A method for evaluating the quality of fresh produce during storage based on a modified color-sensitive sensor array

Abstract

The present invention relates to a method for evaluating the quality of fresh food during storage period based on a modified color-sensitive sensor array. First, an MSNs@dye nano color-sensitive sensor array or a sensitized color-sensitive sensor array is prepared; then fresh food samples of different types and quality grades are selected, and the color-sensitive sensor array reacts with the fresh food samples to obtain characteristic variables of the color-sensitive unit, and a reverse neural network model for evaluating the quality of fresh food is constructed. Finally, the sample to be tested reacts to obtain the characteristic variables of the sample to be tested, and substitutes them into the reverse neural network model to complete the rapid evaluation of the quality of fresh food. The method for evaluating the quality of fresh food during storage period of the present invention uses a modified color-sensitive sensor array to detect fresh food, and cooperates to establish a dedicated reverse neural network model, and realizes rapid evaluation of the quality of fresh food by obtaining information on characteristic volatile components of fresh food, thereby solving the problems of complex detection process, strong subjectivity of results, low sensitivity and poor stability in the prior art.

Description

Technical Field

[0001] The present invention relates to the field of non-destructive detection of fresh produce quality during storage, and in particular to a fresh produce quality evaluation method based on a modified color-sensitive sensor array during storage. Background Art

[0002] Fresh produce has become one of the most in-demand foods in China. Given the unique characteristics of fresh produce, the utilization of nondestructive rapid testing technologies is crucial to ensure its quality, effectively prevent food safety issues caused by counterfeit and substandard goods, and better meet the requirements for nondestructive fresh produce quality authentication. This is crucial for improving the quality and safety of fresh produce and enhancing its quality authentication. Fresh produce quality characteristics primarily include texture, color, tenderness, pH, and freshness, with freshness being the most important indicator for evaluating the quality and safety of fresh produce. Research on freshness testing has been extensive both domestically and internationally, with sensory evaluation being the most common method. Sensory evaluation involves a panel of professional evaluators scoring and evaluating each indicator according to a standardized criteria. While this method is the simplest and most convenient method for assessing freshness, it is subjective and insensitive to low-concentration gases produced during the early stages of freshness. Therefore, it requires a comprehensive assessment based on physical and chemical properties and microbiological indicators.

[0003] Chemical evaluation involves analyzing chemical indicators such as volatile basic nitrogen (TVB-N), pH, potassium, and thiobarbituric acid (TBA) to determine the freshness of fresh produce. TVB-N is the most authoritative indicator of freshness and is typically determined using semi-micro nitrogen determination or microdiffusion. my country stipulates that a TVB-N value exceeding 20 mg / kg is inedible. Furthermore, during the spoilage period, spoilage bacteria utilize protein degradation products to grow and decompose nitrogenous substances, producing odorous gases such as trimethylamine (TMA). Therefore, TMA can be used as a characteristic gas of fresh produce spoilage. Using these traditional methods to analyze fish freshness is often time-consuming, labor-intensive, and somewhat destructive. For a time, the use of electronic noses to analyze fish freshness was a hot topic. These devices primarily utilize odor sensors, data processing equipment, and analysis software to capture and detect gases by simulating the human olfactory system. It not only allows for simple comparative analysis of the odor information of different samples, but also builds a database by collecting standard sample information, and utilizes chemometric statistical analysis methods to conduct qualitative and quantitative analysis of unknown samples, offering the advantages of rapid, convenient, and non-destructive testing. However, electronic nose gas sensors are expensive to manufacture, have difficulty distinguishing similar substances, and are easily affected by high humidity environments. Most importantly, they have low sensitivity to gases such as thiols and amines, which are often characteristic of fresh food spoilage.

[0004] Compared to electronic nose technology, using gas array sensors to detect the freshness of fresh produce has certain advantages. First, electronic nose technology uses metal oxide sensors. The gas-sensitive materials in these sensors primarily rely on physical adsorption and van der Waals forces, which are extremely weak intermolecular forces, to react with the gas being measured. Color sensors, on the other hand, use chemically reactive dyes such as metalloporphyrins. These dyes, which possess functional groups that strongly react with the gas being measured, bind to the gas molecules through strong forces such as metallic bonds, π-π bonds, and hydrogen bonds, resulting in higher sensitivity. Second, when metal sensors are used in high ambient temperatures or after prolonged operation, the sensor's response baseline value can easily shift, leading to inaccurate measurements. Furthermore, while metal sensors adsorb the target gas, they are also affected by water vapor in the environment, causing them to age due to moisture. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method for evaluating the quality of fresh produce during storage period based on a modified color-sensitive sensor array, so as to solve the problems existing in the prior art of complex detection process, strong subjectivity of results, low sensitivity and poor stability.

[0006] To solve the above technical problems, the technical solution of the present invention is: a method for evaluating the quality of fresh produce during storage based on a modified color-sensitive sensor array, the innovation of which is that it includes the following steps:

[0007] S1: Preparation of color-sensitive sensor array:

[0008] First, a modified color-sensitive dye is obtained by adding MSNs or performing sensitization, and then the modified color-sensitive dye is fixed on a substrate to prepare an MSNs@dye nano color-sensitive sensor array or a sensitized color-sensitive sensor array.

[0009] S2: Establishment of fresh storage quality model:

[0010] First, fresh samples of different types and quality levels are selected, wherein the types of fresh samples include oysters and pork, and the different quality levels of the samples include the freshness levels of fresh samples of different storage periods;

[0011] Secondly, use a camera to obtain an image of the color-sensitive sensor array before the reaction, and save the image in a computer;

[0012] Subsequently, the fresh sample is placed in a reaction container, and the color-sensitive sensor array prepared in step S1 and the fresh sample are stored in the same sealed reaction container to allow the color-sensitive sensor array and the volatile odor substances in the fresh sample to fully react. After the reaction, an image of the color-sensitive sensor array after the reaction is captured using a camera, and the obtained image is stored in a computer. The computer is used to locate the positions of the color-sensitive units in the images before and after the reaction, extract their color features, and subtract the grayscale means of each color-sensitive unit before and after the reaction. The difference in grayscale means is obtained, which is the characteristic variable of the color-sensitive unit;

[0013] Finally, the characteristic variables obtained from all fresh samples are combined to obtain a characteristic matrix. The characteristic matrix is ​​used as input and the quality grade of the detected fresh samples as output to construct a reverse neural network model for fresh produce quality evaluation, which is the required fresh produce storage period quality model.

[0014] S3: Rapid evaluation of fresh produce quality:

[0015] The sample to be tested is reacted according to the method described in step S2 to obtain the characteristic variables of the sample to be tested, and the characteristic variables of the sample to be tested are substituted into the fresh storage period quality model constructed in step S2 to obtain the quality grade information of the sample to be tested and complete the rapid evaluation of the fresh quality.

[0016] Furthermore, in step S1, the preparation of the MSNs@dye nano color-sensitive sensor array is specifically as follows:

[0017] First, 100 mL of anhydrous ethanol and 300 mL of deionized water were mixed in a 500 mL three-necked flask, and then 6 mL of ammonia water was added. The entire system was continuously stirred at 25 ° C. Then, 25 mg of hexadecyltriethylammonium bromide was added. After the solution was mixed evenly, 0.2 g of tetraethoxysilane was added. The reaction time was 10-14 hours. Finally, the filtered suspension was calcined in a muffle furnace at 540 ° C for 4-8 hours to eliminate the template to obtain MSNs nanomaterials.

[0018] Then, the MSNs nanomaterial is added to ethanol to obtain an MSNs dispersion, and the MSNs dispersion is then mixed with a prepared color-sensitive solution to obtain a mixed solution, wherein the color-sensitive solution is any one of an ethanol solution containing a pH indicator, a dichloromethane solution containing a metal porphyrin, or a dichloromethane solution containing fluoroborane pyrrole, and polyethylene glycol 600 is added to the mixed solution. After stirring, an MSNs@dye nano-color-sensitive dye is obtained, and the MSNs@dye nano-color-sensitive dye is fixed on a substrate to prepare an MSNs@dye nano-color-sensitive sensor array, and the nano-color-sensitive sensor array is stored in a sealed bag for standby use;

[0019] The preparation of the enhanced color sensitive sensor array is specifically as follows:

[0020] First, an acid-base solution is mixed with a pH indicator to obtain a sensitized solution of the pH indicator. The sensitized solution of the pH indicator is then mixed with a metal porphyrin solution to obtain a sensitized color-sensitive dye. The acid-base solution is an ethanol solution containing TBAH or TsOH. Then, the sensitized color-sensitive dye is spotted on a substrate and placed in a 50-60°C oven to obtain a sensitized color-sensitive sensor array. The sensitized color-sensitive sensor array is taken out of the oven and stored in a sealed bag for future use.

[0021] Furthermore, the color-sensitive solution is prepared in X amounts, where X is a positive integer. In the color-sensitive solution, the pH indicator and ethanol are used in a ratio of 20 mg:10 mL, the metalloporphyrin and dichloromethane are used in a ratio of 20 mg:10 mL, and the fluoroborane and dichloromethane are used in a ratio of 20 mg:10 mL.

[0022] Furthermore, the concentrations of the color-sensitive solution and the MSNs dispersion are both 2 mg / mL, the color-sensitive solution and the MSNs dispersion are mixed in a volume ratio of 1:5, the amount of polyethylene glycol 600 added is 25% of the volume of the mixed solution, the concentration of TBAH in the ethanol solution is 1 moL / L, the concentration of TBAH in the ethanol solution is 0.8 moL / L, and the amount of MSNs@dye nano color-sensitive dye or sensitized color-sensitive dye fixed on the substrate is 1.5-2 μL.

[0023] Further, the pH indicator is any one of cresol red, bromophenol blue, bromocresol green, bromocresol purple, chlorophenol red, basic fuchsin or neutral red, the metal porphyrin is any one of tetraphenylporphyrin zinc, tetraphenylporphyrin manganese or 1,7-dimethyl-3,5-distyryl (3,5-di-tert-butyl-4-hydroxyphenyl)-8-(4-benzoic acid methyl ester)-4,4'-difluoro-4-boron-3a,4a-diaza-s-and diindene, the fluoroboryl pyrrole is 8-(4-carbazolephenyl)-4,4-difluoroboryldipyrrolmethane, 8-(4-nitrophenyl)-4,4-difluoro-2,6-dibromoboryldipyrrolmethane, 8-(4-nitrophenyl)-4,4-difluoro-2,6-dibromoboryldipyrrolmethane, Any one of bis[(6-methoxy-2-naphthyl)-4,4-difluoroboron dipyrromethane, 8-(6-methoxy-2-naphthyl)-4,4-difluoroboron dipyrromethane, bis[8-phenyldipyrromethane]nickel(II), bis[8-(4-methylformylphenyl)dipyrromethane]nickel(II) or bis[8-(6-methoxy-2-naphthyl)dipyrromethane]nickel(II), the metalloporphyrin in the metalloporphyrin solution is 5,10,15,20-tetrakis(4-methoxyphenyl)-21H,23H-porphine iron(III) chloride or 5,10,15,20-tetraphenyl-21H,23H-porphine zinc, and the substrate is a silica gel plate or qualitative filter paper.

[0024] Furthermore, when the mixed solution and polyethylene glycol 600 were stirred, the stirring speed was 350 r / min, the stirring time was 10 min, and the temperature was increased to 90° C. at a rate of 5° C. / 5 min and maintained for 60 min.

[0025] Furthermore, in step S2, when the fresh sample is oyster, the amount used is 15-18 g, and when the fresh sample is pork, the amount used is 10-15 g; the detection time of the color-sensitive sensor array is 10-20 minutes.

[0026] Furthermore, in step S2, the feature variable extraction steps are as follows: first, the position of each color-sensitive unit of the color-sensitive sensor array is located by a computer; then, the image of the color-sensitive sensor array before and after the reaction is obtained by the camera is decomposed into R channel, G channel and B channel grayscale images; then, the grayscale mean value of each sensitive unit before and after the reaction is subtracted, that is, △R=R a -R b , △G=G a -G b , △B=B a -B b , where subscript a is the value after reaction and subscript b is the value before reaction

[0027] and ED are the characteristic variables of a color-sensitive unit. X color-sensitive units obtain Y characteristic variables, where Y = 4X;

[0028] The number of fresh samples used to construct the fresh storage quality evaluation model is selected as N, which includes n quality grades and m samples for each grade, that is, N = n*m, where n is a positive integer not less than 2, and m and N are both positive integers;

[0029] The steps for establishing the fresh storage period quality evaluation model are as follows: the characteristic matrix is ​​recorded as S, S is N*Y, the characteristic variable matrix S is used as the input value, and the quality grade matrix T corresponding to the fresh sample is used as the output value to construct a reverse neural network model for fresh storage period quality evaluation;

[0030]

[0031]

[0032] The advantages of the present invention are: the method for evaluating the quality of fresh produce during storage uses a modified color-sensitive sensor array to detect fresh produce, and cooperates with the establishment of a dedicated reverse neural network model to obtain information on characteristic volatile components of fresh produce to achieve rapid evaluation of fresh produce quality, thereby solving the problems of complex detection process, strong subjectivity of results, low sensitivity and poor stability in the prior art.

[0033] The present invention causes the color-sensitive material to react with the volatile components in fresh food through coordination to generate a complex, which further changes the molecular structure of the color-sensitive material and causes the color of the color-sensitive material to change.

[0034] The present invention prepares a nano color-sensitive sensor array by coupling MSNs with color-sensitive materials, breaking through the difficulty of controllable preparation of nano color-sensitive sensor arrays. The surface of MSNs contains unsaturated metal binding sites, which can form complexes with volatile substances to achieve adsorption of volatile substances, thereby improving the sensitivity of detection. In addition, the pH value of the pH indicator is adjusted so that the indicator reaches its respective mutation range, and the optimal acid and base addition amount of the pH indicator is determined to enhance the sensitivity to characteristic gases.

[0035] The reverse neural network model for rapid evaluation of fresh produce quality established by the present invention uses the signal values ​​collected by the nano color-sensitive sensor array as input and uses the different storage periods of fresh produce as output to establish an evaluation model for the quality of fresh produce during storage. This model has good versatility, improves the efficiency of quality evaluation, and is of great significance for the evaluation of fresh produce quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0037] Figure 1 This is a scanning electron microscope image of MSNs in the present invention.

[0038] Figure 2 This is a transmission electron microscope image of MSNs in the present invention.

[0039] Figure 3 This is the energy dispersion spectrum of MSNs in the present invention.

[0040] Figure 4 It is the adsorption-desorption hysteresis line of MSNs in the present invention.

[0041] Figure 5 This is the pore size distribution diagram of MSNs in the present invention.

[0042] Figure 6 Schematic diagram of the MSNs@dye nano color-sensitive sensor array in the present invention.

[0043] Figure 7 Schematic diagram of the enhanced color-sensitive sensor array in the present invention.

[0044] Figure 8 Characteristic images of the MSNs@dye nano color-sensitive sensing array of oysters during different storage periods in the present invention.

[0045] Figure 9 is the TVB-N value of oysters at different storage times in the present invention.

[0046] Figure 10 This is the result of principal component analysis (PCA) of oysters in the present invention.

[0047] Figure 11 This is a scatter plot of the correlation between the TVB-N measurement value and the MSNs@dye nano color-sensitive sensor array prediction value in the present invention.

[0048] Figure 12 It is a scatter plot of the correlation between the measured value and the predicted value of the color-sensitive material not doped with MSNs in the present invention.

[0049] Figure 13 These are characteristic images of the color-sensitive sensor array of pork during different storage periods in the present invention.

[0050] Figure 14 These are the PCA model results for pork at different storage periods in the present invention.

[0051] Figure 15 These are the KNN model results for pork at different storage periods in the present invention.

[0052] Figure 16 These are the LDA model results for pork at different storage periods in the present invention.

[0053] Figure 17-Figure 19 These are the PLS model results for pork at different storage periods in the present invention.

[0054] Figure 20 This is a comparison chart of the Euclidean distance before and after sensitization in the present invention.

[0055] Figure 21 It is the RGB response value before sensitization in the present invention.

[0056] Figure 22 It is the RGB response value after sensitization in the present invention. DETAILED DESCRIPTION

[0057] The following embodiments may enable those skilled in the art to more fully understand the present invention, but the present invention is not limited to the scope of the embodiments.

[0058] The method for evaluating the quality of fresh produce during storage based on the modified color-sensitive sensor array of the present invention is implemented by the following steps:

[0059] Preparation of color-sensitive sensor array:

[0060] First, a modified color-sensitive dye is obtained by adding MSNs or performing sensitization, and then the modified color-sensitive dye is fixed on a substrate to form an MSNs@dye nano color-sensitive sensor array or a sensitized color-sensitive sensor array. In the present invention, MSNs refers to mesoporous silica nanospheres.

[0061] The preparation of MSNs@dye nano color-sensitive sensor array is as follows:

[0062] First, 100 mL of anhydrous ethanol and 300 mL of deionized water were mixed in a 500 mL three-necked flask, and then 6 mL of ammonia water was added. The entire system was continuously stirred at 25°C, and then 25 mg of hexadecyltriethylammonium bromide was quickly added. After the solution was mixed evenly, 0.2 g of tetraethoxysilane was added. The reaction time was 10-14 h. Finally, the filtered suspension was calcined in a muffle furnace at 540°C for 4-8 h to eliminate the template to obtain MSNs nanomaterials.

[0063] pass Figure 1 、 Figure 2 It can be seen that the calcined white powder is composed of spherical nanoparticles with a diameter of about 250nm. These nanoparticles have porous properties. This structure can provide more adsorption sites for the adsorption of volatile substances and improve the enrichment rate of volatile substances. Figure 3 The presence of Si and O elements indicates that the spherical nanoparticles are likely composed of silicon dioxide. The presence of a small amount of C is due to the presence of hydrocarbons in the air and the large amount of carbon in the conductive band. Figure 4 shows a typical H1 adsorption-desorption hysteresis loop, indicating the existence of a system in which the porous adsorbent exhibits capillary condensation. In addition, there is a saturated adsorption platform on the adsorption isotherm of the H1 type hysteresis loop, reflecting a relatively uniform pore size distribution. Figure 5 The pore size distribution of white nanospheres is shown in Figure 2. The pore size of MSNs is mainly concentrated around 12 nm. The results show that MSNs have the ability to adsorb small molecules for chemical reactions.

[0064] Then, the prepared MSNs nanomaterials were added into ethanol to obtain an MSNs dispersion, and the dosage relationship of the MSNs nanomaterials and ethanol was 20 mg:10 mL.

[0065] Nine color-sensitive solutions were prepared, consisting of three dichloromethane solutions containing metalloporphyrins and six dichloromethane solutions containing fluoroborane.

[0066] The nine color-sensitive solutions are respectively labeled as color-sensitive solution A (S1), color-sensitive solution B (S2), color-sensitive solution C (S3), color-sensitive solution D (S4), color-sensitive solution E (S5), color-sensitive solution F (S6), color-sensitive solution G (S7), color-sensitive solution H (S8), and color-sensitive solution I (S9).

[0067] Color-sensitive solution A is a dichloromethane solution containing zinc tetraphenylporphyrin, and the dosage ratio of zinc tetraphenylporphyrin to dichloromethane is 20 mg:10 mL.

[0068] Color-sensitive solution B is a dichloromethane solution containing tetraphenylporphyrin manganese, and the dosage ratio of tetraphenylporphyrin manganese to dichloromethane is 20 mg:10 mL.

[0069] Color-sensitive solution C is a dichloromethane solution containing 1,7-dimethyl-3,5-distyryl(3,5-di-tert-butyl-4-hydroxyphenyl)-8-(4-benzoic acid methyl ester)-4,4'-difluoro-4-boron-3a,4a-diaza-s-indane. The dosage ratio of 1,7-dimethyl-3,5-distyryl(3,5-di-tert-butyl-4-hydroxyphenyl)-8-(4-benzoic acid methyl ester)-4,4'-difluoro-4-boron-3a,4a-diaza-s-indane to dichloromethane is 20 mg:10 mL.

[0070] Color-sensitive solution D is a dichloromethane solution containing 8-(4-carbazolephenyl)-4,4-difluoroborondipyrrolmethane, and the usage ratio of 8-(4-carbazolephenyl)-4,4-difluoroborondipyrrolmethane to dichloromethane is 20 mg:10 mL.

[0071] Color-sensitive solution E is a dichloromethane solution containing 8-(4-nitrophenyl)-4,4-difluoroborondipyrrolmethane, and the dosage ratio of 8-(4-nitrophenyl)-4,4-difluoroborondipyrrolmethane to dichloromethane is 20 mg:10 mL.

[0072] Color-sensitive solution F is a dichloromethane solution containing 8-(6-methoxy-2-naphthyl)-4,4-difluoroborondipyrrolmethane, and the usage ratio of 8-(6-methoxy-2-naphthyl)-4,4-difluoroborondipyrrolmethane and dichloromethane is 20 mg:10 mL.

[0073] Color-sensitive solution G is a dichloromethane solution containing bis[8-phenyldipyrromethene]nickel(II), and the dosage ratio of bis[8-phenyldipyrromethene]nickel(II) to dichloromethane is 20 mg:10 mL.

[0074] Color-sensitive solution H is a dichloromethane solution containing bis[8-(4-methylformylphenyl)dipyrromethene]nickel(II), and the amount ratio of bis[8-(4-methylformylphenyl)dipyrromethene]nickel(II) to dichloromethane is 20 mg:10 mL.

[0075] Color-sensitive solution I is a dichloromethane solution containing bis[8-(6-methoxy-2-naphthyl)dipyrromethene]nickel(II), and the usage ratio of bis[8-(6-methoxy-2-naphthyl)dipyrromethene]nickel(II) to dichloromethane is 20 mg:10 mL.

[0076] Nine dichloromethane solutions containing metalloporphyrins and one dichloromethane solution containing fluoroborane pyrrole were mixed with MSNs dispersions at a volume ratio of 1:5 to obtain eight mixed solutions. 25% of the volume of polyethylene glycol 600 was added to each mixed solution. Subsequently, the mixed solutions were stirred at 50°C for 10 min at a speed of 350 r / min. The temperature was raised to 90°C at a rate of 5°C / 5 min and maintained for 60 min. After sufficient reaction, eight MSNs@dye nano-color-sensitive dyes were obtained, which were respectively labeled S1@MSNs, S2@MSNs, S3@MSNs, S4@MSNs, S5@MSNs, S6@MSNs, S7@MSNs, S8@MSNs, and S9@MSNs.

[0077] Finally, 2 μL of MSNs@dye nano color-sensitive dye was aspirated using a microcapillary and fixed on a silica gel plate to prepare a MSNs@dye nano color-sensitive sensor array. Figure 6 As shown, the nano color-sensitive sensor array is stored in a sealed bag for future use.

[0078] The preparation of the enhanced color sensitive sensor array is specifically as follows:

[0079] First, an acid-base solution is mixed with a pH indicator to obtain a sensitized solution of the pH indicator, and then the sensitized solution of the pH indicator is mixed with a metal porphyrin solution to obtain a sensitized color-sensitive dye with higher stability. The acid-base solution is an ethanol solution containing TBAH or TsOH. In the present invention, TBAH refers to tetrabutylammonium hydroxide and TsOH refers to p-toluenesulfonic acid.

[0080] Finally, use a microcapillary to draw 2 μL of the enhanced color-sensitive dye and fix it on the qualitative filter paper, and place it in a 50-60 ° C oven to obtain the enhanced color-sensitive sensor array. After taking the enhanced color-sensitive sensor array out of the oven, store it in a sealed bag for later use.

[0081] Six pH indicators were prepared and labeled as P1, P4, P9, P3, P8 and P11, wherein P1 was a 2 mg / mL cresol red solution, P4 was a 2 mg / mL bromocresol green solution, P9 was a 2 mg / mL basic fuchsin solution, P3 was a 2 mg / mL bromophenol blue solution, P8 was a 2 mg / mL chlorophenol red solution, and P11 was a 2 mg / mL neutral red solution. Two metalloporphyrin solutions were then prepared and labeled as T4 and T9, wherein T4 was a 2 mg / mL 5,10,15,20-tetrakis(4-methoxyphenyl)-21H,23H-porphine iron(II) chloride. I) solution, T9 is a 2 mg / mL 5,10,15,20-tetraphenyl-21H,23H-porphine zinc solution, then 0.5 μL of an ethanol solution containing 1 mol / L TsOH is mixed with 1 mL of P1, 0.7 μL of an ethanol solution containing 1 mol / L TsOH is mixed with 1 mL of P4, 0.7 μL of an ethanol solution containing 1 mol / L TBAH is mixed with 1 mL of P9, T9 is mixed with P3 at a ratio of 1:0.8, T4 is mixed with P8 at a ratio of 1:0.8, and T4 is mixed with P11 at a ratio of 1:0.8 to prepare the following: Figure 7 The 3*2 enhanced color-sensitive sensor array shown in the figure has the Euclidean channel data values ​​before and after enhancement as shown in the figure. Figure 20 As shown, the RGB response values ​​before and after sensitization are as follows Figure 21 、 Figure 22 shown.

[0082] The evaluation method of the present invention is further described below by using the MSNs@dye nano color-sensitive sensor array to evaluate the quality of oysters during storage and the sensitized color-sensitive sensor array to evaluate the quality of pork during storage.

[0083] Evaluation of oyster storage quality based on MSNs@dye nano color-sensitive sensor array:

[0084] The fresh samples selected were oysters, purchased from Zhenjiang Jimailong Supermarket. All oyster samples were relatively uniform in size, weighing 15-18 g / piece. They were placed in a 4°C refrigerator for 1 day (first group), 3 days (second group), 5 days (third group), 7 days (fourth group), and 9 days (fifth group). They were divided into 5 groups, with 30 samples in each group, for a total of 150 fresh samples. The fresh samples were not subjected to moisture absorption treatment.

[0085] Then, a camera was used to obtain the image of the MSNs@dye nano color-sensitive sensor array before reaction, and the obtained image was saved in a computer.

[0086] Subsequently, each oyster sample and the MSNs@dye nanocolor-sensitive sensor array were placed together in a reaction container, and the prepared MSNs@dye nanocolor-sensitive sensor array was fixed on the top of the reaction container. The MSNs@dye nanocolor-sensitive sensor array was allowed to fully react with the volatile odor substances of the oyster for 10 minutes. Finally, a camera was used to obtain an image of the nanocolor-sensitive sensor array after the reaction, and the obtained image was saved in a computer.

[0087] First, referring to the GB 5009.228-2016 standard, the TVB-N values ​​of 150 fresh samples were determined using the fully automated Kjeldahl method. After the volatile gases of the oyster samples were captured using the MSNs@dye nanocolorimetric sensor array, the fresh samples were crushed, and 10 g of the edible portion of the oysters was taken and placed in a distillation tube. The mixture was mixed with 75 mL of deionized water, shaken, and soaked for 30 minutes. Subsequently, 1 g of magnesium oxide was added to the treated sample. A distiller was immediately connected to the distiller. The receiving solution was 30 mL with a concentration of 20 g / L, and 10 drops of a mixed indicator solution (methyl red ethanol solution: bromocresol green ethanol solution = 1:5) were added. The distillation time was 180 seconds. Finally, the receiving solution was titrated with 0.1 mol / L hydrochloric acid standard titrant. The TVB-N content was calculated and expressed in mg / 100 g.

[0088] Figure 6 are the characteristic images of MSNs@dye nano color-sensitive sensing arrays during different storage periods; Figure 7 is the TVB-N value of oysters at different storage times. Figure 8 is the principal component analysis (PCA) result of the sample information collected by the MSNs@dye nano color sensitive sensor array. Figure 7 It can be seen that as the storage time increases, the TVB-N value of oysters also increases, which is proportional to the freshness of the oysters. Figure 8 The PCA discriminant model for oyster samples at five different storage periods is shown in Figure 1, which is composed of PC1 (42.20%), PC2 (31.61%), and PC3 (8.43%). The first three principal components account for 82.24% of the total. This represents the majority of the variable information. The figure shows that oysters at different storage periods exhibit a clustering trend, particularly those stored for one day, which can be clearly distinguished from the other oysters in the PCA score plot. Oyster samples stored at later stages of storage, while exhibiting some overlap in clustering, still exhibit a clustering trend. This phenomenon can be explained by both the inherent differences among oyster samples and the fact that oyster spoilage is a continuous process.

[0089] Use a computer to locate the position of each color-sensitive unit in the nano-color-sensitive sensor array; decompose the image of the nano-color-sensitive sensor array before and after the reaction into R channel, G channel and B channel grayscale images; subtract the grayscale mean of each sensitive unit before and after the reaction, △R=R a -Rb , △G=G a -G b , △B=B a -B b , △H=H a -H b , △S=S a -S b , △V=V a -V b , △L=L a -L b , △A=A a -A b , △B=B a -B b , where subscript a is the value after reaction, subscript b is the value before reaction, and according to Euclidean distance was calculated; △R, △G, △B, △L, △A, △B, △H, △S, △V, and ED were the characteristic variables of a color-sensitive unit, and 90 characteristic variables were obtained for 9 color-sensitive units. Five storage periods, each containing 30 samples, totaling 150 samples, were analyzed using 90 principal components (PCs) obtained from PCA analysis of these 150 samples. The principal components (PCs) were used as the input of the feature matrix S, and the TVB-N content matrix T corresponding to the samples was used as the output to construct a BPNN model for oyster storage quality evaluation.

[0090]

[0091]

[0092] Figure 11 The correlation scatter plot between TVB-N measured values ​​and predicted values ​​of MSNs@dye nano color-sensitive sensor array is shown. The plus sign and circle represent the training and predicted data, respectively. It can be seen that the TVB-N measured values ​​are highly correlated with the predicted values ​​of MSNs@dye nano color-sensitive sensor array. The root mean square error of cross validation (RMSECV) is 0.2719 and Rc is 0.9971. The root mean square error of the predicted value (RMSEP) is 0.9064 and Rp is 0.9628. Figure 12 The correlation between the measured and predicted values ​​of the color-sensitive material without MSN doping was shown, with an RMSECV value of 1.3877 and an Rc of 0.9318; an RMSEP value of 1.6625 and an Rp of 0.9272. Compared with the color-sensitive material without MSNs doping, the correlation coefficients of the calibration set and the prediction set after doping were significantly improved, and the root mean square errors of the calibration set and the prediction set were significantly reduced. Therefore, after doping with MSNs, the VOC information of oysters with different storage periods collected by the nano color-sensitive sensor array had a high correlation with the TVB-N value measured by the Kjeldahl nitrogen method.

[0093] Evaluation of pork storage quality based on enhanced color sensor array:

[0094] The fresh samples selected were pork tenderloin, purchased from the Zhenjiang fresh food market. All pork tenderloin samples were relatively uniform in size, weighing 10-15 g / slice. They were placed in a 4°C refrigerator for 0 days (first group), 1 day (second group), 3 days (third group), 5 days (fourth group), and 7 days (fifth group). They were divided into 5 groups, with 30 samples in each group, for a total of 150 samples. The samples were not subjected to moisture absorption treatment.

[0095] Then, a camera is used to obtain an image of the sensitized color-sensitive sensor array before reaction, and the obtained image is saved in a computer.

[0096] Subsequently, each pork sample and the sensitized color-sensitive sensor array were placed together in a reaction container, and the prepared sensitized color-sensitive sensor array was fixed on the top of the reaction container. The sensitized color-sensitive sensor array was allowed to fully react with the volatile odor substances in the pork for 10 minutes. Finally, a camera was used to obtain an image of the sensitized color-sensitive sensor array after the reaction, and the obtained image was saved in a computer.

[0097] Figure 13 These are the characteristic images of the color-sensitive sensor array of pork during different storage periods.

[0098] A computer is used to locate the position of each color-sensitive unit in the enhanced color-sensitive sensor array; digital image processing technology is used to quantify the color changes of the sensor chip image, segment its color display area, and extract the characteristic values ​​in the sensor chip. A multivariate nonlinear model of characteristic information is established based on multi-sensor information, sensor signal change slope, correlation signal of color-sensitive materials in the sensor, and multi-sensor correlation signal to distinguish different grades of oolong tea. The image of the enhanced color-sensitive sensor array before and after the reaction is acquired by the camera is decomposed into R channel, G channel, and B channel grayscale images; △R = R a -R b , △G=G a -G b , △B=B a -B b , △H=H a -H b , △S=S a -S b , △V=V a -V b , △L=L a -L b , △A=A a -A b , △B=B a -B b , where subscript a is the value after reaction and b is the value before reaction; and according to Euclidean distance was calculated; △R, △G, △B, △L, △A, △B, △H, △S, △V, and ED were the characteristic variables of a color-sensitive unit, and 90 characteristic variables were obtained for 9 color-sensitive units. Pork tenderloin samples were stored for five periods, with 30 samples in each storage period, for a total of 150 samples. 90 principal components (PCs) were obtained after PCA analysis of these 150 samples as input of the feature matrix S, and the sample-corresponding TVB-N content matrix T was used as output to construct a BPNN model for oyster storage quality evaluation.

[0099]

[0100]

[0101] Figure 14 The PCA discriminant model for pork samples at five different storage periods is shown in Figure 1, which is composed of PC1 (73.77%), PC2 (15.08%), and PC3 (3.60%). The first three principal components account for 92.45% of the total variance. This represents the majority of the variable information. The figure shows a clustering trend among pork samples at different storage periods, particularly pork stored for 0 and 1 day, which can be clearly distinguished from the rest of the pork in the PCA score plot. While there is some overlap in the clustering of pork samples stored at later stages, there is still a clustering trend. This phenomenon can be explained by both the inherent differences among the pork samples and the fact that pork spoilage is a continuous process.

[0102] Figure 15 It is a KNN discriminant model for pork samples of five storage periods. KNN is a method that stores all sample data of the training set under the premise that similar samples are close to each other in the simulation space, and calculates the distance between all unknown samples and the training samples one by one, finds the closest K samples, and thus identifies the pork tenderloin samples to be classified. KNN uses the component vector of PCA as the input vector, pork samples of different storage periods as the output variables of the model, and the recognition rate of pork samples of different storage periods as the discriminant condition of the test model. Three-fifths of all samples are used as training sets, and the remaining samples are used as prediction sets. PC and K values ​​are two factors that have a significant impact on the results of the KNN model, such as Figure 15 As shown in Figure 3, when K is 7 and the number of principal components is 5, the best recognition rates of the calibration set and the prediction set both reach 98.99% and 96.67%.

[0103] Figure 16 This is the LDA model result. When the number of PCs is 12, the recognition rates of the training set and prediction set are 97.78% and 96.67% respectively.

[0104] Figure 17-Figure 19The results of the PLS model are shown in Figure 2. When the number of PCs is 7, the model effect is optimal, with the lowest root mean square error of interaction validation (RMSECV) value of 0.423 and Rp of 95.62%.

[0105] The discrimination results of KNN, LDA and PLS confirmed that the sensitized color-sensitive sensor array constructed in the present invention can detect the quality of pork at different storage periods.

[0106] Those skilled in the art will appreciate that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating the quality of fresh food during storage period based on a modified color-sensitive sensor array, characterized in that: The steps include: S1: Preparation of color-sensitive sensor array: First, a modified color-sensitive dye is obtained by adding MSNs or performing sensitization, and then the modified color-sensitive dye is fixed on a substrate to prepare a MSNs@dye nano color-sensitive sensor array or a sensitized color-sensitive sensor array; The preparation of MSNs@dye nano color-sensitive sensor array is as follows: First, 100 mL of anhydrous ethanol and 300 mL of deionized water were mixed in a 500 mL three-necked flask, and then 6 mL of ammonia water was added. The whole system was continuously stirred at 25 ° C, and then 25 mg of hexadecyltriethylammonium bromide was added. After the solution was mixed evenly, 0.2 g of tetraethoxysilane was added. The reaction time was 10-14 h. Finally, the filtered suspension was calcined in a muffle furnace at 540 ° C for 4-8 h to eliminate the template to obtain MSNs nanomaterials. Then, the MSNs nanomaterial is added to ethanol to obtain an MSNs dispersion, and the MSNs dispersion is mixed with a prepared color-sensitive solution to obtain a mixed solution, wherein the color-sensitive solution is any one of an ethanol solution containing a pH indicator, a dichloromethane solution containing a metal porphyrin, or a dichloromethane solution containing fluoroboryl pyrrole, and polyethylene glycol 600 is added to the mixed solution, and after stirring, an MSNs@dye nano color-sensitive dye is obtained, and the MSNs@dye nano color-sensitive dye is fixed on a substrate to prepare an MSNs@dye nano color-sensitive sensor array, and the nano color-sensitive sensor array is stored in a sealed bag for standby use; The color-sensitive solution is prepared in X quantities, where X is a positive integer. In the color-sensitive solution, the ratio of pH indicator to ethanol is 20 mg:10 mL, the ratio of metal porphyrin to dichloromethane is 20 mg:10 mL, and the ratio of fluoroborane to dichloromethane is 20 mg:10 mL; The concentrations of the color-sensitive solution and the MSNs dispersion are both 2 mg / mL, the color-sensitive solution and the MSNs dispersion are mixed at a volume ratio of 1:5, the amount of polyethylene glycol 600 added is 25% of the volume of the mixed solution, the concentration of TBAH in the ethanol solution is 1 moL / L, the concentration of TBAH in the ethanol solution is 0.8 moL / L, and the amount of MSNs@dye nano color-sensitive dye or sensitized color-sensitive dye fixed on the substrate is 1.5-2 μL; The preparation of the enhanced color sensitive sensor array is specifically as follows: First, an acid-base solution is mixed with a pH indicator to obtain a sensitized solution of the pH indicator, and then the sensitized solution of the pH indicator is mixed with a metal porphyrin solution to obtain a sensitized color-sensitive dye, wherein the acid-base solution is an ethanol solution containing TBAH or TsOH, and then the sensitized color-sensitive dye is spotted on a substrate and placed in a 50-60° C. oven to obtain a sensitized color-sensitive sensor array, and the sensitized color-sensitive sensor array is taken out of the oven and stored in a sealed bag for standby use; S2: Establishment of quality model during fresh storage period: First, fresh samples of different types and different quality levels are selected, wherein the types of fresh samples include oysters and pork, and the different quality levels of samples include the freshness levels of fresh samples of different storage periods; Secondly, use a camera to obtain an image of the color-sensitive sensor array before it reacts, and save the image in a computer; Subsequently, the fresh sample is placed in a reaction container, and the color-sensitive sensor array prepared in step S1 and the fresh sample are stored in the same sealed reaction container, so that the color-sensitive sensor array and the volatile odor substances of the fresh sample are fully reacted. After the reaction, a camera is used to obtain an image of the color-sensitive sensor array after the reaction, and the obtained image is saved in a computer. The position of the color-sensitive unit of the image before and after the reaction is located by the computer, and its color characteristics are extracted. The grayscale mean value of each color-sensitive unit before and after the reaction is subtracted, and the difference in grayscale mean value is the characteristic variable of the color-sensitive unit; Finally, the characteristic variables obtained from all fresh samples are combined to obtain the characteristic matrix. The reverse neural network model for fresh food quality evaluation is constructed with the characteristic matrix as input and the quality grade of the detected fresh samples as output, which is the required fresh food storage period quality model. S3: Rapid evaluation of fresh food quality: The sample to be tested is reacted according to the method described in step S2 to obtain the characteristic variables of the sample to be tested, and the characteristic variables of the sample to be tested are substituted into the fresh storage period quality model constructed in step S2 to obtain the quality grade information of the sample to be tested and complete the rapid evaluation of the fresh quality.

2. The method for evaluating the quality of fresh produce during storage period based on the modified color-sensitive sensor array according to claim 1, characterized in that: When the mixed solution and polyethylene glycol 600 were stirred, the stirring speed was 350 r / min, the stirring time was 10 min, and the temperature was increased to 90° C. at a rate of 5° C. / 5 min and maintained for 60 min.

3. The method for evaluating the quality of fresh produce during storage period based on the modified color-sensitive sensor array according to claim 1, characterized in that: In step S2, when the fresh sample is oyster, the amount used is 15-18 g, and when the fresh sample is pork, the amount used is 10-15 g; the detection time of the color-sensitive sensor array is 10-20 min.

4. The method for evaluating the quality of fresh produce during storage period based on the modified color-sensitive sensor array according to claim 1, characterized in that: In step S2, the steps of extracting characteristic variables are as follows: first, using a computer to locate the position of each color-sensitive unit in the color-sensitive sensor array; then, decomposing the image of the color-sensitive sensor array before and after the reaction acquired by the camera into R channel, G channel and B channel grayscale images; then, subtracting the grayscale mean of each sensitive unit before and after the reaction, that is, △R=R a -R b , △G=G a -G b , △B=B a -B b , where subscript a is the value after reaction, subscript b is the value before reaction, and according to ED= Calculate the Euclidean distance. △R, △G, △B and ED are the characteristic variables of a color-sensitive unit. X color-sensitive units obtain Y characteristic variables, where Y=4X; The number of fresh samples used to construct the fresh storage quality evaluation model is selected to be N, including n quality grades, and m samples for each grade, that is, N=n*m, wherein n is a positive integer not less than 2, and m and N are both positive integers; The steps of establishing the fresh storage period quality evaluation model are as follows: the characteristic matrix is ​​recorded as S, S is N*Y, the characteristic variable matrix S is used as the input value, and the quality grade matrix T corresponding to the fresh sample is used as the output value to construct a reverse neural network model for fresh storage period quality evaluation; 。

Images