Method for identifying lichee variety or lichee producing area based on electronic nose and electronic tongue

Abstract

The invention belongs to the technical field of food detection, and provides a method for identifying lichee varieties or lichee producing areas based on an electronic nose and an electronic tongue. According to the method, the electronic nose and the electronic tongue are adopted to measure litchi samples of different varieties and different producing areas, and the response values of the electronic nose and the electronic tongue are taken as variables to perform principal component analysis, so that different varieties of litchis can be identified and distinguished, and the same variety of litchis of different producing areas can also be distinguished. Besides, the response values measured by the electronic nose and the electronic tongue and the physicochemical index content of the sample are used for constructing a model by using a partial least square regression method, so that the physicochemical indexes related to the litchi flavor can be rapidly measured by using the electronic nose and the electronic tongue, the operation is simple, the accuracy is high, and the time and the cost for measuring the physicochemical indexes are greatly reduced; the litchi variety can be further identified. The invention provides a method for identifying the variety and the production place of the litchi, provides a scheme for rapidly determining the physicochemical indexes based on the electronic nose and the electronic tongue, and has good application prospect and value.

Description

Technical Field

[0001] The present invention belongs to the technical field of food detection. More specifically, it relates to a method for identifying litchi varieties or litchi origins based on an electronic nose and an electronic tongue. Background Art

[0002] Litchi (Litchi chinensis Sonn.) belongs to the genus Litchi in the family Sapindaceae and is the fruit of an evergreen arbor. Litchi originated in southern China and has a cultivation history of more than 2300 years. It is a famous and special fruit in South China and one of the famous fruits in Lingnan. It has important economic and cultural values. The pulp of litchi is sweet and delicious and is deeply loved by consumers. The variety and origin of litchi are important factors affecting the quality and taste of litchi. There are many types of litchi, and there are many litchi varieties. The main varieties include Feizixiao, Guiwei, Nuomici, Heiye, Huaizhi, Baitangying, etc. Each litchi has its own unique characteristics. Feizixiao has bright colors and a unique taste that combines sweetness and slight sourness; Guiwei has a slight osmanthus fragrance and its flesh is crispy; Nuomici is soft, glutinous and sweet. In addition, research has found that due to the unique natural environment and cultivation techniques in some origins, unique litchi flavors have been formed. For example, the "Guanyinlv" litchi in Dongguan, Guangdong is famous for its sweet taste and unique aroma when its peel is greenish-yellow, which is closely related to the microclimate conditions in its planting area; Zengcheng Guailv, due to its strict requirements for a specific low-temperature environment, its unique flavor is difficult to be replicated by other production areas; the high sweetness and creamy flesh of the "Nuomici" variety in Dongguan are inseparable from its original cultivation environment. Since the flavor quality and taste of different litchi varieties are significantly different, there are also large differences in price. Many litchis are very similar in appearance and smell, which makes it particularly difficult to distinguish them, resulting in the phenomenon of selling litchi varieties with lower prices under the guise of those with higher prices in the market. Therefore, it is necessary to identify litchis of different varieties and origins.

[0003] The traditional method for determining litchi varieties is mainly through manual identification. The appraiser distinguishes different litchi varieties through smell or taste. However, its sensory analysis has strong subjectivity, poor repeatability, limited sensitivity and recognition, and cannot distinguish litchi varieties and origins on a large scale, so it is not suitable for actual application scenarios. At present, there are few methods for identifying different litchis and constructing models. Chinese Patent CN116645665A discloses a method for litchi variety identification and classification based on Spark and deep learning. By constructing a Resnet34_CBAM model and combining distributed computing technology, rapid classification of a large number of litchi images is realized. This method mainly uses the RGB image data of different litchis for distinction, requires high hardware resources, has high requirements for the generalization ability of the model, and has certain limitations for image identification of litchis with different maturities.

[0004] Currently, the sales mode of litchi mainly focuses on fresh fruit sales, and the proportion of processed litchi in the total output is 13%. The mature period of litchi is concentrated in the hot and humid summer, which makes litchi extremely prone to browning and has a fast decay rate, resulting in a short shelf life. Processing litchi is an effective way to extend its shelf life and meet market demand. When selecting litchi as a processing raw material, its aroma and taste are important indicators for evaluating its quality. At present, in addition to artificial sensory evaluation, the most common detection methods for food aroma are HPLC, HPLC-□MS, GC, GC-□MS and other technologies, which conduct qualitative and quantitative research on single components or a class of flavor components in the sample. The sample pretreatment is complex, the analysis process is time-consuming and laborious, and it is difficult to meet the actual needs of the litchi processing industry. The detection of food taste mainly relies on artificial sensory evaluation and the detection of related physical and chemical indicators, such as the determination of sugar-acid ratio, sugar components, organic acids, amino acid types and contents. The measurement of physical and chemical indicators usually requires relatively complex pretreatment, a long detection time, and complex operations. The results of artificial sensory evaluation of flavor are affected by subjective factors, lack repeatability, and are difficult to meet the large-scale and automated trends in actual production. Moreover, there are interactions between flavors, such as the effects of taste masking, modification, and contrast, which cannot fully, objectively, and comprehensively reflect the taste characteristics of the sample. How to batch and quickly characterize the overall quality of litchi juice in actual production has become a difficult problem in the litchi processing field.

[0005] In summary, there is an urgent need for a method that can simultaneously identify litchi varieties and origins to address quality problems caused by the confusion of varieties and origins in the market and provide more reliable quality assurance for consumers. In addition, developing a method for batch and rapid identification of the physical and chemical indicators of litchi to achieve the identification of the overall quality of litchi flavor quality and litchi juice is also an urgent problem to be solved in the current litchi processing industry and is of great significance in the litchi processing field. Summary of the Invention

[0006] The present invention aims to solve the current lack of technologies for rapid identification of litchi varieties and origins, as well as the lack of methods for batch and rapid identification of the physical and chemical indicators of litchi, and provides a method for identifying litchi varieties or litchi origins and detecting the physical and chemical indicators of litchi based on an electronic nose and an electronic tongue.

[0007] The first object of the present invention is to provide a method for identifying litchi varieties or litchi origins based on an electronic nose and an electronic tongue.

[0008] The second object of the present invention is to provide a method for rapidly detecting the physical and chemical indicators related to litchi flavor based on an electronic nose and an electronic tongue.

[0009] The above objects of the present invention are achieved by the following technical solutions:

[0010] The present invention provides a method for identifying litchi varieties or litchi origins based on an electronic nose and an electronic tongue, comprising the following steps:

[0011] S1. Use an electronic tongue to detect litchi samples of different varieties or origins to obtain response value data of the electronic tongue sensors; use an electronic nose to detect the volatile gases of litchi samples of different varieties or origins to obtain response value data of the electronic nose sensors;

[0012] S2. Perform principal component analysis on the sensor response value data of the electronic nose and the electronic tongue obtained in step S1;

[0013] S3. Take the litchi sample to be tested for electronic tongue detection and electronic nose detection to obtain the response value data of the litchi sample to be tested, and compare the response value of the litchi sample to be tested with the response values of litchi samples of different varieties or origins in step S2 to determine the variety or origin of the sample to be tested;

[0014] The electronic nose is a PEN3 type electronic nose;

[0015] The electrodes of the electronic tongue include one or more of platinum, gold, palladium, tungsten, titanium, and silver.

[0016] As an alternative embodiment, in step S1, put litchi samples of different varieties or origins into a headspace bottle, enrich the volatile gases in the headspace, and use an electronic nose to detect the litchi samples of different varieties or origins to obtain response value data of the electronic nose sensors.

[0017] As an alternative embodiment, the electrodes of the above-mentioned electronic tongue include platinum, gold, palladium, tungsten, titanium, and silver, auxiliary electrodes, and reference electrodes, and the reference electrode is an Ag / AgCl electrode.

[0018] As an alternative embodiment, in step S1, before the electronic tongue detection, activate the electronic tongue sensors with a KCl solution, then wash the sensors with water for 5 - 15 s, and then start the detection.

[0019] As an alternative embodiment, the concentration of the above-mentioned KCl solution is 0.01 - 0.02 mol / L.

[0020] As an alternative embodiment, the sensors of the electronic nose include W1C, W5S, W3C, W6S, W5C, W1S, W1W, W2S, W2W, and W3S.

[0021] Specifically, the W1C is sensitive to aromatic components and benzene; the W5S is sensitive to nitrogen oxides, the W3C is sensitive to ammonia and aromatic components, the W6S is selective to hydrides, the W5C is sensitive to short-chain alkane aromatic components, the W1S is sensitive to methyl groups, the W1W is sensitive to sulfides, the W2S is sensitive to alcohols and aldehyde-ketones, the W2W is sensitive to aromatic components and organic sulfides, and the W3S is sensitive to long-chain alkanes.

[0022] As an alternative embodiment, the detection conditions of the electronic nose are as follows: the cleaning time is 90 - 150 s, the zeroing time is 2.5 - 10 s, the sample preparation time is 5 - 20 s, the injection flow rate is 100 - 200 mL / min, and the detection time is 60 - 150 s.

[0023] As an alternative embodiment, the detection conditions of the electronic nose are as follows: the cleaning time is 120 s, the zeroing time is 5 s, the sample preparation time is 5 s, the injection flow rate is 150 mL / min, and the detection time is 120 s.

[0024] As an alternative embodiment, the sensor amplification factor of the electronic tongue is set to 80 - 120.

[0025] As an alternative embodiment, the sensor amplification factor of the electronic tongue is set to 100.

[0026] As an alternative embodiment, the response value is the average response value at 89 - 91 s.

[0027] As an alternative embodiment, the response value is the average response value at 90 s.

[0028] As an alternative embodiment, the litchi sample is litchi juice obtained by pulping litchi pulp and filtering.

[0029] As an alternative embodiment, the litchi sample is litchi juice obtained by pulping litchi pulp and filtering with a 100 - 300 mesh (preferably 200 mesh) gauze.

[0030] Specifically, the litchi varieties are Xiantao Litchi, Giant Beauty, Qiangang Glutinous Rice Cakes, Jinggang Red Glutinous, Seedless Litchi, and Huaizhi.

[0031] As an alternative specific implementation, take the litchi sample to be tested for electronic nose detection. When the response value of the electronic nose W1C sensor is 3.51 - 3.82, the response value of the W5S sensor is 2.32 - 2.62, the response value of the W3C sensor is 0.92 - 0.96, the response value of the W6S sensor is 0.99 - 1.45, the response value of the W5C sensor is 2.95 - 3.26, the response value of the W1S sensor is 5.10 - 5.78, the response value of the W1W sensor is 2.13 - 2.65, the response value of the W2S sensor is 7.56 - 7.92, the response value of the W2W sensor is 1.68 - 1.95, and the response value of the W3S sensor is 1.56 - 1.87, determine that the litchi variety is Wuheli;

[0032] When the response value of the electronic nose W1C sensor is 2.11 - 2.74, the response value of the W5S sensor is 2.52 - 2.81, the response value of the W3C sensor is 0.91 - 1.13, the response value of the W6S sensor is 1.02 - 1.32, the response value of the W5C sensor is 1.59 - 1.89, the response value of the W1S sensor is 2.56 - 2.87, the response value of the W1W sensor is 2.01 - 2.35, the response value of the W2S sensor is 4.35 - 4.67, the response value of the W2W sensor is 1.56 - 1.85, and the response value of the W3S sensor is 1.43 - 1.68, determine that the litchi variety is Jumeiren;

[0033] When the response value of the electronic nose W1C sensor is 3.95 - 4.21, the response value of the W5S sensor is 2.13 - 2.65, the response value of the W3C sensor is 0.91 - 0.97, the response value of the W6S sensor is 1.03 - 1.33, the response value of the W5C sensor is 3.01 - 3.35, the response value of the W1S sensor is 5.45 - 5.68, the response value of the W1W sensor is 2.10 - 2.35, the response value of the W2S sensor is 9.85 - 11.62, the response value of the W2W sensor is 1.58 - 1.89, and the response value of the W3S sensor is 1.30 - 1.92, determine that the litchi variety is Xiantao litchi;

[0034] When the response value of the electronic nose W1C sensor is 6.22 - 6.64, the response value of the W5S sensor is 20.51 - 26.85, the response value of the W3C sensor is 0.91 - 0.96, the response value of the W6S sensor is 1.06 - 1.98, the response value of the W5C sensor is 4.95 - 5.32, the response value of the W1S sensor is 9.32 - 10.69, the response value of the W1W sensor is 6.23 - 6.89, the response value of the W2S sensor is 9.49 - 9.68, the response value of the W2W sensor is 4.00 - 4.36, and the response value of the W3S sensor is 2.53 - 2.86, determine that the litchi variety is Huaizhi;

[0035] When the response values of the electronic nose W1C sensor are from 5.03 to 5.64, the response value of the W5S sensor is 19.3050, the response values of the W3C sensor are from 0.91 to 0.96, the response values of the W6S sensor are from 0.96 to 1.36, the response values of the W5C sensor are from 3.86 to 4.35, the response values of the W1S sensor are from 8.64 to 8.98, the response values of the W1W sensor are from 5.98 to 6.65, the response values of the W2S sensor are from 11.98 to 12.53, the response values of the W2W sensor are from 4.01 to 4.56, and the response values of the W3S sensor are from 1.48 to 1.92, it is determined that the litchi variety is Qiangang Nuomici;

[0036] When the response values of the electronic nose W1C sensor are from 4.78 to 5.32, the response values of the W5S sensor are from 28.10 to 29.12, the response values of the W3C sensor are from 0.91 to 0.98, the response values of the W6S sensor are from 0.98 to 1.38, the response values of the W5C sensor are from 3.12 to 4.32, the response values of the W1S sensor are from 7.03 to 7.86, the response values of the W1W sensor are from 5.86 to 6.33, the response values of the W2S sensor are from 12.91 to 13.68, the response values of the W2W sensor are from 3.10 to 3.98, and the response values of the W3S sensor are from 0.92 to 1.53, it is determined that the litchi variety is Jinggang Hongnuo.

[0037] As an alternative specific implementation, take the litchi sample to be tested for electronic tongue detection. When the response values of the electronic tongue P1 sensor are from 0.33 to 0.39, the response values of the P2 sensor are from 0.32 to 0.38, the response values of the P3 sensor are from 0.07 to 1.11, the response values of the P4 sensor are from 0.46 to 0.49, the response values of the P5 sensor are from 0.45 to 0.50, the response values of the P6 sensor are from 0.07 to 0.11, the response values of the P7 sensor are from 0.47 to 0.52, the response values of the P8 sensor are from 0.46 to 0.49, the response values of the P9 sensor are from 0.07 to 0.11, the response values of the P10 sensor are from 0.49 to 0.54, the response values of the P11 sensor are from 0.48 to 0.52, the response values of the P12 sensor are from 0.10 to 0.16, the response values of the P13 sensor are from 0.47 to 0.51, the response values of the P14 sensor are from 0.42 to 0.48, the response values of the P15 sensor are from 0.19 to 0.26, the response values of the P16 sensor are from 0.47 to 0.52, the response values of the P17 sensor are from 0.46 to 0.51, and the response values of the P18 sensor are from 0.09 to 0.15, it is determined that the litchi variety is Seedless Litchi;

[0038] When the response value of the electronic tongue P1 sensor is 0.37 - 0.39, the response value of the P2 sensor is 0.36 - 0.41, the response value of the P3 sensor is 0.08 - 0.15, the response value of the P4 sensor is 0.51 - 0.72, the response value of the P5 sensor is 0.51 - 0.62, the response value of the P6 sensor is 0.07 - 0.13, the response value of the P7 sensor is 0.50 - 0.66, the response value of the P8 sensor is 0.50 - 0.71, the response value of the P9 sensor is 0.07 - 0.14, the response value of the P10 sensor is 0.53 - 0.69, the response value of the P11 sensor is 0.53 - 0.62, the response value of the P12 sensor is 0.10 - 0.16, the response value of the P13 sensor is 0.51 - 0.59, the response value of the P14 sensor is 0.46 - 0.51, the response value of the P15 sensor is 0.18 - 0.22, and the response value of the P16 sensor is 0.52 - 0.61, and the response value of the P17 sensor is 0.51 - 0.56, and the response value of the P18 sensor is 0.09 - 0.12, it is determined that the litchi variety is Giant Beauty;

[0039] When the response value of the electronic tongue P1 sensor is 0.39 - 0.42, the response value of the P2 sensor is 0.38 - 0.44, the response value of the P3 sensor is 0.08 - 0.13, the response value of the P4 sensor is 0.52 - 0.62, the response value of the P5 sensor is 0.52 - 0.31, the response value of the P6 sensor is 0.07 - 0.12, the response value of the P7 sensor is 0.51 - 0.59, the response value of the P8 sensor is 0.51 - 0.61, the response value of the P9 sensor is 0.07 - 0.11, the response value of the P10 sensor is 0.55 - 0.59, the response value of the P11 sensor is 0.55 - 0.59, the response value of the P12 sensor is 0.10 - 0.16, the response value of the P13 sensor is 0.52 - 0.61, the response value of the P14 sensor is 0.48 - 0.52, the response value of the P15 sensor is 0.19 - 0.22, the response value of the P16 sensor is 0.55 - 0.61, the response value of the P17 sensor is 0.52 - 0.59, and the response value of the P18 sensor is 0.09 - 0.16, it is determined that the litchi variety is Xiantao Litchi;

[0040] When the response value of the electronic tongue P1 sensor is 0.41 - 0.45, the response value of the P2 sensor is 0.41 - 0.49, the response value of the P3 sensor is 0.08 - 0.12, the response value of the P4 sensor is 0.54 - 0.62, the response value of the P5 sensor is 0.54 - 0.59, the response value of the P6 sensor is 0.07 - 0.11, the response value of the P7 sensor is 0.53 - 0.59, the response value of the P8 sensor is 0.53 - 0.58, the response value of the P9 sensor is 0.07 - 0.09, the response value of the P10 sensor is 0.57 - 0.62, the response value of the P11 sensor is 0.57 - 0.59, the response value of the P12 sensor is 0.09 - 0.14, the response value of the P13 sensor is 0.56 - 0.62, the response value of the P14 sensor is 0.51 - 0.56, the response value of the P15 sensor is 0.19 - 0.24, and the response value of the P16 sensor is 0.57 - 0.64, and the response value of the P17 sensor is 0.56 - 0.59, and the response value of the P18 sensor is 0.09 - 0.13, it is determined that the litchi variety is Huaizhi;

[0041] When the response value of the electronic tongue P1 sensor is 0.42 - 0.46, the response value of the P2 sensor is 0.42 - 0.48, the response value of the P3 sensor is 0.09 - 0.12, the response value of the P4 sensor is 0.55 - 0.58, the response value of the P5 sensor is 0.55 - 0.61, the response value of the P6 sensor is 0.07 - 0.11, the response value of the P7 sensor is 0.52 - 0.59, the response value of the P8 sensor is 0.52 - 0.58, the response value of the P9 sensor is 0.07 - 0.12, the response value of the P10 sensor is 0.57 - 0.61, the response value of the P11 sensor is 0.57 - 0.59, the response value of the P12 sensor is 0.09 - 0.13, the response value of the P13 sensor is 0.56 - 0.63, the response value of the P14 sensor is 0.51 - 0.61, the response value of the P15 sensor is 0.20 - 0.31, the response value of the P16 sensor is 0.58 - 0.68, the response value of the P17 sensor is 0.57 - 0.66, and the response value of the P18 sensor is 0.09 - 0.23, it is determined that the litchi variety is Qiangang Nuomici;

[0042] When the response values of the electronic tongue P1 sensor are between 0.39 and 0.43, the response values of the P2 sensor are between 0.39 and 0.45, the response values of the P3 sensor are between 0.09 and 0.11, the response values of the P4 sensor are between 0.52 and 0.31, the response values of the P5 sensor are between 0.52 and 0.57, the response values of the P6 sensor are between 0.07 and 0.09, the response values of the P7 sensor are between 0.50 and 0.58, the response values of the P8 sensor are between 0.50 and 0.53, the response values of the P9 sensor are between 0.07 and 0.21, the response values of the P10 sensor are between 0.55 and 0.61, the response values of the P11 sensor are between 0.54 and 0.63, the response values of the P12 sensor are between 0.09 and 0.21, the response values of the P13 sensor are between 0.52 and 0.61, the response values of the P14 sensor are between 0.47 and 0.56, the response values of the P15 sensor are between 0.21 and 0.32, the response values of the P16 sensor are between 0.55 and 0.62, and the response values of the P17 sensor are between 0.54 and 0.62, and the response values of the P18 sensor are between 0.09 and 0.12, it is determined that the litchi variety is Jingganghongnuo.

[0043] Specifically, the litchis from different origins are the Feizixiao litchi varieties from different provinces, including Feizixiao from Hainan Province, Feizixiao from Yunnan Province, and Feizixiao from Guangdong Province.

[0044] As an alternative specific implementation, take the litchi sample to be tested for electronic nose and electronic tongue detection. (P1-18 are electronic tongue sensors) When the response value of the electronic nose W1C is 6.02 - 6.62, the response value of the W5S sensor is 6.93 - 7.35, the response value of the W3C sensor is 0.87 - 1.53, the response value of the W6S sensor is 1.22 - 2.31, the response value of the W5C sensor is 4.93 - 5.31, the response value of the W1S sensor is 8.24 - 8.89, the response value of the W1W sensor is 5.62 - 6.32, the response value of the W2S sensor is 17.07 - 18.32, the response value of the W2W sensor is 2.83 - 3.21, the response value of the W3S sensor is 2.28 - 2.98, the response value of the electronic tongue P1 sensor is 0.43 - 1.35, the response value of the P2 sensor is 0.43 - 1.23, the response value of the P3 sensor is 0.09 - 0.19, the response value of the P4 sensor is 0.55 - 1.22, the response value of the P5 sensor is 0.55 - 1.20, the response value of the P6 sensor is 0.08 - 0.21, the response value of the P7 sensor is 0.50 - 0.32, the response value of the P8 sensor is 0.50 - 0.92, the response value of the P9 sensor is 0.07 - 0.25, the response value of the P10 sensor is 0.62 - 0.86, the response value of the P11 sensor is 0.61 - 1.12, the response value of the P12 sensor is 0.12 - 0.53, the response value of the P13 sensor is 0.59 - 0.95, the response value of the P14 sensor is 0.54 - 0.69, the response value of the P15 sensor is 0.20 - 0.53, the response value of the P16 sensor is 0.60 - 1.3, the response value of the P17 sensor is 0.58 - 0.75, and the response value of the P18 sensor is 0.10 - 0.42, it is determined to be the Feizixiao litchi variety in Hainan Province;

[0045] When the response value of the electronic nose W1C is 6.21 - 7.11, the response value of the W5S sensor is 18.29 - 19.56, the response value of the W3C sensor is 0.87 - 1.63, the response value of the W6S sensor is 1.20 - 2.41, the response value of the W5C sensor is 5.04 - 6.13, the response value of the W1S sensor is 7.72 - 8.23, the response value of the W1W sensor is 7.71 - 8.01, the response value of the W2S sensor is 17.78 - 18.95, the response value of the W2W sensor is 3.12 - 4.20, the response value of the W3S sensor is 2.31 - 2.41, the response value of the electronic tongue P1 sensor is 0.43 - 0.98, the response value of the P2 sensor is 0.43 - 0.87, the response value of the P3 sensor is 0.09 - 1.53, the response value of the P4 sensor is 0.54 - 1.42, the response value of the P5 sensor is 0.54 - 0.95, the response value of the P6 sensor is 0.07 - 1.84, the response value of the P7 sensor is 0.50 - 1.64, the response value of the P8 sensor is 0.50 - 1.69, the response value of the P9 sensor is 0.07 - 1.24, the response value of the P10 sensor is 0.63 - 0.78, the response value of the P11 sensor is 0.62 - 1.15, the response value of the P12 sensor is 0.11 - 0.35, the response value of the P13 sensor is 0.60 - 0.94, the response value of the P14 sensor is 0.55 - 0.84, the response value of the P15 sensor is 0.19 - 0.48, the response value of the P16 sensor is 0.60 - 1.35, the response value of the P17 sensor is 0.57 - 0.94, the response value of the P18 sensor is 0.09 - 0.22, it is determined as the Feizixiao litchi variety in Yunnan Province;

[0046] When the response value of the electronic nose W1C is 3.94 - 4.52, the response value of the W5S sensor is 35.36 - 36.45, the response value of the W3C sensor is 0.88 - 1.32, the response value of the W6S sensor is 1.08 - 2.53, the response value of the W5C sensor is 3.11 - 4.52, the response value of the W1S sensor is 5.91 - 6.87, the response value of the W1W sensor is 10.80 - 11.45, the response value of the W2S sensor is 9.54 - 10.58, the response value of the W2W sensor is 4.35 - 5.34, the response value of the W3S sensor is 1.90 - 2.65, the response value of the electronic tongue P1 sensor is 0.39 - 0.79, the response value of the P2 sensor is 0.39 - 0.98, the response value of the P3 sensor is 0.09 - 0.46, the response value of the P4 sensor is 0.53 - 0.87, the response value of the P5 sensor is 0.53 - 1.12, the response value of the P6 sensor is 0.07 - 0.96, the response value of the P7 sensor is 0.44 - 1.25, the response value of the P8 sensor is 0.44 - 1.36, the response value of the P9 sensor is 0.06 - 0.31, the response value of the P10 sensor is 0.63 - 0.95, the response value of the P11 sensor is 0.62 - 1.22, the response value of the P12 sensor is 0.11 - 0.51, the response value of the P13 sensor is 0.61 - 1.23, the response value of the P14 sensor is 0.55 - 0.94, the response value of the P15 sensor is 0.18 - 0.48, the response value of the P16 sensor is 0.59 - 1.41, the response value of the P17 sensor is 0.57 - 0.97, and the response value of the P18 sensor is 0.09 - 0.14, it is determined as the Feizixiao litchi variety in Guangdong Province.

[0047] The present invention also provides a method for rapidly detecting flavor-related physical and chemical indexes of litchi based on an electronic nose and an electronic tongue, comprising the following steps:

[0048] (1) Detect the litchi sample with the electronic tongue to obtain the response value data of the electronic tongue sensor; detect the volatile gas of the litchi sample with the electronic nose to obtain the response value data of the electronic nose sensor; measure the content of flavor-related physical and chemical indexes of the litchi sample;

[0049] (2) Use the response value and the data of the content of flavor-related physical and chemical indexes to establish a partial least squares regression model of the response value and the content of flavor-related physical and chemical indexes;

[0050] (3) Take the litchi sample to be tested for detection with the electronic nose and the electronic tongue to obtain the response value of the litchi sample to be tested, and substitute the response value into the model in step (2), then the content of flavor-related physical and chemical indexes of the litchi sample to be tested can be obtained;

[0051] The flavor-related physical and chemical indexes are sugar-acid ratio, sugar components, organic acids, and characteristic volatile substances.

[0052] Specifically, the sugar component is the sum of fructose, glucose, and sucrose; the organic acids are the sum of tartaric acid, malic acid, and citric acid.

[0053] As an alternative embodiment, the characteristic volatile substances are those obtained by screening the volatile components of the sample measured by gas chromatography-mass spectrometry using orthogonal partial least squares discriminant analysis.

[0054] As an alternative embodiment, the variety of litchi samples measured in step (1) is the same as that of the litchi samples to be tested in step (3).

[0055] As an alternative embodiment, the characteristic volatile substances are those obtained by detecting the volatile substances contained in litchi by using a gas chromatography-mass spectrometer after headspace solid-phase microextraction of the litchi samples, and then screening through OPLS-DA (orthogonal partial least squares discriminant analysis) with the criteria of VIP > 1 and a statistically significant two-tailed t-test (P < 0.05).

[0056] As an alternative specific embodiment, the formula for the physical and chemical indexes of the model in step (2) is:

[0057] Sugar-acid ratio content = -4.1730×W1C - 0.9771×W5S + 0.3123×W3C + 1.8554×W6S - 4.1367×W5C + 13.2519×W1S + 6.2793×W1W - 2.007×W2S - 10.0370×W2S + 14.9546×W3S + 5.1563×P1 + 5.1863×P2 - 0.0409×P3 + 7.1273×P4 + 7.0794×P5 + 0.2125×P6 + 5.1164×P7 + 5.0956×P8 - 0.3390×P9 + 4.7192×P10 + 4.3903×P11 + 2.5084×P12 + 5.3195×P13 + 4.4933×P14 + 1.5606×P15 + 6.2368×P16 + 5.6657×P17 + 1.8505×P18.

[0058] Total sugar content = 1.9418×W1C + 0.2468×W5S - 0.0410×W3C - 0.1967×W6S + 1.1886×W5C - 1.9771×W1S - 1.5198×W1W - 0.0262×W2S + 1.2060×W2W - 0.3461×W3S - 0.1383×P1 - 0.1427×P2 + 0.0108×P3 - 0.2469×P4 - 0.2464×P5 - 0.0361×P6 - 0.1048×P7 - 0.1055×P8 + 0.0048×P9 - 0.1180×P10 - 0.0911×P11 - 0.1456×P12 - 0.1061×P13 - 0.0800×P14 - 0.0743×P15 - 0.2130×P16 - 0.2089×P17 - 0.1341×P18。

[0059] Total acid content = -0.9117×W1C + 1.5049×W5S + 0.0001×W3C - 0.0646×W6S - 0.7493×W5C - 1.0279×W1S + 0.1831×W1W - 3.2316×W2S + 0.0821×W2W - 0.2200×W3S - 0.0038×P1 - 0.0038×P2 + 0.0004×P3 + 0.0080×P4 + 0.0073×P5 - 0.0002×P6 - 0.0138×P7 - 0.0136×P8 - 0.0018×P9 + 0.0088×P10 + 0.0091×P11 + 0.0026×P12 + 0.0146×P13 + 0.0129×P14 - 0.0047×P15 + 0.0051×P16 + 0.0052×P17 - 0.0011×P18。

[0060] Content of characteristic volatile substances = 9.0383×W1C + 1.9907×W5S - 0.6472×W3C - 3.1953×W6S + 9.9741×W5C - 25.1167×W1S - 12.8103×W1W + 3.6136×W2S + 18.7819×W2W - 28.8936×W3S - 9.4119×P1 - 9.4882×P2 + 0.0723×P3 - 13.4475×P4 - 13.3468×P5 - 0.3774×P6 - 9.6659×P7 - 9.6275×P8 + 0.6293×P9 - 8.8184×P10 - 8.1545×P11 - 4.3461×P12 - 9.9008×P13 - 8.3071×P14 - 3.0778×P15 - 11.8370×P16 - 10.7503×P17 - 3.5783×P18。

[0061] The present invention has the following beneficial effects:

[0062] The present invention uses an electronic nose and an electronic tongue to measure different litchi samples including different varieties and different origins, and uses the response values ​​of 10 sensors of the electronic nose and the response values ​​of 6 inert electrodes of the electronic tongue as variables to perform principal component analysis, which can not only identify and distinguish different varieties of litchi, but also distinguish the same variety of litchi from different origins. In addition, by using the partial least squares regression method to construct a model based on the response value and the content of the sample's physical and chemical indexes, the electronic nose and the electronic tongue are used to quickly measure the physical and chemical indexes related to the flavor of litchi, which is simple to operate and has high accuracy, greatly simplifies the time and cost of measuring the physical and chemical indexes, and can further identify the litchi variety. The present invention provides a method for identifying litchi varieties and origins, and provides a solution for quickly measuring physical and chemical indexes based on the electronic nose and the electronic tongue, which has good application prospects and value. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 These are the electronic tongue response value graphs of different varieties of litchi (A is the seedless litchi, B is the giant beauty, C is the fairy peach litchi, D is the Huaizhi, E is the Qiangang glutinous rice cake, and F is the Jinggang red glutinous rice cake).

[0064] Figure 2 The principal component analysis results of different litchi varieties based on electronic tongue signals.

[0065] Figure 3 The electronic nose analysis results of different varieties of litchi (Figure A is the radar chart of electronic nose analysis of different varieties of litchi, and Figure B is the PCA analysis result of the electronic nose sensor signal).

[0066] Figure 4 The principal component analysis results of different litchi varieties based on combined signals.

[0067] Figure 5 Modeling results of sensor signals and sugar-acid ratio (Figure A is the modeling result of electronic tongue sensor signals, and Figure B is the modeling result of electronic nose combined with electronic tongue sensor signals).

[0068] Figure 6 The modeling results of sensor signals and total sugar (Figure A is the modeling result of electronic tongue sensor signals, and Figure B is the modeling result of electronic nose combined with electronic tongue sensor signals).

[0069] Figure 7 The modeling results of sensor signals and total acid (Figure A is the modeling result of electronic tongue sensor signals, and Figure B is the modeling result of electronic nose combined with electronic tongue sensor signals).

[0070] Figure 8Modeling results of sensor signals and volatile components (Figure A shows the modeling results of electronic nose sensor signals, and Figure B shows the modeling results of electronic nose combined with electronic tongue sensor signals).

[0071] Figure 9 PCA analysis results of Feizixiao from different origins based on combined signals.

[0072] Figure 10 Evaluation results of the quantitative prediction model for the sugar-acid ratio content based on combined signals (Figure A shows the results of the training set, and Figure B shows the results of the prediction set).

[0073] Figure 11 Evaluation results of the quantitative prediction model for the total sugar content based on combined signals (Figure A shows the results of the training set, and Figure B shows the results of the prediction set).

[0074] Figure 12 Evaluation results of the quantitative prediction model for the total acid content based on combined signals (Figure A shows the results of the training set, and Figure B shows the results of the prediction set).

[0075] Figure 13 Evaluation results of the quantitative prediction model for the content of characteristic volatile components based on combined signals (Figure A shows the results of the training set, and Figure B shows the results of the prediction set). Specific implementation manners

[0076] The following further illustrates the present invention in conjunction with the accompanying drawings of the specification and specific embodiments, but the embodiments do not limit the present invention in any form. Unless otherwise specified, the reagents, methods, and equipment used in the present invention are conventional reagents, methods, and equipment in the technical field.

[0077] Unless otherwise specified, the reagents and materials used in the following examples are all commercially available.

[0078] In the following examples, the litchis used for experiments were all taken from the Guangdong Province, and the varieties were Xiantao Li, Jumeiren, Qiangang Nuomici, Jinggang Hongnuo, Seedless Li, and Huaizhi. Among them, Xiantao Li was taken from Xuwen County, Zhanjiang City, Jumeiren was taken from Yangxi County, Yangjiang City, and Qiangang Nuomici, Jinggang Hongnuo, Seedless Li, and Huaizhi were all taken from Conghua District, Guangzhou City. After picking, they were quickly transported to the laboratory and stored in a -80°C refrigerator.

[0079] In the following examples, the Folin phenol reagent was purchased from San Land Chemical Company.

[0080] Fumaric acid (HPLC≥98%), malic acid (HPLC≥98%), and DL-tartaric acid (HPLC≥98%) were purchased from Shanghai Yuanye Bio-Technology Co., Ltd., and the product numbers were B20404, B20937, and B25642 respectively.

[0081] In the following examples, the electronic tongue (Isenso, instrument model Supertongue) is manufactured by Shanghai Ruibin International Trade Co., Ltd.; the refractometer (PR-201α) is manufactured by ATAGO Co., Ltd. of Japan; the pH meter (FE20) is manufactured by Mettler Toledo Instruments (Shanghai) Co., Ltd.

[0082] In the following examples, the analytical methods used are as follows:

[0083] Principal Components Analysis (PCA) is a statistical method for observing the correlations among multiple variables. It can effectively reduce the dimensionality of the original data of the electronic tongue, extract characteristic data, and then use fewer variables to explain as much as possible the variance of the original data to be studied. The sensor response signals obtained are preprocessed using PCA to quickly and intuitively understand the distribution of the sample dataset under study.

[0084] Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) is a multivariate statistical analysis method mainly used for classification and feature selection, and is particularly useful in metabolomics and omics data analysis. OPLS-DA combines two techniques: Partial Least Squares Regression (PLSR) and Orthogonal Signal Correction (OSC), aiming to distinguish different groups of samples and identify the key variables affecting group classification.

[0085] Partial least squares regression (PLSR) is an efficient information extraction method. PLSR requires the establishment of two matrices: X (sensor response signal matrix) and Y (corresponding actual physical quantity, such as concentration). The goal is to find those components of the input matrix X that are as correlated as possible with the input variables, while achieving the best correlation with the target values in the Y matrix. For this reason, the PLSR algorithm is commonly used for data with trends, similar to predicting the value of an unknown point after establishing a standard curve.

[0086] The data obtained from the experiments in the following examples were statistically analyzed using Microsoft Office Excel 2021, multivariate data statistical analysis was performed using SIMCA-P 14.1, data analysis and modeling were carried out using The Unscrambler X 10.4, and plotting was done using Origin 2021.

[0087] Example 1 Detection Methods for Electronic Nose, Electronic Tongue, and Physicochemical Indicators

[0088] I. Sample Preparation

[0089] Take out the whole fruits from the -80°C refrigerator, select appropriate lychees for thawing under running water. When the lychees can just be peeled, peel and pit the lychees, then make a pulp, filter it with a 200-mesh gauze to obtain lychee juice for testing.

[0090] II. Electronic nose detection

[0091] The PEN3 type electronic nose sensor of German AIRSENSE company contains 10 metal oxide semiconductor chemical sensing elements. They have different sensitivities to different types of volatile substances, and the specific performance descriptions are shown in Table 1.

[0092] Table 1 PEN3 type e-nose sensor sequence and performance description

[0093]

[0094] Take 10 mL of the processed lychee juice and put it into a 30 mL sample injection bottle. Set the headspace enrichment time to 90 min, the cleaning time to 120 s, the zeroing time to 5 s, the sample preparation time to 5 s, the sample injection flow rate to 150 mL / min, and the detection time to 120 s. Each group of samples is tested in parallel three times, and air purging and standardization procedures are carried out after each test.

[0095] III. Electronic tongue detection

[0096] The electronic tongue system (Super Tongue) consists of four parts: a sensor array, a signal conditioning system, a test platform, and application software. The sensor array consists of six working electrodes and an auxiliary electrode to form an independent unit mechanism, which together with a reference electrode forms a complete sensor array; the signal conditioning system consists of a signal excitation unit, a signal conditioning unit, and a data acquisition unit.

[0097] Before the experiment, activate the electronic tongue sensor with 0.01 mol / L KCL solution to stabilize the measurement signal. Then, wash the sensor with deionized water for 10 s, take 10 mL of lychee juice and put it into a 50 mL special beaker for the electronic tongue, add 20 mL of water for dilution, set the sensor amplification factor to 100, and conduct the detection.

[0098] IV. Detection methods for physical and chemical indicators

[0099] (1) Detection of sugar-acid ratio

[0100] The sugar-acid ratio is the ratio of soluble solids to titratable acid.

[0101] Soluble solids: Measured at room temperature using a hand-held refractometer, expressed as TSS (°Brix).

[0102] Titratable acid: Refer to the national standard GB 12456-2021 "Determination of Total Acid in Foods". Use a pipette to suck 50 mL of litchi juice and place it in a 150 mL conical flask. Turn on the power of the pH meter, and calibrate the pH meter after it stabilizes. Press the pH reading switch, shake the bottle body, and quickly titrate with 0.1 mol / L sodium hydroxide standard titrant. Observe the change of the solution pH at any time. When approaching the titration end point, slow down the titration speed. Until the pH of the solution reaches 8.2, record the value of the volume of the sodium hydroxide standard titrant consumed, and take 3 parallel measurements for calculation.

[0103] (2) Detection of sugar components

[0104] The content of glucose, fructose, and sucrose in litchi is determined by the high performance liquid chromatography method in the national standard (GB 5009.8-2016). Weigh 2 g of litchi pulp into a 100 mL volumetric flask, dissolve it with 50 mL of water, slowly add 5 mL of zinc acetate solution and 5 mL of potassium ferrocyanide solution respectively, then add water to the scale for volume fixation, ultrasonicate for 30 min, centrifuge to obtain the supernatant, and filter it through a 0.45 μm microporous filter membrane into a sample bottle. Use a differential detector for determination. The temperature of the differential detector is 40 °C, the chromatographic column is an amino chromatographic column, the column length is 250 mm, the inner diameter is 4.6 mm, the film thickness is 5 μm, the mobile phase is acetonitrile: water = 7:3, the flow rate is 1 mL / min, the column temperature is 40 °C, and the injection volume is 20 μL. Each sample is measured 3 times in parallel, and the content of glucose, fructose, and sucrose is calculated according to the standard curve.

[0105] (3) Detection of organic acids

[0106] Refer to the method of the national standard (GB5009.157-2016). Accurately pipette 5.00 mL of litchi juice, add 0.2 mL of 1 mol / L phosphoric acid, dilute it to 10 mL with distilled water, centrifuge at 10000 r / min for 10 min, and filter the sample solution through a 0.45 μm filter membrane. Chromatographic conditions: Mobile phase: 0.1% phosphoric acid solution, 100% methanol; Flow rate: 1.0 mL / min; Injection volume: 20 uL; Ultraviolet detection wavelength: 210 nm.

[0107] (4) Volatile components

[0108] Extraction: Use the headspace solid-phase microextraction (HS-SPME) method. Age the solid-phase microextraction head in the gas chromatograph injection port. The aging temperature is 240 °C, the aging time is 1 h, and the carrier gas flow rate is 1.00 mL / min. Accurately transfer 5 mL of the sample into a 20 mL headspace vial, add 1 g of NaCl to promote the volatilization of volatile compounds, add 10 μL of the internal standard cyclohexanone (0.05 mg / mL), seal it, and equilibrate for 10 min. Insert the SPME extraction head 1 cm into the conical flask, push out the fiber head, adsorb the headspace at 40 °C in a metal bath for 30 min, desorb at 240 °C in the gas chromatograph injection port for 5 min, and perform GC-MS analysis.

[0109] GC conditions: DB-WAX capillary column (30 m × 0.25 mm × 0.25 μm); injection port temperature 240 °C, carrier gas is He (purity 99.99%), flow rate 1.00 mL / min, splitless injection. Temperature program: initial temperature 40 °C, hold for 3 min, heat at a rate of 5 °C / min to 170 °C, and then heat at a rate of 8 °C / min to 230 °C. MS conditions: electron impact ionization (EI), ion source temperature 230 °C, transfer line temperature 240 °C, mass scan range 35 - 500 m / z.

[0110] Qualitative analysis: retrieved from the NIST spectral library, and the aroma components were qualitatively analyzed using retention time and matching degree (≥85%). Quantitative analysis: calculated based on the peak area (assuming the absolute correction factor of each volatile compound is 1.0), and the calculation formula is as follows:

[0111]

[0112] In the formula: C x —— content of volatile compound, μg / mL; C1—— mass concentration of internal standard, μg / mL; V1—— volume of internal standard, mL; S x —— peak area of volatile compound; S1—— peak area of internal standard; V0—— volume of sample, mL.

[0113] Analysis of the determination results of electronic tongue in Example 2

[0114] Six kinds of litchis, namely Xiantao Lychee, Jumieren, Qiangang Glutinous Rice Creeper, Jinggang Red Glutinous, Seedless Lychee and Huaizhi, were detected by electronic tongue according to the method in Example 1, and the response value curves of the six kinds of litchis were obtained, as shown in Figures A - F of Figure 1 Furthermore, the principal component analysis was performed on the measured electronic tongue response values. The PCA analysis results based on the electronic tongue response signals are shown in Figure 2 As shown, the contribution rates of the first two principal components to the variance are 69.2% and 23.7% respectively, and a total of 92.9% of the effective information of the litchi juice samples can be explained, indicating that different varieties of litchis can be distinguished on the whole. P1 - 18 are the 18 sensors of the electronic tongue sensor. After extracting the response values, when the electronic tongue response values of the litchi of unknown variety match the corresponding characteristic matrix in Table 2, it can be determined as the corresponding variety.

[0115] Table 2 Electronic tongue response characteristic matrix

[0116]

[0117]

[0118] Analysis of the determination results of electronic nose in Example 3

[0119] Six kinds of litchi, namely Xiantao Litchi, Jumeiren Litchi, Qiangang Glutinous Rice Cake Litchi, Jinggang Red Glutinous Litchi, Seedless Litchi and Huaizhi Litchi, were detected by an electronic nose according to the method in Example 1. The radar charts of the electronic nose analysis of the six kinds of litchi are as shown in Figure 3 Figure A. Further, the measured electronic nose response values were subjected to principal component analysis. The PCA analysis results based on the electronic nose response signals are as shown in Figure 3 Figure B. The contribution rates of the first two principal components to the variance were 81.2% and 10.2% respectively, and a total of 91.4% of the effective information of the litchi juice samples could be explained, indicating that different varieties of litchi could be distinguished as a whole.

[0120] When the response value of the electronic nose W1C sensor is 3.7442, the response value of the W5S sensor is 2.7192, the response value of the W3C sensor is 0.9574, the response value of the W6S sensor is 1.1327, the response value of the W5C sensor is 3.0237, the response value of the W1S sensor is 5.4226, the response value of the W1W sensor is 2.3113, the response value of the W2S sensor is 7.8436, the response value of the W2W sensor is 1.8830, and the response value of the W3S sensor is 1.7003, it can be determined that the litchi variety is Seedless Litchi.

[0121] When the response value of the electronic nose W1C sensor is 2.2714, the response value of the W5S sensor is 2.6564, the response value of the W3C sensor is 0.9719, the response value of the W6S sensor is 1.0889, the response value of the W5C sensor is 1.8641, the response value of the W1S sensor is 2.7466, the response value of the W1W sensor is 2.2153, the response value of the W2S sensor is 4.5045, the response value of the W2W sensor is 1.7509, and the response value of the W3S sensor is 1.5140, it can be determined that the litchi variety is Jumeiren Litchi.

[0122] When the response value of the electronic nose W1C sensor is 4.0081, the response value of the W5S sensor is 2.5988, the response value of the W3C sensor is 0.9564, the response value of the W6S sensor is 1.1366, the response value of the W5C sensor is 3.2080, the response value of the W1S sensor is 5.6324, the response value of the W1W sensor is 2.3368, the response value of the W2S sensor is 10.1022, the response value of the W2W sensor is 1.8058, and the response value of the W3S sensor is 1.7230, it can be determined that the litchi variety is Xiantao Litchi.

[0123] When the response value of the W1C sensor of the electronic nose is 6.4114, the response value of the W5S sensor is 23.0024, the response value of the W3C sensor is 0.9044, the response value of the W6S sensor is 1.1878, the response value of the W5C sensor is 5.1302, the response value of the W1S sensor is 10.4905, the response value of the W1W sensor is 6.6320, the response value of the W2S sensor is 9.6586, and the response value of the W2W sensor is 4.1920, and the response value of the W3S sensor is 2.7138, it can be determined that the litchi variety is Huaizhi.

[0124] When the response value of the W1C sensor of the electronic nose is 5.2575, the response value of the W5S sensor is 19.3050, the response value of the W3C sensor is 0.9500, the response value of the W6S sensor is 1.1955, the response value of the W5C sensor is 4.0732, the response value of the W1S sensor is 8.7844, the response value of the W1W sensor is 6.0428, the response value of the W2S sensor is 12.2634, the response value of the W2W sensor is 4.3162, and the response value of the W3S sensor is 1.8348, it can be determined that the litchi variety is Qiangang Nuomici.

[0125] When the response value of the W1C sensor of the electronic nose is 4.9887, the response value of the W5S sensor is 28.4475, the response value of the W3C sensor is 0.9594, the response value of the W6S sensor is 1.1899, the response value of the W5C sensor is 3.9831, the response value of the W1S sensor is 7.4431, the response value of the W1W sensor is 6.0786, the response value of the W2S sensor is 13.3491, the response value of the W2W sensor is 3.4733, and the response value of the W3S sensor is 1.8192, it can be determined that the litchi variety is Jinggang Hongnuo.

[0126] Example 4 Analysis results of different varieties of litchi by electronic nose combined with electronic tongue

[0127] Combined with the response signals of the electronic nose and the electronic tongue for analysis, the principal component analysis results of different varieties of litchi based on the combined signals are as Figure 4 shown. The variance contribution rates of the first and second principal components are 69.6% and 13.4% respectively, and the total contribution rate is 83%, indicating that the combined signals of the electronic nose and the electronic tongue can well distinguish different varieties of litchi. From Figure 4 it can be seen that all varieties of litchi can be distinguished, without overlapping parts, and the distinguishing effect is better than the PCA results using only the sensor signals of the electronic nose or the electronic tongue.

[0128] Example 5 Determination of physical and chemical indexes of different varieties of litchi

[0129] The physicochemical indexes of different litchi varieties (Xiantao Litchi, Giant Beauty, Qiangang Glutinous Rice Creeper, Jinggang Red Glutinous Rice, Seedless Litchi and Huaizhi) were detected by the method in Example 1, including sugar-acid ratio, total sugar, total acid and characteristic volatile components. The measurement results are shown in Table 3. The results show that:

[0130] The sugar-acid ratio is the ratio of soluble solids (TSS) to titratable acid (TA). The sugar-acid ratios of the 6 litchi varieties range from 94.99 to 180.69. The total sugar content in the table is the sum of the contents of fructose, glucose and sucrose. The total sugar contents of the 6 litchi varieties range from 13.66 to 21.22 g / 100 g. The total acid content is the sum of the contents of tartaric acid, malic acid and citric acid. The total acid contents of the 6 litchi varieties range from 204.60 to 557.36 mg / 100 mL. Among the 6 litchi varieties, 231 volatile substances were detected by HS-GC-MS. By performing OPLS-DA analysis and using VIP>1 and a statistically significant two-tailed t-test (P<0.05) as the criteria, a total of 31 characteristic volatile components were screened out. The contents of the characteristic volatile components of the 6 litchi varieties range from 11.51 to 66.58 μg / mL.

[0131] Table 3 Contents of sugar-acid ratio, total sugar, total acid and characteristic volatile components of different litchi varieties

[0132]

[0133] Example 6 Establishment of the PLSR model between sensor response signals and physicochemical indexes

[0134] Based on the sensor response values of the electronic nose and electronic tongue and the contents of the physicochemical indexes obtained in Examples 1-4, a PLSR model was established. In the partial least squares regression (PLSR) model, the correlation coefficient and RMSEP are two important indexes for evaluating the performance and prediction ability of the model. The closer the correlation coefficient is to 1 and the smaller the RMSEP value, the more reliable the model is.

[0135] In the PLSR model of the sugar-acid ratio ( Figure 5 ), the model prediction correlation coefficient using only the sensor response results of the electronic tongue is 0.5819, and the RMSEP value is 17.9407; the model prediction correlation coefficient based on the response values of the electronic nose and electronic tongue is 0.7559, and the RMSEP value is 11.3681.

[0136] In the PLSR model of the total sugar ( Figure 6 ), the model prediction correlation coefficient using only the sensor response results of the electronic tongue is 0.5698, and the RMSEP value is 1.9554; the model prediction correlation coefficient based on the response values of the electronic nose and electronic tongue is 0.7043, and the RMSEP value is 1.7559.

[0137] In the PLSR model of total acid ( Figure 7 ), the model prediction correlation coefficient using only the response results of the electronic tongue sensor is 0.6066, and the RMSEP value is 83.4741; the model prediction correlation coefficient based on the response values of the electronic nose and the electronic tongue is 0.7575, and the RMSEP value is 58.7662.

[0138] In the PLSR model of characteristic volatile components ( Figure 8 ), the model prediction correlation coefficient using only the response results of the electronic nose sensor is 0.8598, and the RMSEP value is 8.9641; the model prediction correlation coefficient based on the response values of the electronic nose and the electronic tongue is 0.8821, and the RMSEP value is 7.9846.

[0139] The above results show that the reliability of the models combining the sensor signals of the electronic nose and the electronic tongue is better than that of using only the electronic nose or the electronic tongue alone. The model prediction correlation coefficients established by combining the signals are all greater than 0.70, indicating that the models can perform effective quantitative characterization.

[0140] Example 7

[0141] The litchi variety selected for the test materials in this example is 'Feizixiao', which comes from three places in Hainan Province, Yunnan Province and Guangdong Province respectively. The Feizixiao litchi from different origins was subjected to electronic nose detection, electronic tongue detection and physical and chemical index detection, and the detection methods were the same as those in Example 1.

[0142] The results of principal component analysis of Feizixiao litchi from different origins based on the sensor response signals of both the electronic nose and the electronic tongue are as Figure 9 shown. The results show that the discrimination effect of the combined signals is similar to that in Example 3. Among them, the variance contribution rates of the first and second principal components are 61.4% and 23.8% respectively, and the total contribution rate is 85.2%. There is no overlap among the Feizixiao litchi from Hainan, Yunnan and Guangdong, proving that the combination of the electronic nose and the electronic tongue can distinguish litchi from different origins. When the detection results of the electronic nose and the electronic tongue of the Feizixiao litchi with unknown origin match the sensor response characteristic matrix in Table 4, it can be determined as the litchi of the corresponding origin.

[0143] Table 4 Sensor response characteristic matrix

[0144]

[0145]

[0146] The detected physical and chemical indexes were analyzed by partial least squares regression method, and a PLSR model was established to quantitatively characterize the sugar-acid ratio, total acid (tartaric acid, malic acid, citric acid), total sugar (fructose, glucose, sucrose) and the content of characteristic volatile substances of Feizixiao litchi from different origins.

[0147] Figure 10 For the evaluation results of the sugar-acid ratio PLSR model, the correlation coefficient of the training set is 0.9366, and the RMSEP value is 1.9139; the correlation coefficient of the prediction set is 0.7760, and the RMSEP value is 5.4200.

[0148] Figure 11 For the evaluation results of the total sugar content PLSR model, the prediction correlation coefficient of the training set is 0.8262, and the RMSEP value is 0.3498; the correlation coefficient of the prediction set is 0.7568, and the RMSEP value is 0.2767.

[0149] Figure 12 For the evaluation results of the total acid content PLSR model, the prediction correlation coefficient of the training set is 0.9380, and the RMSEP value is 8.8154; the correlation coefficient of the prediction set is 0.9554, and the RMSEP value is 7.3608.

[0150] Figure 13 For the evaluation results of the characteristic volatile content PLSR model, the prediction correlation coefficient of the training set is 0.9170, and the RMSEP value is 3.4979; the correlation coefficient of the prediction set is 0.9437, and the RMSEP value is 2.7772.

[0151] The above results show that the PLSR model established based on the combined signals of electronic nose and electronic tongue can effectively quantify and characterize the sugar-acid ratio, total sugar, total acid, and characteristic volatile substance content of different litchis.

[0152] The quantitative prediction model obtained through the above calculation and verification is:

[0153] Sugar-acid ratio content = -4.1730×W1C - 0.9771×W5S + 0.3123×W3C + 1.8554×W6S - 4.1367×W5C + 13.2519×W1S + 6.2793×W1W - 2.007×W2S - 10.0370×W2S + 14.9546×W3S + 5.1563×P1 + 5.1863×P2 - 0.0409×P3 + 7.1273×P4 + 7.0794×P5 + 0.2125×P6 + 5.1164×P7 + 5.0956×P8 - 0.3390×P9 + 4.7192×P10 + 4.3903×P11 + 2.5084×P12 + 5.3195×P13 + 4.4933×P14 + 1.5606×P15 + 6.2368×P16 + 5.6657×P17 + 1.8505×P18.

[0154] Total sugar content = 1.9418×W1C + 0.2468×W5S - 0.0410×W3C - 0.1967×W6S + 1.1886×W5C - 1.9771×W1S - 1.5198×W1W - 0.0262×W2S + 1.2060×W2W - 0.3461×W3S - 0.1383×P1 - 0.1427×P2 + 0.0108×P3 - 0.2469×P4 - 0.2464×P5 - 0.0361×P6 - 0.1048×P7 - 0.1055×P8 + 0.0048×P9 - 0.1180×P10 - 0.0911×P11 - 0.1456×P12 - 0.1061×P13 - 0.0800×P14 - 0.0743×P15 - 0.2130×P16 - 0.2089×P17 - 0.1341×P18。

[0155] Total acid content = -0.9117×W1C + 1.5049×W5S + 0.0001×W3C - 0.0646×W6S - 0.7493×W5C - 1.0279×W1S + 0.1831×W1W - 3.2316×W2S + 0.0821×W2W - 0.2200×W3S - 0.0038×P1 - 0.0038×P2 + 0.0004×P3 + 0.0080×P4 + 0.0073×P5 - 0.0002×P6 - 0.0138×P7 - 0.0136×P8 - 0.0018×P9 + 0.0088×P10 + 0.0091×P11 + 0.0026×P12 + 0.0146×P13 + 0.0129×P14 - 0.0047×P15 + 0.0051×P16 + 0.0052×P17 - 0.0011×P18。

[0156] Content of characteristic volatile substances = 9.0383×W1C + 1.9907×W5S - 0.6472×W3C - 3.1953×W6S + 9.9741×W5C - 25.1167×W1S - 12.8103×W1W + 3.6136×W2S + 18.7819×W2W - 28.8936×W3S - 9.4119×P1 - 9.4882×P2 + 0.0723×P3 - 13.4475×P4 - 13.3468×P5 - 0.3774×P6 - 9.6659×P7 - 9.6275×P8 + 0.6293×P9 - 8.8184×P10 - 8.1545×P11 - 4.3461×P12 - 9.9008×P13 - 8.3071×P14 - 3.0778×P15 - 11.8370×P16 - 10.7503×P17 - 3.5783×P18。

[0157] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be equivalent replacement methods and are all included in the protection scope of the present invention.

Claims

1. A method for identifying litchi varieties or litchi origins based on electronic nose and electronic tongue, characterized in that: The steps include: S1. The electronic tongue detects litchi samples of different varieties or origins to obtain the response value data of the electronic tongue sensor; the electronic nose detects volatile gases of litchi samples of different varieties or origins to obtain the response value data of the electronic nose sensor; S2. performing principal component analysis using the sensor response value data of the electronic nose and electronic tongue obtained in step S1; S3. Take the litchi sample to be tested for electronic tongue detection and electronic nose detection to obtain the response value data of the litchi sample to be tested, and compare the response value of the litchi sample to be tested with the response value of the litchi sample of different varieties or origin in step S2 to determine the variety or origin of the sample to be tested; The electronic nose is a PEN3 type electronic nose; The electrodes of the electronic tongue include one or more of platinum, gold, palladium, tungsten, titanium, and silver.

2. The method according to claim 1, characterized in that: The sensors of the electronic nose include W1C, W5S, W3C, W6S, W5C, W1S, W1W, W2S, W2W and W3S.

3. The method according to claim 1 or 2, characterized in that: The electronic nose detection conditions are as follows: cleaning time is 90-150s, zeroing time is 2.5-10s, sample preparation time is 5-20s, injection flow rate is 100-200mL / min, and detection time is 60-150s.

4. The method according to claim 1, characterized in that: The sensor amplification factor of the electronic tongue is set to 80-120.

5. The method according to claim 1, characterized in that: The response value is the average value of the response at 89-91s.

6. The method according to claim 1, characterized in that: The litchi sample is litchi juice obtained by pulping litchi pulp and filtering it.

7. The method according to claim 1, characterized in that: The litchi varieties are Xiantao litchi, Jumeiren litchi, Qiangang Niumici litchi, Jingganghongnuo litchi, Wuzili litchi and Huaizhi litchi.

8. The method according to claim 1, characterized in that: The lychees from different origins are Feizixiao lychee varieties produced in different provinces, including Feizixiao in Hainan Province, Feizixiao in Yunnan Province and Feizixiao in Guangdong Province.

9. A method for rapidly detecting lychee flavor-related physical and chemical indices based on electronic nose and electronic tongue, characterized in that: The steps include: (1) using an electronic tongue to detect litchi samples and obtaining response value data of the electronic tongue sensor; using an electronic nose to detect volatile gases of litchi samples and obtaining response value data of the electronic nose sensor; and determining the content of flavor-related physical and chemical indicators of litchi samples; (2) Using the response values ​​and the content data of flavor-related physical and chemical indices, a partial least squares regression model of the response values ​​and the content of flavor-related physical and chemical indices was established; (3) taking the litchi sample to be tested and performing electronic nose and electronic tongue detection to obtain a response value of the litchi sample to be tested, and substituting the response value into the model of step (2) to obtain the flavor-related physical and chemical index content of the litchi sample to be tested; The flavor-related physical and chemical indicators are sugar-acid ratio, sugar components, organic acids and characteristic volatile substances.

10. The method according to claim 9, characterized in that: The characteristic volatile substances are: characteristic volatile substances obtained by measuring the volatile components of the sample using gas chromatography-mass spectrometry and screening using orthogonal partial least squares discriminant analysis.

Images