high-temperature Daqu liquors with the same flavor can be achieved.

[0060] In order to better evaluate the recognition ability of the colorimetric array sensor for high-temperature Daqu liquor, the chemometric random forest algorithm was used to establish a high-temperature Daqu liquor discrimination model. The training set and prediction set accuracy results of the model are shown in Table 2, where the direct sample is the sample without the catalyst solution (that is, TMB, 30wt% H2O2 and liquor sample, sodium acetate buffer were mixed in a volume ratio of 1:1:1:7), and the sensor array is a combination of RGB data matrices extracted under the conditions of direct sample addition with 6 single catalysts and without catalyst. It can be seen from Table 2 that the training set discrimination accuracy of TPP-Mg, TPP-Zn, TPP-Cu, TPP and other sensor units for the same flavor liquor can reach 100%, and the prediction set accuracy is above 80%, and the recognition ability is significantly better than other sensor units. Surprisingly, the sensor array constructed by combining 7 sensor units has more advantages than single sensor units such as TPP-Mg and TPP-Zn. Not only can the results of the same aroma discrimination reach 100%, but even the accuracy of discriminating 31 types of liquor can reach 100%. This shows that the colorimetric array sensor constructed by combining multiple sensor units not only has a huge amount of data and rich response colors, but can also effectively avoid the errors of a single sensor and significantly improve the recognition ability of high-temperature Daqu liquor.

[0061] 3. Finally, the independent soft mode algorithm of data-driven clustering (DD-SIMCA) is used to identify adulteration of the same fragrance:

[0062] Import the RGB data of the sensor array into MATLAB (in this embodiment, the higher-priced liquors of the same flavor and brand are used as real samples, and the others are used as adulterated samples for identification), run DDSGUI in the command window, and input the RGB data of the real sample and the adulterated sample in the pop-up window to obtain the identification graphs of the three flavors. After beautifying and modifying the format, the DD-SIMCA results of the three flavors are obtained, see Figure 3 .

[0063] Figure 3 The DD-SIMCA acceptance graph of three types of liquor is shown. The line in the graph is the discrimination line, that is, the threshold of the real sample. The real sample is located in the lower left corner of the discrimination line in the graph, which is reflected in the samples such as Moutai Feitian, Baiyunbian 20 years, Guojiao 1573 and Wuliangye. The adulterated sample is located in the upper right corner of the discrimination line, belonging to other grades of liquor with similar taste in addition to the famous liquor in the same aroma. The identification results show that the real samples and adulterated samples in the three aromas can be accurately identified, and the classification identification rate reaches more than 99%, which finally shows that this method can achieve the identification of liquors of different grades with the same aroma.

[0064] Table 2 The result of discriminating high-temperature Daqu Baijiu by random forest algorithm

[0065]

[0066] The embodiments of the present invention are described in detail above, but the present invention is not limited to the above embodiments. In addition, all equivalent changes and modifications made without departing from the method and scope of the present invention should be included in the scope of the present invention.

Claims

1. A method for identifying high-temperature Daqu liquor based on the color change of TMB oxidized by porphyrin catalyst-mimicking peroxidase activity, wherein a color developer and a porphyrin catalyst are added to the liquor sample to obtain a standard solution, and then an RF model and / or a DD-SIMCA model are constructed based on the RGB values ​​extracted from the standard solution image, ultimately achieving the classification and / or authenticity identification of the liquor to be tested.

2. The method according to claim 1, characterized in that In the method, the classification of the liquor to be tested is achieved by the following steps: Step 1: Solution preparation (1) Preparation of porphyrin catalyst solution Accurately weigh the porphyrin catalyst, add an organic solvent thereto and dissolve it by ultrasonication to obtain a porphyrin catalyst solution for standby use; (2) Preparation of blank solution The blank solution was prepared by mixing TMB solution, hydrogen peroxide, porphyrin catalyst solution and sodium acetate buffer in a certain volume ratio; (3) Preparation of standard solution TMB solution, hydrogen peroxide, porphyrin catalyst solution, sodium acetate buffer and known liquor samples were mixed in a certain volume ratio to prepare a standard solution; Step 2: Construction of sensor array Place several 96-well plates in parallel, drip the pre-prepared blank solution into the first column; drip the standard solution into the second column, and finally use a camera to capture the image in the 96-well plate, extract the RGB value of the image, and form the corresponding sensor array; Step 3: Build RF model According to the sensor array in step 2, an RF model of a known liquor sample is established; Step 4: Classification and prediction of liquor to be tested The TMB solution, hydrogen peroxide, porphyrin catalyst solution, sodium acetate buffer solution and the liquor sample to be tested are mixed in a certain volume ratio to prepare a test solution; The test solution is dropped into a 96-well plate, the image in the 96-well plate is captured by a camera, the RGB values ​​of the image are extracted, and then the above-mentioned RF model is used to classify and predict the liquor samples to be tested.

3. The method according to claim 1, characterized in that In the method, the authenticity identification of the liquor to be tested is realized by using the independent soft mode algorithm DD-SIMCA of data-driven clustering, specifically: Step 1: Solution preparation (1) Preparation of porphyrin catalyst solution Accurately weigh the porphyrin catalyst, add an organic solvent thereto and dissolve it by ultrasonication to obtain a porphyrin catalyst solution for standby use; (2) Preparation of blank solution The blank solution was prepared by mixing TMB solution, hydrogen peroxide, porphyrin catalyst solution and sodium acetate buffer in a certain volume ratio; (3) Preparation of standard solution TMB solution, hydrogen peroxide, porphyrin catalyst solution, sodium acetate buffer and known liquor samples were mixed in a certain volume ratio to prepare a standard solution; Step 2: Construction of sensor array Place several 96-well plates in parallel, drip the pre-prepared blank solution into the first column; drip the standard solution into the second column, and finally use a camera to capture the image in the 96-well plate, extract the RGB value of the image, and form the corresponding sensor array; Step ③: Establish the DD-SIMCA model of known liquor samples; According to the sensor array in step ②, a DD-SIMCA model of known liquor samples is established; Step ④: Mix TMB solution, hydrogen peroxide, porphyrin catalyst solution, sodium acetate buffer and the liquor sample to be tested in a certain volume ratio to prepare a test solution; The test solution is dropped into a 96-well plate, the image in the 96-well plate is captured by a camera, and the RGB values ​​of the image are extracted. The DD-SIMCA model is used to identify the authenticity of the tested liquor.

4. The method according to claim 1, 2 or 3, characterized in that: The porphyrin catalyst is one or more of tetraphenylporphyrin magnesium, tetraphenylporphyrin zinc, tetraphenylporphyrin, tetraphenylporphyrin copper, tetrakis (3-hydroxyphenyl) porphyrin and tetrakis-p-tolylporphyrin iron.

5. The method according to claim 4, characterized in that When the porphyrin catalyst is a plurality of tetraphenylporphyrin magnesium, tetraphenylporphyrin zinc, tetraphenylporphyrin, tetraphenylporphyrin copper, tetrakis (3-hydroxyphenyl) porphyrin and tetrakis-p-tolylporphyrin iron, the specific operations of step 2 and step ② are: ① Place several 96-well plates side by side, and drop the prepared blank solution into the first column; ② Starting from the second column, the standard solutions prepared from the above-mentioned different single porphyrin catalyst solutions are added dropwise; ③ The catalyst-free solution is dripped into each subsequent column. The catalyst-free solution is prepared by mixing TMB solution, hydrogen peroxide, sodium acetate buffer, and known liquor samples in a certain volume ratio; ④Finally, use a camera to capture the images of all the 96-well plates mentioned above, and then extract the RGB values ​​of the images to form the corresponding sensor array.

6. The method according to claim 2 or 3, characterized in that: The organic solvent is DMF or DMSO; and / or The volume ratio of TMB solution, hydrogen peroxide, porphyrin catalyst solution, sodium acetate buffer and liquor to be tested in the test liquid is 1:1:1:6:1; and / or The volume ratio of TMB solution, hydrogen peroxide, porphyrin catalyst solution and sodium acetate buffer in the blank solution is 1:1:1:7; and / or The volume ratio of TMB solution, hydrogen peroxide, sodium acetate buffer and known liquor sample in the catalyst-free solution is 1:1:7:1; and / or The TMB solution is a 0.03 mol / L TMB DMSO solution; and / or The concentration of the porphyrin catalyst solution is 2×10 -5 ~5×10 -4 mol / L.

7. The method according to claim 2, characterized in that: The specific operation of establishing the RF model in step 3 is: The random forest RF algorithm built by the "TreeBagger" function in Matlab2018b is used to model the RGB data in csv, and "oobvarimp" is set in this function to enable variable importance evaluation of out-of-bag samples. The number of decision trees nTree is optimized from 1 to 100 according to the RGB data. Based on the optimized nTree, samples accounting for 80% of the total sample volume are randomly selected as training sets, and the remaining samples are used as prediction sets to build a random forest model. Classification and prediction are then performed based on the accuracy of the model's training set and prediction set.

8. The method according to claim 3, characterized in that In the step ③ and step ④, the independent soft mode algorithm DD-SIMCA of data-driven clustering is finally used to identify the adulteration of liquor. The specific operation is as follows: Import the RGB data of the sensor array into MATLAB, run DDSGUI in the command window, and input the RGB data of known liquor samples in the pop-up window one by one to obtain the DD-SIMCA identification model diagram of the known liquor samples. When identifying the samples to be tested, after inputting the RGB data of the samples to be tested, the authenticity of the samples to be tested can be identified through the relationship between the threshold values ​​of the liquor samples to be tested and the known liquor samples in the diagram.

Images