Method for predicting organic acid in pitaya by using electronic nose

Abstract

The invention discloses a method for predicting organic acid in dragon fruits by using an electronic nose, and belongs to the technical field of food classification, the method comprises the following steps: 1, selecting a calibration set sample and a prediction set sample dragon fruit for processing, and respectively obtaining dragon fruit pulp; step 2, obtaining volatile component characteristics of pitaya pulp through a sensor, taking residual calibration set samples, and detecting metabolites in the pitaya pulp; step 3, determining flavor characteristics of different varieties of dragon fruits, establishing a dragon fruit variety classification calibration model, and establishing a regression calibration model for predicting the organic acid of the dragon fruits according to the obtained organic acid; and 4, predicting the organic acid contained in the pitaya pulp. According to the method, the characteristics of the volatile components of the pitaya pulp are obtained by testing different sensitivities of the three different varieties of pitaya to various sensors of the electronic nose, so that the variety classification of the pitaya is carried out, the organic acid in the pitaya pulp is predicted, and the accuracy and efficiency of the classification of the variety of the pitaya are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of food classification, and particularly relates to a method for predicting organic acids in pitaya using an electronic nose. Background Art

[0002] Pitaya originated from the tropical regions of Central America. Due to its adaptability to tropical and subtropical climate conditions, it is widely planted in many suitable regions around the world, and the industrial scale continues to expand. Pitaya is rich in citric acid, gamma-aminobutyric acid, various minerals and unique phytochemicals, with high nutritional value. Its pulp is sweet, with a delicate or crispy texture, and has both medicinal and ornamental values. The market demand has increased significantly, making it a core category in the tropical fruit industry. The contents of citric acid and gamma-aminobutyric acid in pitaya pulp are key factors shaping its variety characteristics. The differences in the contents of citric acid and gamma-aminobutyric acid in different varieties of pitaya have an important impact on flavor and quality. Red-skinned and red-fleshed pitaya has a relatively high citric acid content, with a strong flavor and antioxidant properties; red-skinned and white-fleshed varieties have a moderate citric acid content, balancing the sugar-acid ratio; while yellow-skinned and white-fleshed varieties have a relatively low citric acid content, with a sweet taste and affect physiological regulation. Gamma-aminobutyric acid enhances stress resistance in red-skinned and red-fleshed varieties, maintains metabolic advantages in red-skinned and white-fleshed varieties, and improves adaptability in yellow-skinned and white-fleshed varieties, promoting the diversification of pitaya quality and industrial upgrading.

[0003] Currently, the differentiation of pitaya varieties mainly relies on manual visual inspection and simple sensory tests, resulting in inconsistent classification criteria and low accuracy. Existing technologies mostly rely on manual evaluation and physical and chemical analysis methods, which have problems such as cumbersome detection processes, long time consumption, and high costs. In addition, the quality traceability system for pitaya is not yet perfect, lacking effective standardized management and variety traceability technologies. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for predicting organic acids in pitaya using an electronic nose to solve the problems faced in the above background art.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] A method for predicting organic acids in pitaya using an electronic nose, the method comprising:

[0007] Step 1: Select pitaya samples for the calibration set and the prediction set. The calibration set samples include red-skinned and red-fleshed pitaya, red-skinned and white-fleshed pitaya, and yellow-skinned and white-fleshed pitaya. Take part of the calibration set samples and all the prediction set samples, wash them clean with tap water, peel them, and take 20 g of pulp in a low-temperature environment and place it in a 50 ml centrifuge tube as the calibration pitaya pulp and the prediction pitaya pulp, and perform enrichment;

[0008] Step 2: Take the calibrated pitaya pulp and test it using an electronic nose sensor. Obtain the volatile component characteristics of the pitaya pulp by taking advantage of the different sensitivities of three different varieties of pitaya to different sensors. Take the remaining samples in the calibration set and detect the metabolites in the pitaya pulp using a high-resolution mass spectrometer;

[0009] Step 3: Use the volatile component characteristics of the pitaya pulp obtained in Step 2 to determine the flavor characteristics of different varieties of pitaya, thereby establishing a calibration model for pitaya variety classification. Then, detect the organic acids affecting the pitaya variety differences from the metabolites in the pulp through a high-resolution mass spectrometer, and establish a regression calibration model for predicting pitaya organic acids through a multiple linear regression model and a stepwise regression algorithm;

[0010] Step 4: Take the predicted pitaya pulp, measure its flavor characteristics using an electronic nose, and substitute it into the regression calibration model established in Step 3 to predict the organic acids contained in the pitaya pulp.

[0011] Furthermore, the method for obtaining the pitaya for the calibration set samples and the prediction set samples in Step 1 is as follows:

[0012] Select pitayas of red-skin and red-flesh pitaya, red-skin and white-flesh pitaya, and yellow-skin and white-flesh pitaya. Use an electronic balance to screen out a certain number of pitayas of different types with the same weight. Then, select nine fresh, mature, and disease-free pitayas from each type of pitaya. Part of them are used as calibration set samples, and the other part are used as prediction set samples.

[0013] Furthermore, the specific method for enrichment in Step 1 is as follows:

[0014] Transfer the selected pitayas to a cleaning tank, wash them with running tap water. After washing, manually peel them, use a high-temperature sterilized stainless steel medicine spoon to scoop out 20 g of pulp and place it in a 50 g centrifuge tube. Seal the centrifuge tube and place it in a normal temperature environment, and let it stand for 30 minutes for enrichment.

[0015] Furthermore, the method for obtaining the volatile component characteristics of the pitaya pulp in Step 2 is as follows:

[0016] S21: Remove the sealing film of the centrifuge tube, and directly insert the sampling needle of the electronic nose device into the sealed centrifuge tube containing the sample;

[0017] S22: Adopt the direct headspace aspiration method to detect the volatile components in the sample through the electronic nose. Among them, the gas in the sample is sampled through the intake pipe of the electronic nose at a set flow rate of 400 mL·min - 1 for gas sampling of the sample, with a sampling time of 1 second / group. When injecting the sample, the electronic nose device automatically performs self-cleaning of the sensor for 100 seconds.

[0018] Further, the method for establishing the calibration model for pitaya variety classification in Step 3 is as follows:

[0019] After obtaining the flavor characteristics of pitayas of different varieties, use R language to perform a multivariate classification model. Set the flavor characteristics of the pitaya pulp as the X variable and the pitaya variety as the Y variable, so as to establish the calibration model for pitaya variety classification.

[0020] Further, the method for establishing the regression calibration model for predicting organic acids in pitayas in Step 3 is as follows:

[0021] Perform metabolome detection on the calibrated pitaya samples after being detected by the electronic nose to obtain organic acids in the pitayas.

[0022] According to the data of each variety obtained by the electronic nose, use R language to perform a multiple linear regression model and a stepwise regression algorithm. Set the electronic nose as the X variable and the organic acid as the Y variable, and use the stepwise regression algorithm to establish the independent variables selected for the model, so as to establish the regression calibration model for predicting organic acids in pitayas.

[0023] Further, the method for predicting the organic acids contained in the pitaya pulp in Step 4 is as follows:

[0024] Take the predicted pitaya samples, collect relevant data using the electronic nose, and input them into the established regression calibration model for predicting organic acids in pitayas, so as to predict the citric acid and gamma-aminobutyric acid that affect the pitaya variety in the pitaya pulp, and establish the regression calibration model between pitaya citric acid and the electronic nose, and the regression calibration model between pitaya gamma-aminobutyric acid and the electronic nose.

[0025] Take the predicted pitaya samples, obtain the flavor characteristics of the pitayas using the electronic nose, and input them into the established regression calibration model between pitaya citric acid and the electronic nose, and the regression calibration model between pitaya gamma-aminobutyric acid and the electronic nose, so as to predict the amounts of citric acid and gamma-aminobutyric acid in the pitaya pulp.

[0026] Advantages of the present invention:

[0027] The present invention uses the electronic nose detection technology to collect the volatile component characteristics of the pitaya pulp through a non-destructive technology, realizes the rapid differentiation of different pitaya varieties, and simultaneously predicts the organic acids in the pitayas; through data processing and model construction, uses the electronic nose to realize the method for pitaya variety classification and organic acid prediction, achieving the rapid non-destructive detection of different varieties and the organic acids in the pulp.

[0028] The present invention utilizes an electronic nose sensor to obtain the volatile component characteristics of pitaya pulp by testing the sensitivities of various sensors of the electronic nose to three different varieties of pitayas, thereby classifying the pitaya varieties and predicting the organic acids in the pitaya pulp. This method not only improves the accuracy and efficiency of pitaya variety classification but also provides a rapid and non-destructive detection means for predicting the organic acid content in pitayas, contributing to optimizing the planting, processing, and quality control processes of pitayas.

[0029] Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for describing the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

[0031] Figure 1 is a flowchart of the method of the present invention;

[0032] Figure 2 is a graph of the results of multiple linear regression analysis for pitaya variety classification and organic acid content prediction in the present invention;

[0033] Figure 3 is a regression calibration model graph of the response of the W1S sensor and the citric acid content in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0034] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, rather than all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0035] A method for predicting organic acids in pitayas using an electronic nose, as Figure 1 shown, the method mainly includes:

[0036] Step 1: Select calibration set samples and prediction set samples of pitayas. The calibration set samples include red-skinned and red-fleshed pitayas, red-skinned and white-fleshed pitayas, and yellow-skinned and white-fleshed pitayas. Take some calibration set samples and all prediction set samples, wash them clean with tap water, peel them, and take 20 g of pulp in a low-temperature environment and place it in a 50-ml centrifuge tube as calibration pitaya pulp and prediction pitaya pulp, and then conduct enrichment. The methods for obtaining the calibration set samples and prediction set samples of pitayas are as follows: Select pitayas of red-skinned and red-fleshed pitayas, red-skinned and white-fleshed pitayas, and yellow-skinned and white-fleshed pitayas, and use an electronic balance to screen out a certain number of pitayas of different varieties with the same weight. Then, select nine fresh, mature, and disease-free pitayas from each variety of pitayas. One part is used as calibration set samples, and the other part is used as prediction set samples. The specific method for enrichment is as follows: Transfer the selected pitayas to a cleaning tank, wash them with flowing tap water, manually peel them after washing, use a high-temperature sterilized stainless-steel medicine spoon to scoop out 20 g of pulp and place it in a 50-g centrifuge tube, seal the centrifuge tube, and place the centrifuge tube in a normal-temperature environment and let it stand for 30 minutes for enrichment. Use sealing film crazily. The sealing film must ensure good airtightness to prevent the leakage of volatile substances in the sample or interference with the external environment.

[0037] Step 2: Take the calibration pitaya pulp and use an electronic nose sensor to test it. Obtain the volatile component characteristics of the pitaya pulp by the different sensitivities of three different varieties of pitayas to different sensors. Take the remaining calibration set samples and use a high-resolution mass spectrometer to detect the metabolites in the pitaya pulp. The method for obtaining the volatile component characteristics of the pitaya pulp is as follows: First, remove the sealing film of the centrifuge tube, and directly insert the sampling needle of the electronic nose device into the sealed centrifuge tube containing the sample. The electronic nose device can be a German PEN3 electronic nose sensor, etc. Ensure good contact between the sampling needle and the sample during insertion without air leakage. Then, use the direct headspace aspiration method to detect the volatile components in the sample through the electronic nose. Among them, the gas in the sample is sampled through the intake pipe of the electronic nose at a set flow rate of 400 mL·min - 1 Sample gas sampling is carried out, and the sampling time is 1 second / group. During sample injection, the electronic nose device automatically performs self-cleaning of the sensor for 100 seconds to ensure the same sensor state for each measurement. During the measurement process, the volatile components of the sample will be detected and converted into electrical signals within 100 seconds of sampling and analysis time. Finally, take the data from 69 to 71 seconds for result summary and analysis. During this data period, the signals captured by the system are the most representative and can effectively reflect the volatile component characteristics of the sample.

[0038] Step 3. Use the volatile component characteristics of pitaya pulp obtained in Step 2 to determine the flavor characteristics of different varieties of pitaya, thereby establishing a calibration model for pitaya variety classification. Then, detect the metabolites in the pulp by a high-resolution mass spectrometer to obtain the organic acids that affect the variety differences of pitaya, and establish a regression calibration model for predicting pitaya organic acids through a multiple linear regression model and a stepwise regression algorithm; the method for establishing the pitaya variety classification calibration model is as follows: after obtaining the flavor characteristics of different varieties of pitaya, use the R language for a multiple classification model, set the flavor characteristics of pitaya pulp as the X variable, and set the pitaya variety as the Y variable, thereby establishing a calibration model for pitaya variety classification. For 3 different types of pitaya, each type is measured 3 times and a classification model is established based on the data. Predict the pitaya variety according to the distribution of the samples in the sketch of the LDA model; the method for establishing the regression calibration model for predicting pitaya organic acids is as follows: perform metabolome detection on the calibrated pitaya samples after being detected by the electronic nose to obtain organic acids in the pitaya. Then, according to the data of each variety of pitaya obtained by the electronic nose, use the R language for a multiple linear regression model and a stepwise regression algorithm, and set the electronic nose sensor W1S as the X variable and the organic acid as the Y variable. Adopt the stepwise regression algorithm to establish the independent variables selected for the model, thereby establishing a regression calibration model for predicting pitaya organic acids.

[0039] Step 4. Take the predicted pitaya pulp, use the electronic nose to measure its flavor characteristics, and substitute it into the regression calibration model established in Step 3 to predict the organic acids contained in the pitaya pulp. Specifically: take the predicted pitaya samples, use the electronic nose to collect relevant data, and substitute it into the established regression calibration model for predicting pitaya organic acids, thereby predicting that the citric acid and gamma-aminobutyric acid that affect the pitaya variety are contained in the pitaya pulp, and establish a regression calibration model between pitaya citric acid and the electronic nose, and a regression calibration model between pitaya gamma-aminobutyric acid and the electronic nose. Then, take the predicted pitaya samples, use the electronic nose to obtain the flavor characteristics of the pitaya, and substitute it into the established regression calibration model between pitaya citric acid and the electronic nose, and the regression calibration model between pitaya gamma-aminobutyric acid and the electronic nose, thereby predicting the amounts of citric acid and gamma-aminobutyric acid in the pitaya pulp. Specific Example 1:

[0041] A. Purchase a batch of red - skinned and red - fleshed pitayas (R), red - skinned and white - fleshed pitayas (W), and yellow - skinned and white - fleshed pitayas (Y) from Fruitday. Use an electronic balance to screen out 20 pitayas with the same weight, and then select 9 fresh, mature, disease - free, and uniformly sized pitayas from them. Among them, 9 are used as calibration set samples, and the other 9 are used as prediction set samples. Wash the selected 9 pitayas with tap water, manually peel them, and use a high - temperature - sterilized stainless - steel medicine spoon to scoop out 20 g of pulp in a low - temperature environment and place it in a 50 - mL centrifuge tube. Then seal the centrifuge tube, and the sealing film should ensure good airtightness. Place the centrifuge tube containing the pitaya samples in a normal - temperature environment and let it stand for 30 minutes for enrichment. 30 minutes after enrichment, gently remove the sealing film of each centrifuge tube, and directly insert the sampling needle of the electronic nose device into the sealed centrifuge tube containing the sample, ensuring good contact between the sampling needle and the sample without air leakage. Using the direct headspace aspiration method, detect the volatile components in the sample through the electronic nose. Through the intake pipe of the electronic nose, sample the gas of the sample at a set flow rate (400 mL·min -1 ) The gas sampling time is 1 second per group. When injecting the sample, the electronic nose device will automatically clean the sensor for 100 seconds to ensure that the sensor state is the same for each measurement. During the measurement process, the volatile components of the sample will be detected and converted into electrical signals within 100 seconds of sampling and analysis time, and the data from 69 - 71 seconds is taken. Obtain the electronic nose data of the calibration pitayas (as shown in Table 1);

[0042] Table 1 Electronic nose data of calibration pitayas

[0043]

[0044]

[0045] Among them, W1C, W5S, W3C, W6S, W5C, W1S, W1W, W2S, W2W, W3S are various types of sensors of the electronic nose.

[0046] B. Obtain a range summary table of the three varieties of pitayas measured by the electronic nose (as shown in Table 2); and establish a classification LDA model through the data. After the model training is completed, the evaluation by the confusion matrix shows that it reaches a high accuracy rate of 92.59% in 27 samples. Among them, the classification accuracy rate of the red - skinned and white - fleshed (W) variety reaches 100%, while 1 sample of each of the red - skinned and red - fleshed (R) and yellow - skinned and white - fleshed (H) varieties is misclassified. In addition, the two - dimensional projection map of the LDA model visually shows the distribution of the three varieties. LDA1 and LDA2 explain 97% and 2.9% of the data variation respectively, and the concentrated distribution of the sample points in their respective regions further confirms the effectiveness of the model.

[0047] Table 2 Summary of the Electronic Nose Range for Red-Fleshed Pitaya, White-Fleshed Pitaya, and Yellow-Fleshed Pitaya

[0048] sensor R W Y g 0.8159-1.0852 0.6948-0.854 0.9803-1.3738 W1C 1.0534-1.1881 1.1799-1.381 0.9875-1.1082 W5S 0.842-1.1881 0.7796-1.3407 0.9837-1.2978 W3C 0.8421-1.0807 0.7255-0.8687 0.985-1.2914 W6S 0.8598-1.0759 0.8004-1.0657 0.9849-1.2395 W5C 0.7537-1.5581 1.245-1.5581 0.7537-1.098 W1S 1.012-1.3242 1.2396-1.7545 0.9544-1.3598 W1W 1.0518-1.4352 1.1855-1.4352 0.8684-1.097 W2S 0.986-2.4352 1.0366-2.129 0.9945-1.0212 W2W 0.7557-1.5552 1.0366-1.5552 0.986-1.0934 W3S 0.7537-1.3598 1.0387-2.1192 0.8713-1.3647

[0049] C. The electronic nose data of the Hainan red-fleshed pitaya samples collected in step A were averaged to obtain 0.9558. Substituting this value into the summary range table established in step B, it can be seen that this sample value falls within the variety range of red-fleshed pitaya R. Then, after classifying the three calibrated pitayas detected by the electronic nose in step A in step B, a metabolomics test was carried out. Using a sterilized stainless-steel spatula, 50 mg of the sample was weighed and placed into a 2.0 mL EP tube. 500 μL of 80% ice methanol solution was added, and a small amount of steel beads were added and pulverized with a grinder. Before the metabolomics experiment, the centrifuge tube containing the pitaya sample was placed in a -20 °C refrigerator and allowed to stand for 30 min to precipitate the proteins in the sample. After centrifuging at 20000 g for 15 min, 400 μL of the supernatant was transferred to another EP tube. After centrifuging at 20000 g for 15 min, the supernatant was transferred into an injection vial for UPLC-HRMS detection. For each sample, 10 - 20 μL of the extract was taken in equal amounts and mixed into a QC sample for UPLC-HRMS detection. Through data collation, it was found that citric acid and γ-aminobutyric acid play key roles in pitaya varieties.

[0050] D. Using R language for multiple linear regression models and stepwise regression algorithms with the electronic nose data measured in step A and the citric acid measured in step C. The W1S channel of the electronic nose sensor was set as the X variable, and citric acid was set as the Y variable. A stepwise regression algorithm was used to establish the independent variables selected for the model. Finally, the regression calibration model for the response of the W1S sensor and the citric acid content was established as: Electronic nose of pitaya y = -4.1288E09x + 5.982E09 (as Figure 2 、 Figure 3 shown), the residual standard error was 5.734. After simple iteration, R 2 was 0.9313.

[0051] E. Substituting the W1S sensor response x = 1.0986 of the predicted Hainan red-fleshed pitaya sample collected by the electronic nose in step A into the regression calibration model established in step D, the citric acid content y of the pitaya was obtained as 1.4500027E09. From the regression model, the R 2 value was 0.2947, indicating that the prediction of citric acid in the pitaya was reliable.

[0052] The present invention utilizes an electronic nose sensor. By testing the different sensitivities of various sensors of the electronic nose to three different varieties of pitayas, the volatile component characteristics of pitaya pulp are obtained, so as to classify the pitaya varieties and predict the organic acids in the pitaya pulp. This method not only improves the accuracy and efficiency of pitaya variety classification, but also provides a rapid and non-destructive detection means for predicting the content of organic acids in pitayas, which helps to optimize the planting, processing and quality control processes of pitayas.

[0053] The above content is only an example and illustration of the concept of the present invention. Those skilled in the art of the present technology can make various modifications or supplements to the described specific embodiments or use similar ways to replace them. As long as they do not deviate from the concept of the invention or exceed the scope defined by this claims, they should all fall within the protection scope of the present invention.

Claims

1. A method for predicting organic acids in pitaya using an electronic nose, characterized in that: The method comprises: Step 1, select the calibration set sample and the prediction set sample pitaya, the calibration set samples include red skin and red flesh pitaya, red skin and white flesh pitaya and yellow skin and white flesh pitaya, take some calibration set samples and all prediction set samples, wash them with tap water and peel them, take 20g of pulp in a low temperature environment and put it in a 50ml centrifuge tube as calibration pitaya pulp and prediction pitaya pulp, and enrich them; Step 2: Take the calibrated pitaya pulp and test it using an electronic nose sensor to obtain the volatile component characteristics of the pitaya pulp by taking the different sensitivities of three different varieties of pitaya to different sensors, and take the remaining calibration set samples to detect metabolites in the pitaya pulp using a high-resolution mass spectrometer; Step 3, using the volatile component characteristics of the pitaya pulp obtained in step 2 to determine the flavor characteristics of different varieties of pitaya, thereby establishing a pitaya variety classification and calibration model, and then detecting metabolites in the pulp by a high-resolution mass spectrometer to obtain organic acids that affect the differences in pitaya varieties, and establishing a regression calibration model for predicting pitaya organic acids by a multiple linear regression model and a stepwise regression algorithm; Step 4: Take the predicted pitaya pulp, use the electronic nose to measure its flavor characteristics, and substitute it into the regression calibration model established in step 3 to predict the organic acids contained in the pitaya pulp.

2. A method for predicting organic acids in pitaya using an electronic nose according to claim 1, characterized in that: The method for obtaining the calibration set sample and the prediction set sample pitaya in step 1 is: Red-skinned and red-fleshed dragon fruit, red-skinned and white-fleshed dragon fruit, and yellow-skinned and white-fleshed dragon fruit were selected, and a certain number of different types of dragon fruits with consistent weight were screened out using an electronic balance. Then nine fresh, ripe, disease-free dragon fruits were selected from each type of dragon fruit, and some of them were used as calibration set samples, and the other part was used as prediction set samples.

3. A method for predicting organic acids in pitaya using an electronic nose according to claim 1, characterized in that: The specific method for enrichment in step 1 is: The selected pitaya was transferred to a cleaning tank, washed with running tap water, peeled manually, and 20 g of pulp was scooped out with a high-temperature sterilized stainless steel spoon and placed in a 50 g centrifuge tube. The centrifuge tube was sealed and placed in a room temperature environment for 30 minutes for enrichment.

4. A method for predicting organic acids in pitaya using an electronic nose according to claim 1, characterized in that: The method for obtaining the volatile component characteristics of pitaya pulp in the step 2 is: S21, removing the sealing film of the centrifuge tube, and using the injection needle of the electronic nose device to directly insert into the sealed centrifuge tube containing the sample; S22. Direct headspace aspiration was used to detect the volatile components in the sample through an electronic nose, where the air inlet of the electronic nose was set at a flow rate of 400 mL min -1 Gas sampling was performed on the samples with a sampling time of 1 second per group. During sampling, the electronic nose device automatically performed self-cleaning of the sensor for 100 seconds.

5. A method for predicting organic acids in pitaya using an electronic nose according to claim 1, characterized in that: The method for establishing the pitaya variety classification and calibration model in step 3 is: After obtaining the flavor characteristics of different varieties of pitaya, a multivariate classification model was developed using R language. The flavor characteristics of pitaya pulp were set as the X variable, and the pitaya variety was positioned as the Y variable, thus establishing a pitaya variety classification and calibration model.

6. A method for predicting organic acids in pitaya using an electronic nose according to claim 1, characterized in that: The method for establishing a regression calibration model for predicting organic acids in pitaya in the step 3 is: The calibrated pitaya samples that have been tested by the electronic nose are subjected to metabolomics testing to obtain organic acids in pitaya; According to the data of each variety obtained by the electronic nose, the multiple linear regression model and stepwise regression algorithm were performed using R language. The electronic nose was set as the X variable and the organic acid was set as the Y variable. The stepwise regression algorithm was used to establish the independent variables for screening the model, thus establishing a regression calibration model for predicting organic acid in pitaya.

7. A method for predicting organic acids in pitaya using an electronic nose according to claim 1, characterized in that: The method for predicting the organic acids contained in the dragon fruit pulp in the step 4 is: Take the predicted pitaya sample, use the electronic nose to collect relevant data, and bring it into the established regression calibration model for predicting organic acids in pitaya, so as to predict that the pitaya pulp contains citric acid and aminobutyric acid that affect the pitaya variety, and establish the regression calibration model of pitaya citric acid and electronic nose, and the regression calibration model of pitaya aminobutyric acid and electronic nose; Take the predicted pitaya sample, use the electronic nose to obtain the flavor characteristics of the pitaya, and bring it into the established regression calibration model of pitaya citric acid and electronic nose, and the regression calibration model of pitaya aminobutyric acid and electronic nose, so as to predict the amount of citric acid and aminobutyric acid in the pitaya pulp.

Images